[House Hearing, 107 Congress]
[From the U.S. Government Publishing Office]
ELECTION NIGHT COVERAGE BY THE NETWORKS
=======================================================================
HEARING
before the
COMMITTEE ON ENERGY AND COMMERCE
HOUSE OF REPRESENTATIVES
ONE HUNDRED SEVENTH CONGRESS
FIRST SESSION
__________
FEBRUARY 14, 2001
__________
Serial No. 107-25
__________
Printed for the use of the Committee on Energy and Commerce
Available via the World Wide Web: http://www.access.gpo.gov/congress/
house
__________
U.S. GOVERNMENT PRINTING OFFICE
71-490DTP WASHINGTON : 2001
_______________________________________________________________________
For sale by the Superintendent of Documents, U.S. Government Printing
Office
Internet: bookstore.gpo.gov Phone: (202) 512-1800 Fax: (202) 512-2250
Mail: Stop SSOP, Washington, DC 20402-0001
COMMITTEE ON ENERGY AND COMMERCE
W.J. ``BILLY'' TAUZIN, Louisiana, Chairman
MICHAEL BILIRAKIS, Florida JOHN D. DINGELL, Michigan
JOE BARTON, Texas HENRY A. WAXMAN, California
FRED UPTON, Michigan EDWARD J. MARKEY, Massachusetts
CLIFF STEARNS, Florida RALPH M. HALL, Texas
PAUL E. GILLMOR, Ohio RICK BOUCHER, Virginia
JAMES C. GREENWOOD, Pennsylvania EDOLPHUS TOWNS, New York
CHRISTOPHER COX, California FRANK PALLONE, Jr., New Jersey
NATHAN DEAL, Georgia SHERROD BROWN, Ohio
STEVE LARGENT, Oklahoma BART GORDON, Tennessee
RICHARD BURR, North Carolina PETER DEUTSCH, Florida
ED WHITFIELD, Kentucky BOBBY L. RUSH, Illinois
GREG GANSKE, Iowa ANNA G. ESHOO, California
CHARLIE NORWOOD, Georgia BART STUPAK, Michigan
BARBARA CUBIN, Wyoming ELIOT L. ENGEL, New York
JOHN SHIMKUS, Illinois TOM SAWYER, Ohio
HEATHER WILSON, New Mexico ALBERT R. WYNN, Maryland
JOHN B. SHADEGG, Arizona GENE GREEN, Texas
CHARLES ``CHIP'' PICKERING, KAREN McCARTHY, Missouri
Mississippi TED STRICKLAND, Ohio
VITO FOSSELLA, New York DIANA DeGETTE, Colorado
ROY BLUNT, Missouri THOMAS M. BARRETT, Wisconsin
TOM DAVIS, Virginia BILL LUTHER, Minnesota
ED BRYANT, Tennessee LOIS CAPPS, California
ROBERT L. EHRLICH, Jr., Maryland MICHAEL F. DOYLE, Pennsylvania
STEVE BUYER, Indiana CHRISTOPHER JOHN, Louisiana
GEORGE RADANOVICH, California JANE HARMAN, California
JOSEPH R. PITTS, Pennsylvania
MARY BONO, California
GREG WALDEN, Oregon
LEE TERRY, Nebraska
CHARLES F. BASS, New Hampshire
David V. Marventano, Staff Director
James D. Barnette, General Counsel
Reid P.F. Stuntz, Minority Staff Director and Chief Counsel
(ii)
C O N T E N T S
__________
Page
Testimony of:
Ailes, Roger, Chairman and Chief Executive Officer, Fox News. 115
Biemer, Paul, Research Triangle Institute.................... 46
Boccardi, Louis D., President and Chief Executive Officer,
Associated Press........................................... 89
Heyward, Andrew, President, CBS News......................... 105
Johnson, Tom, Chairman and Chief Executive Officer, CNN...... 111
Konner, Joan, Professor of Journalism and Dean Emerita,
Graduate School of Journalism, Columbia University......... 34
Lack, Andrew, President, NBC News............................ 118
Risser, James V., Retired Director, Knight Fellowship
Program, Stanford University............................... 39
Savaglio, Ted C., Director, Voter News Service............... 85
Wattenberg, Ben J., Senior Fellow, American Enterprise
Institute.................................................. 43
Westin, David, President, ABC News........................... 95
Material submitted for the record by:
Boccardi, Louis D., President and Chief Executive Officer,
Associated Press, letter dated November 16, 2000, to
Congressman Billy Tauzin................................... 173
Dingell, Hon. John D., a Representative in Congress from the
State of Michigan:
Letter dated March 26, 2001, to Andrew Heyward,
President, CBS News, enclosing questions for the
record, and responses to same.......................... 174
Letter dated March 26, 2001, to Joan Konner, Professor of
Journalism and Dean Emerita, Graduate School of
Journalism, Columbia University, enclosing questions
for the record, and responses to same.................. 178
Letter dated March 26, 2001, to James V. Risser, Retired
Director, Knight Fellowship Program, Stanford
University, enclosing questions for the record, and
responses to same...................................... 180
Savaglio, Ted C., Director, Voter News Service, letter dated
April 13, 2001, to Hon. John D. Dingell, enclosing response
for the record............................................. 167
Tomlin, David, Assistant to the President, letter dated
February 5, 2001, to Hon. John D. Dingell and Hon. Edward
J. Markey, enclosing response to questions................. 170
Westin, David, President, ABC News, letter dated April 9,
2001, to Hon. John D. Dingell, enclosing response for the
record..................................................... 176
(iii)
ELECTION NIGHT COVERAGE BY THE NETWORKS
----------
WEDNESDAY, FEBRUARY 14, 2001
House of Representatives,
Committee on Energy and Commerce,
Washington, DC.
The committee met, pursuant to notice, at 11:07 a.m., in
room 2123, Rayburn House Office Building, Hon. W.J. ``Billy''
Tauzin (chairman) presiding.
Members present: Representatives Tauzin, Bilirakis, Barton,
Upton, Stearns, Gillmor, Greenwood, Cox, Deal, Largent, Burr,
Whitfield, Ganske, Norwood, Shimkus, Wilson, Shadegg,
Pickering, Fossella, Davis, Blunt, Bryant, Ehrlich, Buyer,
Radanovich, Pitts, Walden, Terry, Bass, Dingell, Waxman,
Markey, Towns, Brown, Deutsch, Rush, Eshoo, Stupak, Engel,
Sawyer, Wynn, Green, McCarthy, Strickland, DeGette, Barrett,
Luther, Capps, Doyle, John, and Harman.
Staff present: Mark Paoletta, majority counsel; Tom
DiLenge, majority counsel; Jan Faiks, majority counsel; Charles
Symington, majority counsel; Ann Washington, majority counsel;
Julie Corcoran, majority counsel; Anthony Habib, legislative
clerk; Yong Choe, legislative clerk; Edith Holleman, minority
counsel; Laura Sheenan, minority counsel; and Chris Knauer,
minority investigator.
Chairman Tauzin. The committee will please come to order.
The Chair recognizes a presence of a quorum and welcomes all of
you to this important oversight investigatory hearing on the
issue of election night coverage of the Presidential election
November 2000. I would ask our guests to settle in and get
comfortable. I apologize for the conditions of the room. Mr.
Dingell and I have commented just before about how we are
outgrowing the size of this room with the size of our
committee, and we apologize to our guests for the limited
space, and to the press for the limited conditions under which
you have to work today, we apologize.
The Chair recognizes himself for an opening statement.
Ladies and gentlemen, today we will be hearing from some very
important witnesses who will give us a real sense of what went
wrong in terms of the election night coverage of the
Presidential election of November 2000, but I would be remiss
if I did not remind all of us that this is not a new problem.
As Mr. Dingell pointed out to me, it was a problem of the
seventies. It was a problem going all the way back to the
Kennedy-Nixon election when Illinois was called in the one
column back and forth several times. It was certainly a problem
in the eighties and we have several charts that I think will
give you an idea of what we experienced in the eighties, when
elections began to be called on the basis of exit polling data
and early projections of winners in the Presidential race were
thought to have a profound effect on local races, particularly
congressional races in the West, when this committee and other
committees of the Congress held hearings, over a dozen
hearings, in the 1980's examining the problem of early exit
poll calls and its effect upon voter turnout in other
elections.
I have a chart that I will ask the staff to put up which
contains some of the headlines that were predominant in the
1980's. You can see these headlines: Networks in Dispute on
Fast Projections; Angered California Voters May Attempt to Beat
the Clock in 1984; Time Zone Fallout; TV Changed the Election
of the Eighties. These are headlines from important newspapers,
New York Times and Washington Post, Christian Science Monitor
and others in the 1980's. What follows is a chart that includes
some of the quotations of that era from the members of the
committees who actually held hearings.
I call your attention, for example, to the chairman of the
1984 House hearings, Representative Timothy Worth of Colorado.
His quote, which will go up in a minute, is that ``The evidence
is overwhelming from our perspective, at least my own, that
there is a responsibility that when you report early people do
not vote,'' he goes on to talk about the early calls made then.
Ed Markey, my good friend in 1980, who sponsored and pushed
forward legislation for a uniform poll closing time, was quoted
as saying ``Then your interest in Election Day is not building
an audience.''
I am saying another philosophy takes over. It has something
to do with projecting the winner. It becomes a race for the
networks, for the news departments on Election Day. And there
are quotes on this chart you can examine from Bob Matsui and
John Glenn, who is a Senator, who was also testifying at those
hearings. Our good friend Nancy Pelosi from California is
quoted saying ``As an organization person I can tell you that
the early projections had a very deleterious effect on the
morale and actual voting that occurred.''
So we have statements in these hearings going back in the
1980's about the problem that occurred with early calls in the
1980's.
If you'll also, however, look at the next chart, you will
see that we were not alone in criticizing the use of exit
polling. The networks themselves were criticizing the use of
exit polling and were questioning whether exit polling was in
fact a valuable tool or a dangerous tool. You'll see quotes
from none other than Walter Cronkite suggesting that exit
polling was a dangerous tool, from the head of CBS announcing
that this was not something that networks ought to be doing
because this was just guessing and projecting rather than
reporting the actual news of an election.
Of course, since the 1980's all of the networks have come
to rely upon those exit polling data more and more. In fact, as
we know, in 1990 some of the networks decided on a single
source for that exit polling, the Voter News Service, in order
to, I suppose, prevent some of the competition among the
networks to be the first with the news. In 1994, that
collaborative effort was broken again when one of networks
chose to go ahead of VNS and make decisions on their own and
their ratings soared because they were first out with
projections, and we arrive today at a time and an age when we
have seen another example of how exiting polling data produced
from VNS may have had a serious effect on the outcome of
elections in some local and other races out West because of the
early and sometimes flawed reporting of those results in the
East.
I want to quote to you from a young legislator back in the
1980's in his statement before this committee. The quote is as
follows: ``further, in today's technologically advanced age we
are experiencing a problem with dissemination of information,
sometimes incorrect, and the media's projections of winners
often before the polls even close.'' This young Member called
for a uniform nationwide poll closing time and a universal time
when voters therefore would cast their vote without the
influence of early calls. That young legislator was none other
than the chairman of this committee today.
The problem existed then and it certainly existed on
election night November 2000. The way this committee came into
recognition that we were still experiencing the problem despite
an agreement in the 1980's that the networks would not try to
make projections based on exit polling until most of the polls
had closed in the State was when we examined the problem of the
networks delaying calls for one of the candidates while making
speedy calls for other candidates. Something appeared to be
wrong, and so we held a news conference and called for this
investigation.
I am pleased to say that not only did we conduct a very
thorough investigation at this committee level but the networks
did so themselves. I want to thank all of the networks for the
work they did in self-examination of the problem. I
particularly want to highlight CNN for hiring outside
consultants to examine and critically evaluate the role of CNN
and other networks in their use of the VNS information on
election night and CBS for using outside counsel in their
report. Let me thank all the networks for the self-examination.
What we have learned from the self-examination by the
networks is that there are serious flaws within the VNS
modeling and those serious flaws produces statistical biases in
favor of Democrats in this case today and against Republicans,
that the statistical flaws tend to overstate the Democratic
vote in the exit poll and understate the Republican vote, and
we have charts again to demonstrate that and we will today.
The good news is that we discovered no evidence of
intentional bias, no evidence of intentional slanting of this
information. What we discovered, to our dismay, is that while
we've been told that exit polling is getting better in the
country, what we have learned is exit polling is getting worse,
that it is less scientific today than it was before, and that
the VNS models in fact produce some very bad information. As
one of the networks told me, ``garbage in, garbage out.'' And
the problem basically that we have to answer today is how can
we at this level, recognizing the very sensitive First
Amendment rights of the reporters and the networks to report
the news as they see fit, recognizing that we would defend your
right to do it wrong if you really wanted to, how can we assist
in getting some new agreements to do it right and how can we
assist through this investigatory hearing to not only settle
some of the outstanding issues that were not settled in the
eighties but perhaps make changes in the law that will help
produce a situation where Americans have a chance to vote
without being influenced by the reporting of the election
itself before the polls are closed.
And so, ladies and gentlemen, we arrive at this point, our
own investigation producing evidence of flawed and biased
modeling, the networks producing similar findings, the networks
being very responsible, I think, and critical of their own VNS
systems, and we arrive at this point where we give the networks
and the VNS representatives and others a chance to explain what
happened and what they suggest we might do in the future to
avoid these problems.
And before I finish, with the agreement of the minority, we
have prepared a brief 10-minute clip in chronological order of
the events of election night 2000. We would like to show you
that clip because it presents the problem, I think, in dramatic
form. This is the way networks were using VNS in November 2000.
[Videotape shown.]
Chairman Tauzin. Thank you. I've been asked by Mr. Dingell
to correct the record, that while we did share the video with
the minority there was no agreement to show it today.
Let me conclude by summarizing quickly again what our staff
discovered for us. What our staff discovered for us is that the
VNS modeling, according to our investigation, is seriously
flawed, that it underestimated in exit polling numbers 32
States for Bush and underestimated only 15 for Gore. But it
overestimated for Bush 15 States and overestimated for Gore 34
States, indicating some clear error in the system, and that is
the ultimate finding of our investigators.
We look forward today for a similar discussion of what the
networks themselves found and what VNS has found and eventual
testimony of the network representatives themselves. The Chair
now yields for an opening statement to my friend from Michigan,
Mr. Dingell.
Mr. Dingell. Mr. Chairman, I thank you. I enjoyed the
video. I hope my friends at the networks found it equally
enjoyable. I will observe that it chronicles a monumental
screw-up which I think has embarrassed an awful lot of people.
I would observe that today we are going over the results of
an election. My first memory of an election was Roosevelt in
1932. More recently, I remember the Literary Digest in 1936
which predicted that Alf Landon would win. As you remember, he
carried two States. And then I had the great enjoyment while I
was going to Georgetown University, just out of the Army, in
seeing a wonderful picture of Harry Truman holding up a copy of
the Chicago Tribune saying ``Dewey Wins.''
The business of predicting in the highly competitive and
complex business of elections is an enormously difficult task,
and I commend you for having these hearings because I believe
they will give us a chance to review what has been done, what
needs to be done, and how it is that we should approach this as
a Nation.
I would simply observe that for good motives and bad,
because of skill or incompetence and sometimes from outright
malevolence and sometimes for quite decent motives, elections
have been called wrong by the media for a long time. The good
news and the bad news are before us. First, the good news.
Because of the massive attention to miscalls by the networks on
election night and perhaps because of this hearing and inquiry,
Mr. Chairman, the networks and others have to varying degrees
taken a hard look at themselves and drawn tough and I think
appropriate conclusions. CNN, CBS and ABC in particular should
be commended for their efforts.
Another good piece of news is that contrary to inflammatory
allegations made in November the inquiry found no evidence of
intentional bias. Clearly had there been such, the credibility
of the networks would have been shattered, and I think properly
so. The very publicly and seemingly prematurely--there have
been expressed fears of some of my colleagues that have not
happily been realized on this matter.
And the final piece of good news is that this hearing may
serve as a wakeup call for all of us here, especially the
Republican leadership, to muster similar effort and energy to
have the House address the real electoral issues of voter
disenfranchisement. That is, I think, perhaps the sorriest
story of this election, and I hope that perhaps you and I, Mr.
Chairman, will be able to lend our skills to that task. I would
note that it tends to point out massive needs for reform of
financing and almost everything else.
Now for the bad news. From the outset it appeared that many
found that the inquiry was an attempt by the Republicans to
shift attention from the well-established election problems in
Florida that cost Vice President Gore the presidency. At the
outset it also appeared that this inquiry would cause
collateral, if not direct damage to the First Amendment
protections to the free press, and from the outset it appeared
that there were many allegations being made with too little
factual basis. I do not believe that these concerns are as
great as they were at the time, and I hope that they will be
eased by the hearing that we are holding today.
What did we learn, Mr. Chairman? That numerous problems
before and after and during election night led to network
errors that affect both parties, Democrats and Republicans?
That critical and later rescinded late night calls of Florida
for George Bush that were a basis of perception for some of the
media and the public that George Bush was a winner and Al Gore
was a spoiler was the networks' fault with Voter News Service
and Associated Press not in support? The answer to that is yes.
That networks can and must make major improvements to gain lost
credibility? Absolutely. The Congress after thorough review of
the pros and cons will need to at least consider requiring
uniform poll closing times? I think that is a certainty. And
that Congress itself is more credible when its investigations
are carefully calibrated and targeted, and public conclusions
are drawn after the investigation is done? I think that is
clear and I think with that you would agree, too, Mr. Chairman.
I want to thank all the witnesses for appearing here today,
particularly Mr. Boccardi of the Associated Press. I would note
that we Democrats will be particularly welcoming Mr. Boccardi.
He refused to allow his staff to talk to our people on this
side although they were happy to be interviewed by various
media outlets. I think we will want to ask him about that
little matter. I know he will have a fine answer for us. I
certainly look forward to hearing from him as AP will probably
have an interesting story to tell.
I thank you for recognizing me, Mr. Chairman. I yield back
the balance of my time.
Chairman Tauzin. I thank my friend. The Chair yields to the
gentleman from Florida, Mr. Bilirakis, for an opening
statement.
Mr. Bilirakis. Thank you, Mr. Chairman. I will be very
brief. I want to also add my welcome to the witnesses and thank
them for taking time away from their very busy schedules to be
here. I think we learned an awful lot from what transpired this
last November, and I might add partially into December. As Mr.
Dingell stated, there was a lot of good news and a lot of bad
news. I think the good news is an awful lot of American people
may be, for the first time, aware of how very significant their
vote might be. This business about my vote might not count or
won't count will probably not be as much in their mind as it
has been in the past.
The bad news is the controversies that have taken place.
The eyes of America really are on this committee, Mr. Chairman,
and I think they clearly expect us to do something. I think
with your leadership and maybe the bipartisan spirit that
hopefully will continue, we will do something right in this
regard. Thank you.
Chairman Tauzin. I thank my friend. The Chair now is
pleased to recognize the gentleman from California, Mr. Waxman,
for an opening statement.
Mr. Waxman. Thank you very much, Mr. Chairman. The networks
have a fundamental obligation to give us unbiased and accurate
information at all times and particularly when they are
reporting election results. But I want to read from an article
that appeared just today in the Los Angeles Times. It is called
``How TV Killed Democracy on November 7.'' It is an editorial
by Todd Gipling. He starts off saying ``Suppose that a first
cousin of Al Gore had been running one of the network news
teams issuing election night projections. Suppose that having
previously recused himself from a columnist job saying his
objectivity would suffer from family loyalty, this cousin had
chatted with Gore six times on Election Day. Suppose the same
cousin had been the first to declare Gore as the winner in
Florida on election night, helping coax the rival networks to
follow suit, leading George W. Bush to call up Gore in order to
concede, thereby helping to create that Gore was the duly
elected President of the United States long before all the
votes had been counted. Can anybody reasonably doubt that the
pundits would be working themselves into a nonstop lather
charging the liberal media as accessories to grand larceny? Can
we imagine, say, Rupert Murdock's Fox news channel right
leaning heads dropping the subject?'' according to Mr. Gipling.
Well, of course, what we know is that this did not happen
with Al Gore, but it did happen with Fox and John Ellis, and of
everything that happened on election night this was the most
important in impact. It created a presumption that George Bush
won the election. It set in motion a chain of events that were
devastating to Al Gore's chances and it immeasurably helped
George Bush maintain the idea in people's minds than he was the
man who won the election.
But I know we are going to look at different ideas. I think
the idea of a uniform poll closing is a good one. There are a
lot of ways to avoid the kinds of problems that we saw on
election night, and that is one of the best ones being
proposed. I think what we are seeing is a result of cost
cutting by the networks in their news divisions. It resulted
now, as we see it in hindsight, in the chaotic result of
election night reporting, which not only was embarrassing to
them but it had an impact on how the American people decided
the election and therefore had an impact on how the election
was ultimately decided.
Thank you for calling this hearing. I look forward to
hearing from the witnesses and getting a chance to question
them.
Chairman Tauzin. I thank my friend. The Chair will take a
moment to advise the committee that the investigators did make
fact findings regarding the 2:16 call in the morning that are
available, and we will discuss them during the course of this
hearing as well.
The Chair now recognizes the gentleman from Texas, Mr.
Barton, for an opening statement.
Mr. Barton. I want to thank the chairman for conducting
this hearing. I think it is timely. I would hope at the end of
the hearing that perhaps on a bipartisan basis we seriously
look at some legislation that would result in a common closing
of the polls around the Nation so that all candidates are
treated equally in terms of the spin that is put on which
States are going which way so that we won't have a situation
that apparently perennially occurs every 4 years where
depending on whose candidate is doing the best at one point in
time one party's candidate feels like they are being
disenfranchised or unduly chastised by the calling of the
election. So the hearings are timely and I hope that we might
have a legislative result occur jointly as a result of hearing,
and I would yield back.
Chairman Tauzin. I thank the gentleman. The Chair
recognizes the gentleman from Massachusetts, Mr. Markey.
Mr. Markey. Thank you, Mr. Chairman. I thank you for having
this hearing. The closest Presidential election in American
history certainly posed great challenges for pollsters,
journalists and commentators, not to mention the candidates
themselves. Early election evening calls in certain States for
certain candidates were driven by reliance on projected voter
tallies lists. In one key State, Florida, the call was clearly
too early as not all polls had closed in that State's Panhandle
region, but even so the call for Gore was based upon faulty
data, and the subsequent call of the entire election for Bush
in the wee hours of the morning was similarly flawed and
premature.
But was there bias? In the immediate aftermath of Election
Day questions were raised as to whether early calls were part
of a vast left wing conspiracy. Were dozens of network
journalists, the staff of the Voter News Service, all of the
network news directors, a score of election night anchors, and
the President's first cousin all co-conspirators in an
intricately designed plot to call key States early for Vice
President Gore? Or on the other hand, was it an ingeniously
designed deception whereby all of the co-conspirators would
provide an electoral feint early in the evening by calling
Florida for Gore simply to throw the conspiracy theorists off
their trail when their true design was to call the election for
George Bush just hours later without adequate supporting data?
A complicated conspiracy theory without question. It is of
course preposterous to believe there was such a plot. Rather, I
believe that the only bias in common for all the networks was
the desire to be first, this competitive urge which blurred
their own judgments. Rather than calling States in a way that
led to voter suppression, I believe that early in the evening
the reporting was consistently of a nature that the election
was indeed going to be very close, which could have led to an
increase in turnout rather than a lowering of turnout.
If one wants to question whether early calls for Gore
suppressed Bush voters out West, couldn't one equally argue
that if it was so apparent to everyone if Gore was going to win
then the Nader vote could increase, leaving Gore out in States
like Oregon and Washington. Why isn't that as equally
plausible? We will never know.
We could indeed have a wide ranging debate over who was
hurt most by election night coverage. Was it Bush because of
the early evening calls of a few States for Gore or was Gore
hurt more because of the subsequent network announcement of the
entire Nation for Bush, which created the presumption during
the entire recount that Gore had already lost? I believe that
any aberration in calling certain States at the time they were
called was based not on deficiencies in journalistic ethics but
rather on the fact that the networks were relying upon the
professionalism and the integrity of the work performed by the
Voter News Service. It was clear that the models utilized by
the VNS were highly flawed and the close election in Florida
amply highlighted for all the networks subsequently the
problems and the methodology utilized by VNS. The problem, in
my view, is not with the network news divisions or their
anchors therefore, but rather with VNS. It is clear this flawed
methodology and resulting shoddy VNS data misled the network
news divisions and caused many of the problems for the networks
and their election night coverage.
In addition, the fact that the networks readily agreed that
they erred in calling Florida before all the polls closed in
Florida is also well known. The networks' reaffirmation not to
call States in the future until all the polls in that State are
closed is welcome and laudable.
My hope is that this hearing will wind up serving a useful
purpose. If we can agree that there was no overt bias, no
networkwide conspiracy, then we should also stop searching for
unconscious messages packed into the choice of adjectives or
the on camera body English of network anchors. Instead let's
see what Congress can do so that in the future nobody can
allege that early calls affected voters elsewhere in the
country. I believe a key part of the solution is legislation
which would establish a uniform poll closing time. Uniform poll
closing bill, H.R. 50, which I was pleased to introduce with
the active leadership of both Chairman Tauzin and ranking
Democrat John Dingell, seeks to give Congress a constructive
way to prevent news reporting of the outcome of one State from
influencing the behavior of voters in States where the polls
are still open. It is both unrealistic and probably a violation
of the first amendment to mandate that the results on the East
Coast not be reported for 3 or more hours while the West Coast
is still voting, but news organizations have repeatedly
expressed a willingness not to report the results in a State
before the polls have closed in that State and not to report
the results of a time zone in a State if part of the State is
still voting in another time zone.
The fact that this pledge was not honored to the letter in
Florida will be noted today. But I believe that the networks
intend to correct this problem, and I also believe if we can
get broad support for uniform poll closing this hearing may
lead to a permanent beneficial change in the way we conduct
national elections.
Thank you, Mr. Chairman, and I also want to thank all of
our witnesses for their voluntary cooperation with the
committee. Thank you.
Chairman Tauzin. I thank my friend. I want to assure him
that neither the committee nor the chairman ascribes to any
vast conspiracy theories, left or right. I yield to my friend
from Florida, Mr. Stearns.
Mr. Barton. Parliamentary inquiry, Mr. Chairman. Is it not
true that under the rules, except for the chairman and the
ranking member, all of members' opening statements are supposed
to be 3 minutes or less.
Chairman Tauzin. That is exactly correct and the Chair will
ask everyone to abide by the 3-minute rule, including the
gentleman from Florida.
Mr. Stearns. Thank you, Mr. Chairman. Let me commend you
for holding these hearings. I thank the witnesses for
attending. I know how busy they are.
I am also proud to be a cosponsor of the Tauzin-Markey
legislation which creates a uniform poll closing. I would like
to put in the record a news release from Florida Secretary of
State Katherine Harris in which she requested the media to
delay predictions of the outcome of elections until 8 p.m.
Eastern Standard Time. ``Florida has six counties in the
Central Time Zone and the Secretary wants all Floridians' votes
to be cast prior to predictions on the winner of races. With
several races too close to call, full voter involvement is
imperative for Floridians to participate in the electoral
process. The last thing we need is to have our citizens in the
Central Time Zone think their votes do not count because it
certainly does. Waiting until 8 p.m. Eastern Standard Time
allows all of Florida the opportunity to decide the outcome of
races within Florida.''
This is dated October 30, 2000 and Mr. Chairman, with your
permission I would like to make that part of record.
Chairman Tauzin. Without objection, it will be made a part
of record.
[The news release follows:]
Florida Department of State
Katherine Harris, Secretary of State
NEWS RELEASE
secretary of state requests patience in predicting winners of races
Tallahasee, FL--Secretary of State, Katherine Harris today
requested the media to delay predictions of the outcome of elections
until after 8 p.m. Eastern Standard Time. Florida has six counties in
the Central Time zone and the Secretary wants all Floridians' votes to
be cast prior to predictions on the winners of races.
With several races too close to call, full voter involvement is
imperative for Floridians to participate in the electoral process.
``The last thing we need is to have out citizens in the Central Time
zone think their vote doesn't count--because it certainly does!''
Waiting until 8 p.m. Eastern Standard Time allows all Floridians
the opportunity to decide the outcome of races within Florida.
Mr. Stearns. Mr. Chairman, the networks and news
organizations are entrusted with delivering citizens with
unbiased truthful reporting. They are supposed to cover and
report the news, not create it. Unfortunately, this did not
happen on November 7, 2000. What we found is, ``a staged
collective drag race on the crowded highways of democracy,''
all stumbling past the finish line to be the first to report.
One would suppose network news organizations would have learned
their lessons from 1984 when they called the Presidential
election before the polls on the West Coast closed. As a
result, they promised us to voluntarily agree not to use exit
polls to call the race until the majority of the polls in that
State had closed. This did not happen.
Mr. Chairman, what we have here is a very important
hearing. Regrettably, by calling Florida for Vice President
Gore before all of the polls had closed in the State, the
networks' projections may have also depressed voter turnout in
portions of the Florida Panhandle, a region of the State which
is a Republican stronghold.
We, as well as the networks, have learned a lot since
election night, so I look forward to today's testimony to know
how and why the vote projections were made and, more
importantly, to learn what steps and procedures the network and
news organizations will take to ensure another election night
debacle does not happen again.
Chairman Tauzin. I thank the gentleman. The Chair recognize
the gentleman, Mr. Brown.
Mr. Brown. I thank the chairman. While the networks deserve
criticism from the Congress and the public, I am amazed that
this committee is holding a hearing about election night
coverage while this Congress and the Bush administration are
not moving legislation to correct the flaws in the system, to
standardize ballots, to establish uniform poll closing, to
modernize election procedures, and equipment.
This Congress must act to end Republican efforts to
suppress minority voters. The revelations of voter intimidation
tactics in Florida are one example of the practices that
national and State GOP officials have been using for more than
20 years to keep voters, especially minority voters, from the
polls. For 8 years as the Ohio Secretary of State I saw the
kind of voter intimidation, suppression and harassment created
and carried out by the Republican Party at the highest levels.
The evidence of voter intimidation in Florida reminded me
of the 1981 gubernatorial race in New Jersey. Sponsored by the
national and State Republican Party, the National Ballot
Security Task Force, comprised of off duty deputy sheriffs and
local policemen, monitored polling places in predominantly
African American precincts. They wore arm bands that identified
them as members of the Ballot Security Task Force. They posted
warning signs that they were patrolling the area and it was a
crime to violate election laws. The Republican Party
acknowledged doing that in a settlement later.
We saw in Florida 2000 a kinder, gentler version of the
Ballot Security Task Force. We know of the purging of thousands
of voters, mostly black voters, illegally from voter rolls. We
know of police checkpoints established near polling places. We
know of requests for additional forms of ID in predominantly
African American precincts. All of these tactics were created
and executed by Republican officials, usually high ranking GOP
officials.
The media had the responsibility to tell the public more
about these voter suppression tactics. These forms of
intimidation diminish the electoral process.
Similar to the suppression tactics, the media repeated some
of the same mistakes when reporting on the Florida recount. I
was in Florida during the recount and witnessed firsthand the
media's reluctance to fully examine statements the Republicans
made about the recount process. In Palm Beach County I stood 20
feet from Governor Pataki of New York as he repeatedly said
four recounts had been conducted. Four recounts had been
conducted. Like birds off a telephone wire, every Republican
elected official repeated this mantra. All fair-minded people
know that the four-recount charge was simply not true.
In another instance I stood by as Senator Lugar from
Indiana stated that his State doesn't do hand recounts. A
simple call to his elections office confirmed that the Hoozier
State does conduct hand recounts. But the media allowed Lugar
and Pataki and countless other Republicans to repeat this
mantra generally unchallenged.
These statements reflected a series of distortions backed
up by a conservative, corporate-owned media too lazy to
scrutinize such allegations and too eager to manufacture drama.
The media have the responsibility to check the facts for their
audience. I asked the news executives here today, scrutinize
our observations, refrain from adopting a ``he said, she said''
approach to news coverage because the ``he said, she said''
coverage causes politicians to exaggerate, to distort and even
to lie.
Florida surely taught us that. Do not accept what we say.
Make us tell the truth. This task is a challenge of today's 24
hours news cycle. I ask you to resist merely filling the time
with talking heads. I hope that the media does its job better.
I hope that we in Congress do our jobs better as well.
Chairman Tauzin. The Chair thanks the gentleman. The Chair
recognizes the gentleman from North Carolina, Mr. Burr, for an
opening statement.
Mr. Burr. Mr. Chairman, I ask unanimous consent that all
members be allowed to enter opening statements into the record.
Chairman Tauzin. Without objection, so ordered.
Mr. Burr. Mr. Chairman, let me thank you. Let me thank all
the members on this committee. Let me thank all the witnesses
that will be here today. I am not here today searching for
answers. I have had an opportunity to read the testimony, to
look over the interviews, to try to analyze the data, and I
have come to a conclusion, the networks screwed up. A
combination of flawed methodology, competitive forces, close
elections, and the pressure to be first overshadowed their
promise to be accurate. Ladies and gentlemen, if we were
handicapping the show ``Survivor'' and who the winner might be,
this might have gone unnoticed. But it wasn't. This was about
the election of the next President of the United States. We do
a lot and have done a lot to protect the Office of the
President. We teach our children that having one President, we
must protect the integrity of the office regardless of the
office holder. Our Founding Fathers entrusted the President
with incredible powers because they understood the importance
of the office in this very young country.
Those same Founding Fathers also entrusted the media with
incredible powers, powers that are so clearly stated that few
suggest that a change is necessary or constitutional. They
believed that for a Nation to grow its people must be informed.
They feared that without specific restrictions the government
might be tempted to filter the information and allow an
important trust to be broken.
Ronald Reagan said it best when he said he never understood
what was so important about the United States Constitution.
Every country had one. He said it wasn't until he read theirs
that he understood what was so powerful about ours. Theirs
starts ``We, the government.'' Ours starts ``We, the people.''
The American people are the single greatest asset of this
country. Their trust in their leadership and their trust in the
media must exist without objection. Without trust that fine
balance created by our Founding Fathers will quickly grow old
and be replaced. The integrity of the Office of the President
will only exist in the history books, and the freedom currently
entrusted to the media will be assaulted as often as people
disagree with the news.
Americans deserve to know that the information that they
hear from the media is legitimate, accurate, and truthful.
Let's give confidence to the American people that our Nation's
free system protects the public interest and does its best to
communicate the truth. I am confident that during the course of
this hearing the American people will listen anxiously to hear
the Members of Congress' commitment to conduct free and fair
elections and for the media's commitment to report factual and
accurate results.
Mr. Chairman, I yield back, but I also would like to
welcome Mr. Biemer, a constituent in North Carolina, and thank
you for the invitation to him.
Chairman Tauzin. The Chair thanks the vice chairman of the
committee for his welcome to Mr. Biemer. We will give you an
official welcome as soon as we can.
The Chair yields to Mr. Deutsch or Mr. Rush. Mr. Deutsch is
next.
Mr. Deutsch. Thank you, Mr. Chairman. Mr. Chairman, in your
opening comments you said the purpose of this hearing is to
investigate what went wrong with the election coverage, and I
think that is an important issue for this Congress to
investigate. But a far more important issue which at this point
this Congress is not investigating is what went wrong with the
election. And in a public setting like this I urge my
colleagues who have the ability in a formal way, which we as
the minority cannot call meetings in any committee of this
Congress or establish any formal committee of this Congress, to
look at the real issue of what happened this past November,
what went wrong with the election. Let me present, I think,
what is probably a much more accurate thing that no one at this
point has mentioned, that the exit polling was probably more
accurate than the counting the actual counting of ballots,
specifically as in Florida.
I spoke to the statistician for VNS, which will be at the
next panel, sitting at least; and I intend to question Dr.
Murray Edelman about this.
One of the realities of what happened in Florida, and again
I'm going to say it, is both the polling and exit polling very
well might have been more accurate than the actual counting. If
Florida was a foreign country and we had American election
observers in Florida and over 100,000 ballots were thrown out,
a majority of which were African Americans, there would be no
American who would accept the results of the Florida election
as a valid result.
And in fact that is what happened in Florida. There is a
direct correlation between ballots that were thrown out and the
racial complexion of individual precincts.
When the Supreme Court made its ruling to stop the counting
of ballots I said publicly and privately at that time that my
hope for the good of the country was that when they counted the
votes George Bush would win. But we now know through the good
work of many news organizations and not this Congress--and,
again, it has not been presented that much in the national
press, but if you read the articles and you understand the
numbers there is no question, there is no question, it is no
longer debatable that if the vote in Florida were counted, Al
Gore would be president of the United States.
So really in a sense we want to talk about projections, and
the results I would actually present to this committee is that,
in fact, the Supreme Court's political decision of stopping the
counting of the votes was in fact influenced by the missed
calls of calling Bush the President. If there was no winner
after November 7 I think the political decision very well might
have been different.
Let me just close on two points. One is, my good friend and
colleague, Mr. Stearns from Florida, mentioned the issue of the
Panhandle in Florida. It's not the first time that that
allegation has been mentioned. There is absolutely no
specific--any kind of empirical data to support those
allegations. They have been continuously discussed, as have
been other issues discussed during the whole post-election
effort, including other colleagues of mine who talked about the
Gore campaign's vast conspiracy regarding overseas absentee
ballots without any factual basis at all.
Chairman Tauzin. The gentleman's time has expired.
Mr. Deutsch. Thank you, Mr. Chairman.
Chairman Tauzin. Thank the gentleman.
The Chair now recognizes the gentleman from Florida, Mr.
Cox, for an opening statement.
I'm sorry, Mr. Greenwood is here.
Mr. Greenwood. Thank you, Mr. Chairman.
It is an everyday occurrence for Members of Congress to be
called and summoned to respond to the media. Each one of us
here receives probably a dozen calls a day from reporters,
radio, television, newspaper, magazines interrogating us,
interviewing us, calling us to task, reminding us editorially
of our responsibilities.
It is, on the other hand, an extraordinary rare occurrence
for Members of Congress to summon representatives of the media
before us and to interrogate them and to remind them of their
responsibility. So I think the fact that we're doing that today
tells us the importance with which we consider this issue and
reminds us of the incredible power of the media and
particularly the medium of television.
We are extraordinarily sensitive to the first amendment
issues here, and there isn't a member of this panel or Member
of Congress that wants to in any way infringe upon that. We
recognize the dangers that lie therein, but we do want to
remind the media of its responsibility. We do want to challenge
it to do a better job 4 years hence; and, in fact, I am certain
that the media may not even need that reminder. It is probably
busy about figuring out how to do that.
It seems to me not a difficult task. The fact that the
television networks have been able to collaborate as they have
with VNS and have a common mechanism by which to call these
exit polls makes it a practical--relatively easy matter,
practically speaking, for the media to collaborate on how to do
it right; and I'm hopeful that this hearing and what falls from
this hearing will produce that result.
Yield back.
Chairman Tauzin. The Chair now yields to the gentleman from
Chicago, Mr. Rush, for an opening statement.
Mr. Rush. Thank you, Mr. Chairman.
Mr. Chairman, I also want to commend you for holding this
hearing. However, Mr. Chairman, my colleague from the State of
Florida, Mr. Deutsch, asked a question; and that question was,
what went wrong? Mr. Chairman, I for one, an individual who has
fought all of his adult life to ensure that everyone have
access to the polling place on Election Day and that their vote
be counted, I for one have some pretty serious concerns about
and regarding about what when wrong on Election Day.
Mr. Chairman, in the State of Florida and also in my State
of Illinois, what went wrong on Election Day was the fact that
we had literally hundreds of thousands of people throughout
this Nation who got up early on the morning of November 7 and
proceeded to the polling place with the thought in mind and
with the objective of voting for their choice for President of
the United States. And, Mr. Chairman, we found out that between
leaving their homes and arriving at a polling place that
literally hundreds of thousands of people had been harassed by
police departments, not only in the State of Florida but other
places, asked to present identification, were told that they
were under investigation for criminal charges, all types of
shenanigans by police departments all across this country.
We found out that people who decided that they wanted to
come to the polling place to exercise their constitutional
rights were in more than one type, more than one way denied
access to the polling place, access to vote because of some
kind of spurious charges against them.
Then, Mr. Chairman, we also find out that those who were
lucky enough to cast a vote, for a lot of different reasons
those vote weren't counted. And my question to you and my
question to the Republican members of this committee and the
Republican leadership in this Congress, when will we have a
Congressional investigation, a congressional hearing on the
issues that those folks who were denied the right to vote, when
will we have a hearing so that we can get to the bottom of the
issue so that we can get some answers about why they were not--
why they were denied the right to vote?
Mr. Chairman, this might be a good hearing. We might be
able to get some answers. But let's not just focus on the media
today. Let's also look at what happened beyond the media. Let's
look at what happened with the police departments all across
this country. Let's look at what happened in the polling place.
Why were African Americans and other minorities denied the
right to vote? Why were they denied the right to have their
votes counted? I know that that's an appropriate concern not
only for this committee but other committees in this Congress.
Again, the question is, when will the Congress ask the
right questions in order that we get the right answers? When
will we have a hearing, a congressional hearing, an official
Congressional hearing to allow for those individuals who were
denied the right to vote on November 7 to come before this
committee, come before other committees, to come before this
Congress, to get their answers in terms of why they were denied
the right to vote?
Thank you. I yield back the balance of my time.
Chairman Tauzin. The Chair thanks the gentleman.
The Chair has a responsibility now, which I do not,
frankly, like to admit, but the gentleman's request involves
the jurisdiction of another committee, believe it or not. It's
one of those small areas we don't have jurisdiction over. But I
thank the gentleman for his request.
The Chair now recognizes the gentleman from California, Mr.
Cox, for an opening statement.
Mr. Cox. Thank you, Mr. Chairman.
In this exceptionally close election the TV networks in
this Nation came face to face with a dilemma that's been
familiar for some time to people in subatomic physics, that in
some cases the act of observation influences and changes
physical reality. To observe something is to influence it; and
that can have very real consequences, as many of my colleagues
have pointed out. The experience of last November illustrates
this dramatically. An overzealous competition to, as Mr. Markey
said, get there first, to predict the Presidential winner
first, inflicts costs on the entire Nation, on the viability of
the democratic system and on journalism itself.
What took place on election night we will hear from this
first panel and from subsequent panels but, to quote from the
outside review conducted for CNN, reflected commercial rather
than journalistic values. ``In calling winners of individual
States based on exit polling and votes from sample precincts,
accuracy and completeness of information were sacrificed to the
pressures of competition.''
Our mission in Congress can't be to police journalistic
ethics. That's why these internal reviews go on. I dare say
that CNN did not, in fact, do much worse than the other
networks, even though the outside review that was conducted of
CNN seemed to be the harshest at all. It's much more likely
rather the case that CNN is the only one that had an outside
rather than internal review. And where there were some outside
reviews conducted, the criticism probably would have been on
par for all of the other networks.
That's good. That's good that these reviews go on, and some
of them are indeed outside reviews. And it's good that Congress
is relieved of that responsibility, because I believe it is not
within the bounds of our legislative jurisdiction. But we do
have responsibility in the area of election reform, and some of
what we can do with election reform is directly influenced by
what the networks do themselves.
I just want to add to what's been said already my view from
California where Democrats and Republicans have for years been
complaining about the calls on the East in close elections.
Congress held hearings on this very subject after the 1980
election. Those hearings went on for years. There was a report
of the House Administration Committee, which I have with me
here, that makes very specific findings on these very topics.
And in 1980 the shoe was on the other foot. It was not
Republicans who were complaining about the early call in
Florida but rather Democrats who were complaining about the
early call of the entire election, which prompted an early
concession from President Carter.
The House Administration Committee concluded early
projections--and this was in the 1980's and this was, of
course, a Democratic majority in Congress--early projections
undermined people's belief in the importance of their vote, a
belief which is essential in a democratic society.
Some of the evidence before Congress at that time from the
State of California included our Secretary of State's
testimony, March Fong Eu, that early projections caused havoc
and had a significant impact on voter turnout which she said
dropped to practically nothing in the last few hours of voting.
The same came from Diane Feinstein, then the mayor of San
Francisco; from Nancy Pelosi; and you said earlier a field poll
showed that 15 percent of nonvoters said they failed to vote
because of early projections.
It is for that reason that I have sponsored with the
chairman and with Mr. Markey the early--or, excuse me, the
uniform poll closing legislation, which I think is going to be
very much in people's minds as we hear the testimony here. We
hope that such procedural reforms, in conjunction with the
reforms being put in place by the networks themselves, can help
us accomplish this objective.
I thank the chairman.
Chairman Tauzin. The gentleman's time has expired. The
Chair thanks the gentleman and recognizes the gentlelady from
California, Ms. Eshoo, for an opening statement.
Ms. Eshoo. Thank you, Mr. Chairman. I join all the members
of the committee in thanking you for holding this important
hearing. I would also like to thank the distinguished witnesses
that have come to us today from both panels and most especially
to the retired director of the very distinguished Knight
Fellowship Program at Stanford University, Mr. Risser.
Let me just try to summarize some of my thoughts on this
hearing and a few things beyond it, which I believe belong to--
in the responsibility column of the entire Congress. I can't
help but think that just as our markets, which are the broadest
and the deepest in the world, that the coin of the realm really
is confidence. That's why investors invest; that's why we're
the envy of the world; and that's why we are very sad to have
lost that jurisdiction in this committee, Mr. Chairman, over
the Securities and Exchange Commission.
But we have to have confidence or the confidence that I
just expressed--the American people have to have confidence in
their system of election, and I think that's really at the
heart of today's hearing.
Most frankly, I didn't know anything about VNS before
November 7. So I think people that are here today are going to
talk about who they contract with, how they conduct their
business, how they are looking into building the confidence of
the American people and how they report to them, not reporting
any biases or individual views but rather reporting very
accurately or as accurately as is humanly possible in our
technologies that provide for so much more of that today in
bringing that information forward to the American people.
We would always judge emerging democracies by their
elections. And a lot of things went wrong in this election. And
while it may not be the purview or the jurisdiction--and most
people listening in today don't even know what the word
jurisdiction means in the country, but they know something went
wrong. This is not whether the Republicans won and got their
candidate into the White House or that Democrats are whining
because their nominee didn't make it. This is an American
issue. This is something that strikes at the heart of
democracy.
I hope that there will be a delegation led by the
leadership from both sides of the aisle with open minds and
open hearts to go to congressional districts to listen to
people. We have a cancer that needs to be put out of our
election system, and that is any American that has been
deprived of casting a vote and that their vote doesn't count.
So, yes, we need to reform. We need to hear from the
networks and what they plan to do. We need to move to, in my
view, a uniform poll closing time. We need to have, in my view,
national ballots. We need to have equity when it comes to
equipment that goes to our polling stations all over the
country. But we also have to look very deep. We have to look
very deeply and be willing to----
One of the greatest marks of America in my view is that we
are willing to acknowledge when we have done things wrong. It
may take 10 years. It may take 20. It may take 50. But we
acknowledge it, and that's part of our greatness. So I think
that the Congress needs to embark on that journey.
Today is an important first step, Mr. Chairman. Thank you
for holding the hearing. And I look forward to working not only
on this but those issues that may be out of the jurisdiction of
this committee because I think the American people are counting
on us to do so. Thank you.
Chairman Tauzin. The Chair thanks the gentlelady for all
her curtesies; and the Chair recognizes the gentleman from
Georgia, Mr. Deal, for an opening statement.
Mr. Deal. Thank you, Mr. Chairman.
Mr. Chairman, today I'm reminded of the biblical admonition
that to whom much is given, much is required. As we delve into
this area of first amendment constitutional rights, likewise
the recognition that even that must be exercised responsibly.
I don't think there's any doubt that our Nation has given
much to the national media, our valuable time and listening to
what you have to say or reading what you have to write and also
our expectations that the news is not self-generated but
reflects true facts that are external. But today this hearing
will focus on news that by its very definition is self-
generated, that of predicting elections; and, therefore, we
must question whether the facts upon which this news is
predicated has any preventible statistical bias. In other
words, why did VNS use sample models that had not been adjusted
for decades and in my part of country, the South, did not
reflect the very apparent change in regional political
alliances that had been manifest in many elections that
preceded the Presidential election of 2000?
So, Mr. Chairman, I look forward today to hearing what the
media intends to do to fulfill the high expectations of the
public and what, if anything, we as elected representatives
should do to assist them in that direction. Thank you, Mr.
Chairman.
Chairman Tauzin. The Chair thanks the gentleman.
The Chair recognizes Mr. Stupak for an opening statement.
Mr. Stupak. Thank you, Mr. Chairman.
After what happened in the November election, it is
critical that Congress take a bipartisan look at how to assure
that every American who wants to can exercise their right to
vote and that his or her vote is accurately counted. The
American people want us to fix the problems with voting and
tabulating machines, with badly designed ballots and careless
election officials who deliberately or otherwise keep people
from voting. And they deserve to have these problems fixed.
They deserve, Mr. Chairman, to have them fixed in a deliberate
and fair manner, one which does not include--does not include
partisan rhetoric.
Mr. Chairman, I want to work with you constructively over
the next 2 years to ensure this committee works in a fair and
bipartisan manner. I would gently, gently point out to the
chairman that the initial statements offered by his office
suggested that there was initially biased coverage of the
networks in favor of Al Gore. I am glad to see in the last
week's press conference you admitted there was no evidence of
intentional bias although I could point out that you seem to
have made a premature call in this case just like the networks
made two premature calls on election night.
Chairman Tauzin. Would the gentleman yield? I'll extend the
time. I want to correct the record.
The Chair did not make a premature call. All the Chair
stated in the initial press conferences was that there was an
obvious bias in the results. We didn't know what caused it. I
was asked whether I thought it was intentional. I said I don't
know. Until we investigate, I can't say. We investigated, and
the investigation indicated no evidence of intentional bias,
and we called it that way. I think we called it as we saw it
correctly then, and we called it correctly today.
I thank the gentleman.
Mr. Stupak. I thank the chairman. I was going to say, but
like the networks who relied upon the best available
information filtered through your own beliefs. And now with
hindsight we find there was nothing in the network's coverage
on election night that indicated that either candidate was on a
roll. There was no credible evidence of the allegation when it
was made, and there is even less now. Members of Congress, like
members of the media, must do their research before they make
serious allegations.
The second allegation was that the networks had effectively
called the entire election for Al Gore by 8 p.m. when they gave
Florida to Gore and suppressed voters in California. Even a
cursory review of the transcripts of the network's coverage of
that night makes it clear that allegation cannot survive.
Selected excerpts from the transcripts do not justify such a
conclusion and come dangerously close to censorship of the
press.
I welcome this hearing and attention our committee has
focused on the Voter News Service and the networks' process for
making election night calls. This scrutiny has forced VNS and
the networks to examine their decisionmaking process and
improve it in the future. No one has an alleged that the
current system is flawless or does not need to be improved.
However, I believe Congress should be investigating the true
and most fundamental flaw exposed by this Presidential
election, the fact that our country needs to improve and
standardize the voting systems in this country to ensure that
everyone's vote is counted in the way the voter intended it to
be.
Nothing the networks did or did not do changed the outcome
of this election. Not the first call of Florida for Vice
President Gore nor the subsequent Florida and overall election
call for President Bush, both of which were retracted. What did
change the outcome of the election was a flawed ballot design
in Palm Beach County, an inadequate ballot and counting method
in Florida and other jurisdictions all across this country.
Mr. Chairman, I look forward to hearing from the witnesses
today. I look forward to working with you and this Congress on
a range of important issues, not the least of which is the
election reform. I only suggest that we put this hearing in its
proper context.
Thank you, Mr. Chairman.
Chairman Tauzin. I thank the gentleman.
The Chair wishes to do two things real quickly. One is to
also clarify the record. We received a note that NBC has
indicated it also included an outside expert in its initial
analysis. Let me invite the representatives of the networks
when they present their testimony to clarify the record to make
sure we know if they were outside experts who were part of that
internal review so that the record might adequately reflect it.
The Chair would also announce that we have two votes on the
floor, and that those are the only two votes of the day, and
that what I would like to do is to take another opening
statement or two while we have the time. Then we will recess
for a half hour and come back at 1 and hopefully complete
opening statements and begin to hear from our witnesses as
rapidly as we can.
My apologies for these interruptions.
The Chair recognizes the gentleman from Oklahoma, Mr.
Largent, for an opening statement.
Mr. Largent. Thank you, Mr. Chairman.
I'll be very brief and simply say that we hold one national
election. We do that every 4 years. And I think protecting the
integrity of that process is the responsibility of every Member
of Congress and specifically as it relates to how the media may
or may not influence the outcome of that one national election
we hold is definitely under the purview of this committee.
So I want to thank you, Mr. Chairman, for holding this
hearing. Because, frankly, when you think about it, to bring up
this painful experience, and it was painful for all of us, to
bring this experience back up before this committee is a hard
thing to do. And I would like to believe that we would be
holding this hearing regardless of what the outcome was,
because protecting the integrity of this election is so
important.
So, Mr. Chairman, I would just like to say I'm looking
forward to hearing from our panels and thank you for having the
courage to hold this hearing because I do believe that
protecting the integrity of this process is important to our
democracy. Thank you.
Chairman Tauzin. Thank my friend.
The Chair recognizes Mr. Sawyer for an opening statement.
Mr. Sawyer. Thank you, Mr. Chairman. I'm going to be brief
and turn in the rest of my opening statement for inclusion in
the record.
I really agree with all of my colleagues. This oversight
hearing provides a great opportunity not only to understand
what happened but to educate Members of Congress on the efforts
that the networks are making to improve the quality of
reporting. Early or erroneous reporting is important because of
the concern that self-fulfilling prophecies can affect the
outcome of an election.
Yet it's very clear that a legislative fix is not the
proper course of action here. The first amendment would
preclude us from prohibiting the media from interviewing voters
as they exit the polls. It would preclude Congress from
prohibiting the media from reporting the results of those
polls. And the first amendment clearly gives the media the
right to choose how and when to report the news. It's a
fundamental protection not just for the first amendment rights
of journalists, but in their role as surrogates for 280 million
American citizens it protects the first amendment rights of all
of us.
I believe that early or erroneous projections need to be
addressed by the networks and not by Congress. In this
competitive news environment, perhaps the most competitive in
human history, the networks have every incentive to make sure
that what happened on election night last year does not happen
again.
I look forward to hearing the recommendations that have
been made with regard to the models and the new policies that
will come before us today, but let me make one final
observation. Statistical analysis of very large data sets is a
subject I've studied in depth over the past dozen years. It is
a highly developed and deeply sophisticated science, one
applied in only the most superficial ways in exit polling and
in on-the-fly election projections. I hope that the statistical
models and methodologies used in electoral applications will
attract the careful scrutiny of both this committee and the
networks and in so doing not only to improve reporting in the
future but, even yet, to understand what actually happened in
this election and to answer Peter Deutsch's question what went
wrong.
Thank you, Mr. Chairman.
Chairman Tauzin. I thank my friend for his insightful
comments and recognize the gentleman from Michigan, the
chairman of the Telecommunications Subcommittee, Mr. Upton, for
an opening statement.
Mr. Upton. Thank you, Mr. Chairman. I appreciate your
calling this hearing and the hard work of both you and the
staff in putting this together, and I also appreciated looking
at the videotape at the beginning of the hearing as well. It
brought a lot of good and bad memories back from that haunting
evening. But it also underscores the need for real election
reform. America was embarrassed by that seesaw night.
I am anxious to hear the testimony by those that rely on
the VNS, Voter News Service. What was particularly tragic to me
was that the UNS, the Upton News Service, did a better job; and
it did a better job without a single field staff or exit poll
anywhere in the country. I would note that the UNS accurately
predicted Florida, Pennsylvania, Tennessee, West Virginia,
Illinois, Missouri, Ohio, Michigan and even easy ones like
Texas, California and New York. No, it didn't end up being
perfect, but it did call the night right. As a news
organization, as news organizations, America wants fairness and
it wants accuracy. And, sadly, we didn't see a lot of it on
November 7.
I yield back my time.
Chairman Tauzin. Thank my friend.
The Chair recognizes the gentlemen from New York, Mr.
Engel, for an opening statement.
Mr. Engel. Thank you, Mr. Chairman.
I think many of my colleagues on the committee would agree
that if we were engaged in a lawsuit we could easily be
entitled to compensation for mental and emotional distress.
Before discussing the news media's action on election
night, let me first say that I believe wholeheartedly that
there is a need for an overhaul of the voting process. Though
new equipment is key, we need also to be looking at whether
registration is handled in a timely manner and what day and
time we vote, how well voting procedures are explained to each
person and how the votes are tallied and verified.
I'm gravely concerned about reports that people were
prevented from voting, many of whom were minorities. The Civil
Rights Commission is holding public hearings at which many
people have testified that they were not allowed to vote
because they weren't on the rolls. Nor were they afforded the
opportunity to vote and sign an affidavit for later
verification. Such allegations must be thoroughly investigated.
Al Gore was not the only one who lost that night. The
American people lost that night, and the news media also lost
that night. For many years public confidence in the news media
has been on the decline. I suspect that it took a nose dive on
election night.
In order to begin winning back that trust, the news media
must take action, and in the spirit of my friend Tim Russert
here are my suggestions.
No. 1, slow down. The American people don't need to know
that President George W. Bush sneezed 30 seconds after he did
it.
Two, check facts. Check your facts. Too often, I see news
stories that are just plain wrong. In the case of the election,
pay greater attention to State law. The election was so close
that a mandatory recount was required, thus making the outcome
murky, not in the bag.
And, three, balance. Strive much harder for balance. When
you interview someone on a controversial issue, get an opposing
point of view. That may make the news story longer, but it will
also make it better.
This is the formula for winning back the people's trust;
and I thank you, Mr. Chairman for these hearings.
Chairman Tauzin. Good job, Eliot.
The Chair recognizes the gentleman, Mr. Whitfield, for an
opening statement.
Mr. Whitfield. Mr. Chairman, thank you very much. And I'm
sure the representatives of the network feel they've been
punished enough by listening to all of us all this time.
I would simply say that there's a lot of cynicism in
America today. I hear a lot about it in my district, and
frequently people say I don't really believe much of what a
politician says today. I don't believe a lot about what I read
in the paper, and I don't believe a lot about what I hear on
television today.
I think that's one of the real tragedies of what happened
in Election 2000 and the reporting of it. Instead of building
up confidence in the American people in our established
institutions like the press, we seem to be tearing that
confidence down. I hope this hearing will focus on that issue,
and I know that the networks will make every effort to correct
it because they have every interest to do so.
I look forward to the balance of the hearing.
Chairman Tauzin. I thank the gentleman.
The gentleman, Mr. Ganske, is recognized for an opening
statement.
Mr. Ganske. Mr. Chairman, here's the crux of the problem:
Exit polling can affect those who haven't voted if the polls
are still open. You, Mr. Chairman, and Mr. Markey have a bill
to forbid exit polling results until the Nation's polls are
closed and then to make uniform closing times across the
country. The goals are laudatory, but the implementation is
difficult because of equity problems with voters on the East
Coast having better voting hours than those on the West unless
have you very long poll hours nationwide. Nevertheless, I
commend you and Mr. Markey for posing a solution to the problem
of exit polling affecting whether voters even bother to vote;
and I look forward to the testimony.
Chairman Tauzin. I thank my friend.
Will my friend yield quickly? I want to correct the record.
The bill does not prohibit exit polling. I think that would be
an infringement on the first amendment.
Mr. Ganske. Exit polling reporting.
Chairman Tauzin. All it simply does is set a uniform poll
closing time. The networks have rather uniformly after their
internal reviews indicated they would probably all agree not to
use exit polling results until after that poll closing time. So
I think their voluntary agreement to do that along with the
bill might go a long way.
I thank the gentleman for his kind comments.
The Chair announces a recess until 1.
[Brief recess.]
Chairman Tauzin. The committee will please come back to
order. We'll ask our guests to take seats and get comfortable.
The Chair will recognize the gentleman from Texas, Mr.
Green, for an opening statement.
Mr. Green. Thank you, Mr. Chairman. I will not give my
total open statement and ask it be submitted for the record.
I will just briefly say that I appreciate the chance that
the committee has to look at this. I know, historically, the
committee has looked at this before. I think from the testimony
I've read the networks agree the early closing, the early
announcement needs to be corrected and also the--some type of
uniform election date. Now, I know--or election closing time.
Do I believe there were mistakes made in the decision-
making process of the networks? Sure. And do I believe the
Voter News Service used bad exit polling? Sure. Do I believe
this issue is a major factor in deciding the outcome of the
Election 2000? Absolutely not.
I think the rest of my statement, Mr. Chairman, will point
out that the biggest concern that I have is that the Florida
election system--and there but for Florida could go any State--
that predominantly the number of discarded ballots were
predominantly minority Democrat voters; and I would hope if not
this committee then this Congress would look at not only our
issues today but also why we have such a large percentage of
overvotes, for example, in--46,000 overvotes as compared to
17,000 for President Bush. I think that we need to look at
that.
We need to look at, for example, Palm Beach County. The
ballot design was not new, but there were 8,000 voters who
chose Vice President Gore and actually had theirs discarded. In
that country, the Democratic candidate for Senate in that
county won by 10 to 1.
So I think there's a lot that we need--this Congress needs
to look at the election and particularly see what we can do to
make sure that when people go vote they know their vote will be
counted.
I particularly appreciate Mr. Deutsch's comments. In
Florida, maybe the exit pollings were right and the actual
counts were not.
So, Mr. Chairman, I yield back my time.
Chairman Tauzin. Thank my friend.
The Chair recognizes Mr. Bryant for an opening statement.
Mr. Bryant. Thank you, Mr. Chairman.
I, too, want to add my thanks for you holding these
hearings. Certainly they've expanded, at least in terms of the
opening statements. We've heard from the other side maybe
beyond the scope--the intended scope of the hearing. But,
nevertheless, we will move forward.
I want to thank the extremely qualified and talented
witnesses we have, of the various panels. As I look back across
the audience and see a couple of panels waiting who are from
television I can't help but think that you can relate now to
what we who are consumers of your product feel during long
commercials. In essence, that's what we're hear--we paid a lot
of money to get up here.
Chairman Tauzin. Will the gentleman yield?
My wife would also like ask you to keep the volumes the
same when you go to those commercials. We never had a chance to
say that. Thank you. Thank the gentleman.
Mr. Bryant. In some ways, it's very similar. But,
hopefully, we will move along now and hear your testimony.
I do have limited comments. A lot of what I would like to
say has already been said. But I would reiterate I think a
couple of points made particularly by Mr. Greenwood and others
that we are talking about the Constitution and the First
Amendment here, freedom of speech and certainly we intend to
tread very, very lightly here.
There are restrictions, as all of us that went to law
school understand, to the first amendment, but they are very
limited; and certainly we're not talking about those here. On
the other side, there is a counter position, that anyone who
speaks should, particularly when it comes to something as
important as elections, should speak responsibly. And our
concern that I share with Mr. Largent and others in terms of
the influence, I think you understand completely the influence
that you do have over the public in so many ways; and to deal
with the institution of electing a Presidency we have to be
extremely careful. I'm sure there were lessons learned by all
here, and they will be taken to heart as we move toward the
next election.
One of the things I think a lot of us will endorse is the
idea of a uniform closing time on Election Day, particularly
when you've got many States like my home State of Tennessee who
have 2 weeks of unfettered early voting. I think as more States
move to that it would be easier to have a uniform closing
period of time, even though it is somewhat complicated. But
also we must insist upon obedience by the media of your
agreement to not call States until the polls are closed.
Finally, I would comment on a couple of quick issues that
have been raised about exit polling being accurate. As we will
learn later in the hearing, some of the counties in some of
these States, the 16 percent in terms of being off in that exit
voting--and I don't think anyone here is saying that any of the
counties, particularly in Florida, the undervote was 16 percent
of the total vote. I hope not. But, in any event, so much of
what happened in Florida, like other States, is a State issue
in terms of how they conduct their polls and how they run their
polls and run their elections and count their votes and all. I
hope we shy away from trying to Federalize every election in
every State. I don't think that's our job.
As I watched the coverage, as many of you did, the post-
election counting in Florida, the three or four counties that I
saw in contests so much seem to be counties that were actually
controlled--election commissions and people who operated and
administer the elections in those counties were controlled by
what appeared to be the Democrat party. And I know we talk
about butterfly ballots--this is again rehash and rehash--but
again that was approved beforehand by a Democrat administrator
in the county.
So these are really State issues, and I hope the States do
look at how they conduct their elections. I do hope they look
at how they write their ballots and the due process they give
before the election occurs for people to object and how the
voting population is educated. So many of these undervotes were
out there, were multiple votes. People voted more than one
candidate. I don't know how you can count those.
But, again, a lot of this comes back to the States. A lot
of this will come back to the individual voters to make sure
they're informed and know how to vote. And if they have a
problem voting at the box they go seek an official out and say,
I voted wrong; or I don't know; I'm confused here.
But so much so much of our democracy depends on individual
responsibility. So I hope we use caution not only in treading
on the first amendment but also in the Federal Government
trying to impose its big body in the State-run elections.
I would--I think I have run out of time--would yield back
any time I might have remaining.
Chairman Tauzin. Thank the gentleman.
The Chair recognizes Mr. Strickland for an opening
statement.
Mr. Strickland. Thank you, Mr. Chairman.
Mr. Chairman, with all due respect to my colleagues, it
seems a little ironic to me that we should be criticizing the
media for saying things to the American public which is not
accurate. Could it be that we are applying a standard to the
media which we ourselves as Members of this body do not
observe?
A case in point: On C-SPAN this morning, the networks were
criticized for their delay in calling Ohio for Mr. Bush. The
accusation was made that Mr. Bush had won Ohio by six points.
But the fact is that Ohio was a much closer election than that.
Mr. Bush won Ohio by less than four points. In fact, the
official results from the Ohio Secretary of State indicate that
Mr. Bush got 50 percent and that Mr. Gore got 46.4 percent.
I believe in this committee we should do nothing that would
interfere in any way, to any degree with first amendment
protections. But we can and we should clarify what has happened
and make sure the public is fully informed about those facts,
and then trust the public to make appropriate use of those
facts.
From my perspective, as it turns out, the early initial
projections in Florida were in fact accurate; and I believe
that the counting which is going on in Florida will eventually
substantiate the fact that the networks were initially correct
in their judgment.
I would hope that the major issue of electoral reform would
capture the attention of this committee. The people in my
district are not complaining to me about election night
projections, but they are wanting to make sure that every
American can vote and that every vote is counted.
What we have today in our country is the unfortunate
situation where many Americans believe that the person who was
elected by the people does not occupy the Office of President.
That is a sad fact, and it is a situation that we should not
ever allow to happen again.
I return the balance of my time.
Chairman Tauzin. The Chair thanks the gentleman and
recognizes Mr. Norwood for an opening statement.
Mr. Norwood. Thank you, Mr. Chairman. I'll be very brief,
but I do want to thank you for having this timely hearing. I
think it's very appropriate. And separate from our friend in
Ohio, our people are complaining about the exit polling and
when it came out. But this--we need to stay on the point here.
Probably all of us believe in the first amendment very
strongly. We believe in your right to free speech, and we
believe in our right to free speech. However, we also believe
in the right of a fair election. And part of that and the part
that we are particularly zeroed in on today and may look at
other parts later, but today we're talking about how does
incorrect exit polling affect the outcome of national
elections.
It isn't just about the State of Florida. I think our
investigators and others have proved without a shadow of a
doubt that the exit polling that caused the networks to predict
winners and losers incorrectly and wrongly was a very flawed
model, and that is important. In the 1980's, it was important
to the Democrats. In year 2000, that was very important to the
Republicans.
I hope all of us on both sides of the aisle, both sides of
this committee, will understand that next time it may be you.
We're not going to any time legislate, I believe, against
networks' freedom of speech or any of their rights. But it is
important to point out to the networks that they do have a very
large responsibility not to incorrectly affect an election,
regardless of who won, regardless of which party you're in.
So I thank you, Mr. Chairman, for this. I believe as we
hear from the witnesses today that everyone will leave here
knowing that the exit polling, the model we visit with today is
simply flawed; and that must be corrected.
Thank you, Mr. Chairman. I yield back.
Chairman Tauzin. I thank the gentleman.
The Chair recognizes Ms. DeGette for an opening statement.
Ms. DeGette. Thank you, Mr. Chairman.
I've been sitting here patiently listening to my colleagues
make their opening statements. And as a Member from the West,
we've been hearing about these problems for years. Because
every time we have a Presidential election we worry in the West
about what will happen to our voting when exit polling from the
East comes in. It has been an issue for many years. It will
only I think serve to increase--as you see, the recent census
data that shows population increasing in the Sunbelt and
particularly in the West.
Now, I think--I've heard all my colleagues. I think we
could sit up here and argue for the next 4 years what exit
polling did. Did it depress votes in the Panhandle? Did it hurt
votes in Colorado or Arizona or California or Oregon or
Washington? Many of these issues will be made moot by different
kinds of balloting that we're having.
I just was looking at my newspaper clips. The Colorado
legislature is about to pass a bill for mail-in balloting like
we have in Washington and Oregon, and I think that will make a
lot of the projecting that the networks are doing an obsolete
task and will have to find different ways to do it.
What I do think, though--and also I think that it would be
terribly wrong of the U.S. Congress to trample on the first
amendment rights of the media in order to try to solve a short-
term problem. And there is no doubt about that in my view.
I do think that there are some things we can do, Mr.
Chairman. I think that that's what this committee should focus
on. We should focus on election reform. We should focus on
looking at closing of polling places and what time we do that.
We should focus on ballots, and all of those things that can
serve to give Americans confidence in the polling process.
That's what we should be doing.
Finally, I will say I don't think this is a problem solely
of the media. I think it's just as much a problem as ourselves,
the public. I would suggest that we as citizens of this country
victimize ourselves by our own need for urgency. In this era of
rapid communication and the insatiable need for instant
information, there are times when I think we need to pause and
evaluate the risks inherent in our demands to get access to
information. That's not something the media can do by itself.
That's something the citizens of this country need to decide
for themselves.
So, you know, without venturing into the topic of
oversaturation, the media is responding to public demand; and I
think we need to ask ourselves, is this need to know the
results of an election as soon as possible denying millions of
voters their right to be a part of the process? If it is, what
can the American public do about it, not the U.S. Congress
putting unconstitutional restrictions on the media.
I'll yield back the balance of my time.
Chairman Tauzin. Thank the gentlelady.
Mr. Pitts is recognized for an opening statement--new
member of the committee.
Mr. Pitts. Mr. Chairman, the right to vote is the most
fundamental civil right in a democracy in representative
government; and when that right is taken away from anyone, the
proper response is shock and outrage. Whether we're talking
about Jim Crow laws or irresponsible or flawed news coverage,
the effect is the same. Somebody's right to cast a meaningful
vote is taken away.
Mr. Chairman, in my State of Pennsylvania the polls close
at 8 p.m. But in Presidential years the lines are so long in
some places that people are still voting as late as 9 and 9:30.
But you only have to be in line by 8. The line closes. But if
you're in line at 8, you're allowed to vote.
On election night in November, two networks called Vice
President--for Vice President Gore at 8:47. That means that
thousands of Pennsylvanians voted 45 minutes after some in the
media had already decided the winner. Interestingly, I think
calling the votes so soon violated even the industry's own
standard for prudent prediction of election outcomes.
Well, you might say that that would not have--it would have
been the same either way for Pennsylvania. After all, in our
system it's winner take all and--but you can't say that in
Florida. In the case of Florida, the Panhandle was open for a
whole hour after the rest of the States stopped voting. How
long were the lines at closing time in the Panhandle? It seems
to me quite likely that the media called Florida for the Vice
President, what, an hour, perhaps 2 hours before everyone was
done voting? And the networks were wrong. How many voters gave
up and went home? How many voters thought the race over and
their vote didn't count? One hundred? Two hundred? One
thousand? Just a couple hundred voters out of thousands would
have made a huge difference in the month that followed.
Clearly, Mr. Chairman, we have no business calling
elections until people are done voting. It's not just a problem
of accuracy. It's a problem of ethics. The right to vote is
sacred in this country, and it deserves the very highest level
of protection that we can give it. By any measure, what
happened that night in November was a reporting disaster. It's
our responsibility, it's our duty, it's the duty of all of us,
especially the media, to make sure that it never happens again.
I thank you, Mr. Chairman, for calling this hearing. I
thank the witnesses for appearing today.
Chairman Tauzin. Thank the gentleman.
Now recognize another new member of our committee, Mr.
Doyle, for an opening statement.
Mr. Doyle. Thank you, Mr. Chairman. I want to thank you for
convening this hearing to further examine the events that took
place at the network level on election night.
As was made clearly evident in the election night and post-
election night coverage, the networks and, by extension, their
election night coverage policies have played and will continue
to play a significant if not integral role in the entire
process of electing a President. That fact in and of itself
warns that we must strive to place the utmost importance on
accuracy as our first and foremost concern.
Mr. Chairman, much of what I was going to say in my opening
statement has been said many times over; and in the interest of
finally hearing our panelists some time today I will ask
unanimous consent that the remainder of my remarks be inserted
in the record and yield back my time.
Chairman Tauzin. Without objection so ordered.
Mr. Terry is recognized for an opening statement.
Mr. Terry. I'll yield back.
Chairman Tauzin. Mr. Walden.
Mr. Walden. Thank you, Mr. Chairman.
I, too, want to thank you for both this hearing and the way
this hearing has been laid out and organized. Because I think
really the thrust here is to look at the Voter News Service
model and the formula and what worked and what didn't work.
It's not to attack the first amendment.
I've spent 15 years in the radio business myself as an
owner and operator. I have been on both sides of this table,
actively reporting election report results on the election
night, gathering them at the courthouse, broadcasting them over
the air, coordinating that coverage, as well as having been a
candidate. I'm aware of how tricky it can be on election night,
how information you get initially may not be the right
information and how careful you have to be at reporting that
information. Numbers get transposed; election results get
confused; clerks sometimes make errors; partial returns may not
reflect the overall outcome of a precinct. So it is very
important for those of us who are in the media to make sure the
information we give out to our viewers and listeners and
readers is accurate.
I also believe that we have a great obligation to make sure
that as election results are put forth, as projections are
given that we don't somehow influence the outcome, make the
news, if you will.
I happen to represent the Second District of Oregon. One of
my predecessors was Representative Al Ullman, who was defeated
in the 1980 election. I was a press secretary of the campaign
of the candidate who defeated him. I remember very well our
feeling when the networks claimed that Jimmy Carter had lost
and the effect that had on the West Coast. Even the Washington
Post I note reported Rep Al Ullman of Oregon, Chairman of the
House Ways and Means Committee, won't be back to Congress next
year. He lost by 1 percentage point. He blames it on the time
zone effect.
The time zone effect is real. And I was disturbed, I guess,
by some of the information--I think it was in CBS Review--where
one of the authors concluded there is no research proving
western voters are dissuaded from voting by results in the
other States. And yet I have before me, Mr. Chairman,
information from the Secretary of State of California, Bill
Jones, a letter from 1998 also signed by our former Democrat
Secretary of State Phil Keisling and others from around the
country, pointing out that, based on Election Day surveys in
California alone, turnout dropped by 2 percent after the
announcement in 1980. According to a study by the Field
Institute, 10 percent of those questioned blamed the news media
projections for their failure to vote.
Chairman Tauzin. For the sake of the record, would the
gentleman offer that into the record?
Mr. Walden. I would be happy to, Mr. Chairman.
Chairman Tauzin. Without objection, it is so ordered.
[The letter follows:]
Secretary of State
State of California
April 10, 1998
Dear Network Executive:
Last October, California Secretary of State Bill Jones and six
fellow colleagues who represent the contiguous western states, wrote to
you urging the networks and other news media from reporting exit
polling and early projections during last November's presidential
election. As was explained to you then, every four years some 50
million Americans, who live in the Western United States, are
essentially delivered the message that their vote may not mean as much
as their fellow citizens who begin their day three hours earlier on the
East Coast.
Each presidential election year Americans in Arizona, California,
Idaho, Montana, Nevada, Oregon, Utah, Alaska and Washington have their
vote devalued by early news media projections, specifically those based
on exit polling information and not concrete results as reported by our
fellow secretaries of state.
Travelling throughout our respective states last fall, prior to
Election Day, we Secretaries of State heard from the voters directly
their concern about this important issue. The message from our voters
is clear. They want an Election Day that is not predetermined by
polling and predictions.
As you recall, in 1980, when ABC, CBS and NBC projected Ronald
Reagan as the winner shortly after 5:00 p.m. Pacific Standard Time
(PST), nearly three hours before the polls closed on the West Coast,
followed shortly after by President Carters concessions speech, it had
a chilling effect on voter turnout. Based on Election Day surveys in
California alone, turnout dropped by about 2 percent after the
announcement in 1980. According to a study by the Field Institute, 10
percent of those questioned blamed the news media projections for their
failure to vote.
This year, once again election officials and campaign workers
reported that many voters left the polls without casting their ballots,
phones in elections offices stopped ringing and get-out-the-vote drives
fell apart as a result of early exit polling and projections.
In fact, in 1996, the problems associated with projecting contest
winners based on early exit polling results was not confined solely to
the West. As our colleague from New Hampshire, Secretary of State Bill
Gardner, can attest, the news media organizations erred in their
prediction of the winner of the state's U.S. Senate election. That is a
mistake that we can assure you no one wants to see repeated.
Although we concede that on Election Day there was little doubt as
to the anticipated outcome of the 1996 presidential election, there
were a great many state and local races that experienced difficulty in
turning out their voters following premature news projections that the
presidential race was over. The Western State's candidates and voters
deserve to play on the same level playing field as their East Coast
counterparts.
We must not allow our Election Day to become the equivalent of a
drive-through democracy, where expediency is gained at the expense of
fairness. Every American--regardless of where in the country he or she
calls home--is entitled to make informed and independent decisions,
uncomplicated by prejudicial reports that ``the election is over,''
based on incomplete information from one or two geographic regions.
We, the undersigned Secretaries of State, recognize the negative
impact of media exit polls on voter turnout in the Western States
during presidential elections. Therefore, we request that all media
networks join together to refrain from releasing any presidential exit
polling results for the 2000 election cycle until the polls close in
our Western States (PST). On behalf of the millions of voters who call
the great American West home, we urge you not to abridge their right to
participate in the process of electing the president and congress of
the United States by broadcasting or reporting early projections on
Election Day.
Thank you for your consideration.
Sincerely,
Bill Jones, California Secretary of State; Olene Walker, Utah
Lieutenant Governor; Ralph Munro, Washington Secretary of
State; Mike Cooney, Montana Secretary of State; Pete T.
Cenarrusa, Idaho Secretary of State; Phil Keisling, Oregon
Secretary of State; Betsy Davis Beamer, Virginia, Secretary of
the Commonwealth; Kathleen E. Arnold, Secretary of the District
of Columbia; Eric Clark, Mississippi Secretary of State; Scott
Moore, Nebraska Secretary of State; Jim Bennett, Alabama
Secretary of State; Antionio O. Garza, Jr., Texas Secretary of
State; Sharon Priest, Arkansas Secretary of State; Jane Dee
Hull, Arizona Secretary of State; Elain F. Marshall, North
Carolina Secretary of State; Joyce Hazeltine, South Dakota
Secretary of State; Joan Anderson Grove, Minnesota Secretary of
State; Candice S. Miller, Michigan Secretary of State; Paul D.
Pate, Iowa Secretary of State; Victoria Buckley, Colorado
Secretary of State; Fran Ulmer, Lieutenant Governor, State of
Alaska; Stephanie Gonzales, New Mexico Secretary of State;
William F. Galvin, Massachusetts Secretary of State; Dean
Heller, Nevada Secretary of State; Edwin J. Freel, Delaware
Secretary of State; William Gardner, New Hampshire Secretary of
State; Lewis A. Massey, Georgia Secretary of State; Yvette
Kane, Pennsylvania Secretary of State; John T. Willis, Maryland
Secretary of State; Bob Taft, Ohio Secretary of State;
Alexander F. Treadwell, New York Secretary of State; Miles
Rapoport, Connecticut Secretary of State; James Milne, Vermont
Secretary of State; Ken Hechler, West Virginia Secretary of
State; Riley Darnell, Tennessee Secretary of State; Dan
Gwadosky, Maine Secretary of State; James R. Langevin, Rhode
Island Secretary of State; Rebecca McDowell-Cook, Missouri
Secretary of State; Lonna R. Hooks, New Jersey Secretary of
State; Diana J. Ohman, Wyoming Secretary of State; Sandra B.
Mortham, Florida Secretary of State; Douglas La Follette,
Wisconsin Secretary of State; Jim Miles, South Carolina
Secretary of State; W. Fox McKeithen, Louisiana Secretary of
State; John Y. Brown III, Kentucky Secretary of State; Ron
Thornbaugh, Kansas Secretary of State; Tom Cole, Oklahoma
Secretary of State; Al Jaeger, North Dakota Secretary of State;
George H. Ryan, Illinois Secretary of State; Sue Ann Gilroy,
Indiana Secretary of State; and Mazie K. Hirono, Hawaii
Secretary of State.
Mr. Walden. As for Florida, I notice one of my colleagues
suggested that there was--the early projections were accurate,
and there was no reason to question them. And yet their own
analysis by the networks and by VNS later determined all kinds
of problems with that analysis that led up to that projection,
including a VNS keypunch operator who had entered incorrectly
voter count data which had the effect of making it appear that
Vice President Gore had won 98 percent of the Duval County vote
tabulated to that time.
It's clear there were problems. I'm not here to castigate
those who are the anchors getting it over the earpiece, trying
to make the best call they can. I think what we have to do is
look at the data that led up to that so we aren't accidentally
affecting the outcome of elections elsewhere and putting out
bad information that's neither good for the broadcast and media
nor for those who are watching.
So, Mr. Chairman, with that I yield back the balance of my
time and thank you for your efforts on this hearing.
Chairman Tauzin. I thank the gentleman.
I think for our final opening statement the Chair
recognizes Mr. Buyer, another new member of our committee.
Mr. Buyer. Thank you, Mr. Chairman. You're correct. Being a
new member of the committee, the years that I served here on
Armed Services and Judiciary we never had this practice of all
members of a committee speaking while we've invited witnesses
to come testify before us. If this is the culture of the Energy
and Commerce Committee, obviously it's one I don't know about.
If it's a common practice of this committee over the years,
it's one I find distasteful. I find it distasteful because
we've invited a lot of people here to come testify at 11, and
here it's 1:30, and they have not yet begun to testify.
Now maybe it's members themselves, when they have the
opportunity to speak in front of cameras, they get a little
excited and they want to pontificate. But I'm a little bothered
by it.
The only comment that I am going to make is this: I'm
bothered by the rancor in this matter. Now whether it's
Republicans or Democrats I'm bothered by it. Because there is
one general concern that all of us have to agree on. This
comment about the first amendment right, you know, we
collectively have decided to live in a Republic, not a
democracy, under the rule of law to ensure that our society has
civil order. So even though we have a Bill of Rights, none of
those rights are truly absolute. So when someone begins to act
irresponsibly, it is an invitation for government intrusion.
So I'm bothered if someone operates in our society in
whatever industry, from Hollywood or to whomever, and they
think that they have certain rights that will protect them and
therefore they can act irresponsibly, I'm bothered. And I'm
equally as bothered that government then would somehow then try
to step in and be intrusive. So I am anxious to hear from
witnesses who have been very patient. I want to thank you for
doing that. Because I think you also want to be responsible and
accountable to yourselves.
Thank you very much. I yield back the balance of my time.
Chairman Tauzin. The Chair thanks the gentleman for
yielding.
The Chair would point out to the gentleman, as a new
member, that it is the rule of the Energy and Commerce
Committee to permit members to give opening statements. It is a
right to speak. There is, however, no obligation of any member
to speak. Members have the right to waive that right and to
insert their written oral statements into the record. The Chair
has always encouraged members to do that, but we will always
respect members' rights under the committee.
[Additional statements submitted for the record follow:]
Prepared Statement of Hon. Charles F. Bass, a Representative in
Congress from the State of New Hampshire
I thank the Chairman for providing the opportunity for this
committee to exercise its oversight responsibility and appreciate the
efforts of the witnesses.
On election night, television news networks Incorrectly and
prematurely reported the results of the race for President. The
networks heavy reliance on flawed data provided by the Voter News
Service and competition to declare results was an embarrassment to the
networks and a disservice to their viewers.
Although, there is no evidence that the results reported by Voter
News Service and the television news networks were intentional or in
any way deliberate attempt to affect ongoing voting. Nevertheless, I am
concerned about the network's heavy reliance on flawed data, their
insatiable urge to be first, and the effect early and inaccurate media
reports have on voter turnout in states where polls have not yet
closed. In addition, that the errors so heavily skewed in one political
direction ought to raise questions about the fairness and accuracy of
all the systems currently employed.
This is not the first time the networks have incorrectly predicted
the outcome of an election. Many New Hampshire residents and those who
follow our state's politics will recall that in 1996, the networks
falsely predicted that the Democratic challenger had defeated incumbent
Republican U.S. Senator Bob Smith.
I look forward to studying the proposed solutions to the problem
and appreciate the self-examination the networks and news services have
undertaken. The leadership they have demonstrated in setting new
guidelines for reporting election coverage encourages me to believe the
events of last November will not be soon repeated.
______
Prepared Statement of Hon. Ed Towns, a Representative in Congress from
the State of New York
Thank you Mr. Chairman. I too watched television on election night
and rode the all night roller coaster on November 7, 2000. I recall my
feelings of joy, disappointment, and feelings of great hope and
happiness as the Networks changed the call on Florida once again.
I too wish that the Networks had been 100% accurate in their
initial projections on election night, but both human and mechanical
mistakes were made which compounded the problem. While technology has
improved our lives in countless ways, it has also given us a sense of
entitlement to instant news, music and other forms of content and
information sometimes sacrificing the facts in order to break a story.
And while I think that is an issue that should be addressed, it is
a discussion for another day.
The networks should be held--and in my opinion--are being held
accountable. I am not of the opinion that these news organizations
knowingly distributed false information to the public, and it is worth
noting that news organizations are held to the higher authority of
public confidence and public opinion. It should also be said that all
networks have an interest in giving viewers the best possible product
due to the incredible amount of competition that exists in today's news
marketplace.
I am pleased that the networks worked hard on these studies and
policed themselves in a responsible manner, because it is not easy to
openly and publicly criticize yourself. The American public deserves
the best possible news service and each of our broadcast witnesses has
an incredible responsibility to protect the right of the Free Press.
Lastly, it has been said that the news organizations acted in a
biased nature that suppressed votes in later time zones, specifically
on the West Coast. Personally, I would welcome the opportunity to
discuss voter suppression in America today. If we are to discuss these
issues, then I would respectfully suggest that the witness list be
expanded to the thousands of voters who were disenfranchised throughout
Florida and in St. Louis, Missouri in addition to the countless other
locales across the country which may have gone unreported. There is a
vast difference between announcing winners and losers from exit polling
data--which is the ultimate responsibility of the networks, not
Congress--and the practices of intimidation and deception which
unfortunately still permeate our electoral process. The problem here is
not the networks, but zealous political operatives who practice voter
suppression as a trade.
Again, I am pleased that the networks stepped forward to claim
responsibility for their actions and I plan to keep a close eye on the
networks to make sure that they live up to their promise of reforming
the way elections are reported on in the future. Coupled with that,
however, I will do my part to ensure that each and every citizen of
this country has the right and ability to make their vote count in
America. Thank you and I yield back the balance of my time.
______
Prepared Statement of Hon. Frank Pallone, a Representative in Congress
from the State of New Jersey
Mr. Chairman, even before all the votes were counted in last year's
Presidential election, Republican Members of Congress were calling
press conferences, where they recited their old mantra of a liberal
media bias. They also went so far as describing a conspiracy on the
part of the networks to ensure victory for Al Gore.
But, Mr. Chairman, what my friends on the other side of the aisle
did not realize at the time and what the facts of the evening soon
revealed, was that perhaps the largest media bias was taking place at
the FOX News Channel. There, the first cousin of Governor George W.
Bush was calling the shots on which states went to each of the
candidates. If you want to talk about a conflict of interest, look no
further than the fact that a blood relative of then-Governor Bush was
the first to call the election for the Republicans. John Ellis, George
W. Bush's first cousin, was manning the decision desk at Fox News, one
of the nation's major news organizations. According to Ellis himself,
he spoke with candidate George W. Bush five times throughout election
day. He also spoke several times to his other first cousin, Florida
Governor Jeb Bush.
Later in the evening it was Ellis who called the election first for
Bush at 2:15 a.m. Within five minutes all the other networks followed
suit. What Mr. Ellis and the other networks did not realize was that it
was really Mr. Gore that should have been awarded the election. It
wasn't until many of the counties in Florida decided not to accurately
count all the votes cast on November 7, that Mr. Bush could be declared
the winner.
Since the Networks made this erroneous call at such a late hour,
many people around the country went to bed incorrectly assuming that
George W. Bush was the president-elect. The next morning, even after
all the networks had retracted their calls, the public perception was
still that Bush was the President-elect and Gore was seeking to
overturn the results, rather than the fact that the contest between the
two candidates was too close to call. If the networks never called this
election for Bush in the early hours of November 8, it is my belief
that the public would have demanded a fair and accurate count in
Florida, which I believe would have made Al Gore our 43rd President.
Clearly mistakes were made on election night, and those mistakes
had a huge impact on public opinion.
I am sure we can all agree that it is now the job of the networks
and the Voter News Service (VNS), through internal reviews that have
been completed and now must be implemented, to ensure that what
happened last year never happens again. But, if you think about it the
networks first numbers were correct. Exit polls, which measure a
voter's intent rather than what was later counted, showed a clear
victory for Al Gore. So rather than debating whether the media is
liberal or conservative, what we in Congress should really be focusing
on is the fact that not all American's votes were counted.
Chairman Tauzin. The Chair is now pleased to welcome our
first panel. As Mr. Buyer said, I apologize for the long delay,
but those are the rules of our committee. And the Chair under
the rules of the committee must inform the witnesses, of
course, that this is an investigatory hearing. As such, it is
the custom--in fact, tradition and practice of our committee
when we are in this mode to take our testimony under oath. So
that I will ask the witnesses in--prior to your testimony to in
fact stand with me while I administer the oath.
The first witness on the first panel will be Ms. Joan
Konner, a Professor of Journalism and Dean Emerita of the
Graduate School of Journalism, Columbia University, New York,
New York.
Ms. Konner, I welcome you. You are aware that the committee
is holding an investigatory hearing and in doing so the
practice is to take testimony under oath. Do you have any
objection to taking--giving your testimony other oath?
Ms. Konner. I do not.
Chairman Tauzin. The Chair advises you under the rules of
the House and the rules of the committee you're entitled to be
advised by counsel. Do you desire to be advised by counsel
during your testimony today?
Ms. Konner. I do not.
Chairman Tauzin. In that case, would you please indeed
raise your right hand; and I will swear you in.
[Witness sworn.]
Chairman Tauzin. You are now under oath. You can now give a
5-minute summary of your written statement, and we welcome it
right now.
TESTIMONY OF JOAN KONNER, PROFESSOR OF JOURNALISM AND DEAN
EMERITA, GRADUATE SCHOOL OF JOURNALISM, COLUMBIA UNIVERSITY
Ms. Konner. Thank you, Mr. Chairman. Thank you, members of
the committee. I am Joan Konner, former dean of the Columbia
Graduate School of Journalism and a professor there now. Before
going to Columbia as dean, I was a long-time television
producer in commercial and public television. I worked in news
and public affairs for NBC News for 12 years and in public
television for 12 years. I have been a news and documentary
writer, reporter, producer, executive producer, a program
director, a vice president of the public television station in
New York, and president of an independent television production
company.
I was asked by CNN, along with Jim Risser and Ben
Wattenberg, to look at what went wrong in its television
coverage of the Presidential election 2000. Our report,
``Television's Performance on Election Night 2000,'' a report
to CNN, has been submitted in full to the committee; and I've
been told that it will appear as part of the record of these
hearings.
Chairman Tauzin. Would the gentlelady suspend just a
minute. She reminds that I need to do that. And at this point,
the Chair would ask that the reports of all the networks and
RTI and VNS be made an official part of the record of this
hearing. Is there any objection? Without objection it is so
ordered.
I thank the gentlelady. Proceed.
Ms. Konner. The CNN report and all of the other reports
that have been issued about election night reporting recognize
that something went terribly wrong just as everyone, including
the public, recognizes that many things went wrong with the
election process itself. CNN executives, correspondents, and
producers themselves describe election night coverage as a
debacle, a disaster, and a fiasco; and in our report we agree.
I would like to address these remarks to two main points,
the first is to the context of our report and the second to
some of its substance.
First about the context. It is important to note at the
outset that this is a report on journalism. We as journalists
and academics were asked by CNN to undertake an independent
review and to answer the questions: What happened on Election
Night 2000? Why did it happen and what might be done to prevent
such mistakes from happening again? Our inquiry, judgment, and
recommendations were based on the ideals, the principles, and
the best practices of journalism.
The report should be taken as an independent peer review of
the quality of the journalism, not as a political or legal
opinion or a statement of public policy. We believe that CNN
should be commended for being the only network to invite a
wholly independent outside evaluation of its events of its
election-day coverage in order to help improve its performance
in the future.
Our panel's criticism of CNN's performance that night was
based on journalistic principles stated in the report that the
central purpose of a free press and a democratic society is to
provide the public with information upon which the people can
form intelligent decisions concerning important public matters
on which they have the power to act; and that public affairs
journalism is the pursuit of truth in the public interest and
its major values are accuracy, fairness, balance,
responsibility, accountability, independence, integrity, and
timeliness.
Those are the standards that informed our judgments, and
they are the standards that define professionalism according to
the written codes of most mainstream organizations and the
journalists that work for them. We believe that all the
journalists involved in the election coverage at CNN subscribe
to those principles. Nevertheless, we have concluded that
because of several key factors, CNN along with the other
television networks failed in their core mission to inform the
public accurately about the outcome of the election.
Specifically, CNN and the other networks failed in reporting
election results in Florida which turned out to hold the key to
the outcome of the election.
We found and reported that the faulty journalism resulted
from excessive speed and hypercompetition, combined with
overconfidence in experts and a reliance on increasingly
dubious polls. We have stated that the desire to be first or at
least not to be consistently behind the others led the networks
to make calls unwisely based on sketchy and sometimes mistaken
information. We reported an impulse to speed over accuracy. And
we attributed that impulse to the business imperatives of
television news to win the highest ratings, which is not a
journalistic standard but a commercial standard.
Ratings, that is the size of the audience, drive the price
of commercials; and the commercials determine the bottom-line
profits of the corporations that own the network. Our report
found several flaws in the system setup to cover the election.
We questioned the overall consent of the Voter News Service,
which was the single source of information on data on which all
the networks relied. Voter News Service was set up as a
partnership among competing news organizations.
This unusual collaboration among competitors was conceived
principally as a cost-cutting measure, although pooling
resources enabled the networks to greatly expand their polling
reach. We believe relying on a single source of information
contradicts well-known, deeply entrenched best journalistic
practices. Relying on a single source eliminates the checks and
balances built into a competitive vote-gathering and vote
system. It eliminates the possibility of a second source for
validating key and possible conflicting information.
The concept of VNS also effectively eliminates competition
in the market for the establishment of a second system, and it
might also stifle journalistic enterprise. We further question
the purpose of then introducing the element of competition
through independent decision desks at each of the networks, all
of whom rely on the same data and information received at
exactly the same time. What results is a speed trap in which
all of the networks are doing their complicated calculations
under maximum competitive pressure in minimum time, usually
making their so-called projections minutes apart.
The compulsion to be first led CNN and others to project
results without checking other possible sources of information.
At the time the call for Bush was made, there were, in fact,
two other sources available: the Associated Press, which does
its own vote count, and the official returns of the States.
We have questioned what purpose this hypercompetition
serves, either journalistic or commercial. It does not serve
the public, the core mission of journalism. Our inquiry also
indicated serious flaws in the polling methods used by VNS,
including exit polling, outdated polling models, and outdated
technology. We note, as others have, that polls inadequately
take into account the growing number of absentee ballots and
early mailed ballots or the variations caused by a wide variety
of factors on non-responses to the quality of the
questionnaire. We know that polls in general are statistical
calculations, not factual realities. And as such they are an
imperfect measure of voter intent and voting, especially in
close elections.
Our recommendations include the following: that exit
polling no longer be used to project or call winners of States;
and that exit polling be used for analysis only; that returns
from sample or key precincts no longer be used for projecting
or calling winners. We believe that model precincts are subject
to too many errors and could lead to faulty calls. We recommend
that all calls be based on actual vote counts and that no calls
be made in States where polls are still open. We recommend that
no call be made until all available sources of information are
checked.
We recommend that the Voter News Service be reexamined,
repaired, or reinvented and that a second service be
commissioned to conduct parallel national polling. We note that
many of these recommendations would probably slow down the
process of reporting, and we believe that is a good outcome. We
believe that slowing down would improve network performance and
would visibly demonstrate that accuracy was more important than
speed in reporting on elections.
Our report expresses the view that the mistakes in the
reporting of the Presidential election, especially in Florida,
were damaging to journalism and to the country. The erroneous
early call for Gore and a later call for Bush declaring him
prematurely the next President based on faulty numbers
undermined the credibility of the news organizations and
distorted the real result of the election at that point.
Some have charged that the networks--some have charged the
networks with bias in their reporting, that is, deliberately or
unwittingly calling or withholding the results of the race to
benefit one candidate over another. We found no evidence to
support that view. We also found no convincing evidence that
calls made before polls were closed within a State or in
another State have an impact on voter turn out.
All of CNN's election coverage was made with the best
journalistic intentions. But mistakes were made, and they have,
along with other networks, contributed to the public atmosphere
of rancor during the first post-election events.
We thank CNN for being willing to undergo this painful
process of external peer review, a familiar and accepted path
to course correction in many other professions. CNN has already
announced policy changes that will help prevent such a lapse in
the future. It demonstrates a serious commitment to more
stringent standards in covering elections and self-restraint,
an example, we hope, other networks will follow.
[The prepared statement of Joan Konner follows:]
Prepared Statement of Joan Konner, Professor of Journalism and Dean
Emerita, Graduate School of Journalism, Columbia University
I am Joan Konner, professor of journalism at the Columbia Graduate
School of Journalism and former dean of the school from 1988 to 1997.
Before going to Columbia, as dean, I was a longtime television producer
in commercial and public television. I worked in news and public
affairs for NBC News for 12 years and in public television for 12
years. I have been a news and documentary writer, reporter and
producer, an executive producer, program director, vice president of
the public television station in New York and President of an
independent television production company.
I was asked by CNN along with Jim Risser and Ben Wattenberg to look
into what went wrong in its television coverage if the Presidential
Election 2000. Our report ``Television's Performance on Election Night
2000: A Report to CNN'' has been submitted in full to the Committee,
and I have been told it will appear as part of the record of these
hearings.
The CNN report, and all the other reports that have been issued
about election night reporting, recognize that something went terribly
wrong, just as everyone, including the public, recognizes that
something went wrong with the election itself. CNN executives,
correspondents, and producers themselves described election night
coverage as ``a debacle,'' a ``disaster'' a ``fiasco.'' In our report
we agree.
I would like to address these remarks to two main points. The first
is the context of this report; and second, to some of the substance of
the report.
First, about context:
It is important to note at the outset that this is report on
journalism. We, as journalists and academics, were asked by CNN to
undertake an independent review and to answer the questions: What
happened on Election Night 2000? Why did it happen? And what might be
done to prevent such mistakes from happening again?
Our inquiry, judgements and recommendations were based on the
ideals, the principals and the best practices of journalism. The report
should be taken as an independent peer review of the quality of the
journalism, not a political or legal opinion or a statement of public
policy. We believe that CNN should be commended for being the only
network to invite a wholly independent, outside evaluation of the
events of its election day coverage in order to help improve its
performance in the future.
Our panel's criticism of CNN's performance that night was based on
journalistic principals stated in the report:
. . . that the central purpose of a free press in a democratic society
is to provide the public with information upon which the people
can form intelligent decisions concerning important public
matters on which they have the power to act;
. . . that public affairs journalism is the pursuit of truth in the
public interest, and its major components are accuracy,
fairness, balance, responsibility, accountability,
independence. integrity and timeliness.
Those are the standards that informed our judgements and they are
the standards that define professionalism, according to the written
codes of most mainstream organizations and the journalists who work for
them. We believe that all the journalists involved in election coverag
at CNN subscribe to those principals. Nevertheless, we concluded that
because of several key factors, CNN, along with the other television
networks, failed in their core mission, to inform the public accurately
about the outcome of the election. Specifically CNN and the other
networks failed in reporting election results in Florida, which turned
out to hold the key to the outcome of the election.
We found and reported that the faulty journalism resulted from
excessive speed and hyper-competition, combined with an overconfidence
in experts and a reliance on increasingly dubious polls. We stated that
the desire to be first, or at least not to be consistently behind the
others, led the networks to make calls, unwisely, based on sketchy and
sometimes mistaken information. We reported an impulse to speed over
accuracy, and we attributed that impulse to the business imperatives of
television news--to win the highest ratings, which is not a
journalistic standard but a commercial standard. Ratings, that is the
size of the audience, drive the price of the commercials, and
commercials determine the bottom line profits of the corporations that
own the news networks.
Our report found several flaws in the system set up to cover the
election.
We questioned the overall concept of the Voters News Service, which
was the single source of information and data, on which all the
networks relied. VNS was set up as a partnership among the competing
news organizations. This unusual collaboration among competitors was
conceived principally as a cost-cutting measure, although pooling
resources enabled the networks to greatly expand their polling reach.
In our report we say that relying on a single source of information
contradicts well-known, best journalistic practices. Relying on a
single source eliminates the checks and balances built into a
competitive vote-gathering and polling system. It eliminates the
possibility of a second source for validating key and possibly
conflicting information. The concept of VNS also effectively eliminates
competition in the market for the establishment of a second system, and
it might also stifle journalistic enterprise.
We further question the purpose of then introducing the element of
competition through independent decision desks, all of who rely on this
same data and information received at exactly the same time. What
results is a speed trap in which all the networks are doing their
complicated calculations under maximum competitive pressure in minimum
time, usually making their so-called competitive projections minutes
apart. The compulsion to be first led CNN and others to project results
without checking other possible sources of information. At the time the
call for Bush was made, there were in fact, two other sources
available, the Associated Press, which does its own vote count, and the
official returns of the state.
We question what purpose this hyper-competition serves, either
journalistic or commercial. It clearly did not serve the public, the
core mission of journalism.
Our inquiry also indicated serious flaws in the polling methods
used by VNS, including exit polling, outdated polling models and
outdated technology. We note, as have others, that polls inadequately
take into account the growing number of absentee ballots and early
mailed ballots, or the variations caused by a wide variety of factors
from non-responses to the quality of the questionnaire. We note that
polls, in general, are statistical calculations, not factual realities;
and as such, that they are an imperfect measure of voter intent and
actual voting, especially in close elections.
Our recommendations included the following:
. . . that exit polling no longer be used to project or call winners of
states; and that exit polling be used for analysis only;
. . . that returns from sample, or key, precincts no longer be used for
projecting or calling winners. We believe that model precincts
are subject to many errors and can lead to faulty calls.
. . . We recommend that all calls be based on actual counted returns;
. . . and that no calls be made in states where polls are still open;
. . . We recommend that no call be made until all available sources of
information are checked.
We recommend that the Voters News Service be reexamined, repaired
or reinvented, and that a second service be commissioned to conduct
parallel national polling.
We note that many of these recommendations would probably slow down
the process of reporting and we believe that is a good outcome. We
believe that slowing down would improve network performance and would
visibly demonstrate that accuracy was more important than speed in
reporting on elections.
Our report expresses the view that the mistakes in the reporting of
the Presidential Election 2000, specifically in Florida, were damaging
to journalism and to the country. The erroneous early call for Gore and
the later call for Bush, declaring him prematurely the next President
based on faulty numbers, undermined the credibility of the news
organizations and distorted the real result of the election at that
point. Some have charged the networks with bias in their reporting,
that is deliberately of unwittingly calling or withholding the results
of a race to benefit one candidate over another. We found no evidence
to support that view. We also found no convincing evidence that calls
made before polls are closed within a state or in another state have an
impact on voter turnout.
All of CNN's election coverage was made with the best journalistic
intentions, but mistakes were made, and they, along with other
networks, created the public atmosphere for the painful post-election
events.
We thank CNN for being willing to undergo this somewhat painful
process of external peer review, the familiar and accepted path to
course correction in many other professions. CNN has already announced
policy changes that will help prevent such a lapse in the future. It
demonstrates a serious commitment to more stringent reporting standards
in covering elections and self-restraint, an example we hope other
networks will follow.
Chairman Tauzin. Thank you very much, Ms. Konner.
And now, Mr. Risser, if you will stand with me. You are
aware the committee is holding an investigative hearing and
when doing so it is the practice to take the testimony under
oath. Do you have an objection to taking testimony under oath?
Mr. Risser. No, sir.
Chairman Tauzin. The Chair advises you that under the rules
of the House and the rules of this committee you have the right
to be advised by counsel. Do you wish to be advised by counsel
during your testimony?
Mr. Risser. No I don't.
Chairman Tauzin. Raise your right hand as I swear you in.
[Witness sworn.]
Chairman Tauzin. Sir, you are now under oath. You may now
proceed with the summary of your testimony.
TESTIMONY OF JAMES V. RISSER, RETIRED DIRECTOR, KNIGHT
FELLOWSHIP PROGRAM, STANFORD UNIVERSITY
Mr. Risser. Mr. Chairman and members of the committee, my
name is James V. Risser. I am the former director of Stanford
University's John S. Knight Fellowships for Professional
Journalists, which is a midcareer sabbatical program for
professional print and broadcast journalists. I retired from
that position and Stanford faculty in September. Before going
to Stanford in 1985, I worked for 20 years as a journalist,
including 15 years here in Washington where I was the bureau
chief of the Des Moines Register.
As you know, I am one of the members of this committee.
You've just heard from Joan Konner. My statement is not long
but I'll try to make it even a little shorter since it might
repeat some of what she said. As you know, our report released
on February 22 concentrates on the most obvious problem of
election night, the two calls and two retractions made by CNN
and others with regard to Florida. But we believe our findings
and recommendations, both directed specifically to CNN, go
farther than just the Florida case in demonstrating more
broadly what is wrong with the current system of reporting
election returns. We also believe they apply in general to the
networks, not to CNN alone by any means.
I take CNN at its word that it prefers being accurate over
being first in the reporting of the State-by-State returns.
That's the only proper stance for an all-news network that
wants to keep its journalism at a high professional level and
its integrity in tact. The fact that CNN set up our committee
shows they do care about this.
But at the same time, intense competition does exist if the
television news business and the system created by CNN and the
other networks for election night allowed speed to gain the
upper hand. The networks believed that they could be both fast
and accurate, but on this election night at least they were
wrong. The reasons that they were wrong were--and again to
summarize some things that Miss Konner also just hit on--they
relied on one source they had jointly funded, Voter News
Service. The figures and data they got from VNS were not always
accurate. Exit polls certainly were not accurate enough and
were perhaps less accurate than in past elections due to
nonresponses and due to the rise in absentee voting. At the
same time, vote tabulation errors were made.
Third, the network decision teams who analyzed VNS data and
decided when a network should call a State were unable to or at
least did not adequately scrutinize that data. On the decision
team chaired by CNN and CBS, there was a failure to consult
other key sources. This combination of factors led the networks
to call Florida twice and retract twice.
As a result, the networks suffered a grievous blow to their
reputation for delivering timely and accurate news. It turned
out that voters could not, as CBS's Dan Rather assured them,
take their networks' calls to the bank. Instead, the networks
found themselves having to twice eat crow as CNN's Jeff
Greenfield said.
All of this for no real good reason, our committee
concluded. We believe Americans are more interested in having
the election returns reported accurately than they are in
whether one network comes in a few minutes ahead of the other.
Very few people know at the time which network is coming in
first and virtually none of them could remember today. I am one
who believes strongly in journalistic competition; and I am
also aware if one network consistently came in far behind the
others, its audience share might suffer. But what we're talking
about, at least in regard to this election, is a matter of a
few minutes' difference here and there in reporting the winner
of a State and viewers simply don't care about that.
Our committee concluded there are serious doubts about the
validity of using exit polls to project winners and that the
Presidential election is a sacred enough rite of democracy that
nothing more than accurate reporting is acceptable. As we said
in our report, ``Exit polls, whether accurate or not, are self-
generated news. Their use by television networks to project
election results is an attempt to forecast what is not yet
known, the actual vote count, but which will be known within a
few hours when the votes are counted.''
Thus we think the networks ought to slow down and get it
right. You have heard already our recommendations to CNN and by
implication to other networks the stopping of the use of exit
polls to project or call winners, instituting a system of
calling winners done only from actual counted returns in which
a State could not be said to have gone for a particular
candidate until enough votes have not been counted to make the
outcome in that State a certainty. No State should be called
when the polls remain open there. Exit polling, if continued,
ought to be studied to make it more accurate. It is a valuable
tool for learning a lot things about the electorate besides
just how they voted.
Voter News Service should be fixed if its use is to be
continued. Other competing sources should be drawn upon as
well. CNN's decision team should be strengthened and should
adopt higher standards based on more information before
recommending a call to the network. We also ask that CNN and
others make an effort to inform viewers what is going on, that
is, how it arrives at its decision to call States, where the
information comes from, what the uncertainties may be and so
forth.
And finally, CNN--and I would hope others--should simply
take more time to get it right, to show by action that the
networks truly favor accuracy over speed and do not wish to
impinge improperly on the election process. CNN's own
representatives can explain for you the response they made to
our report. While it does not adopt everything, we suggest it
goes a very long way toward meeting many of our most important
findings and recommendations and would put CNN on the road
toward doing the kind of job in the next election that I am
sure it wanted to do in this one. Thank you.
[The prepared statement of James V. Risser follows:]
Prepared Statement of James V. Risser, Retired Director, Knight
Fellowship Program, Stanford University
Mr. Chairman and members of the committee, my name is James V.
Risser. I am the former director of Stanford University's John S.
Knight Fellowships for Professional Journalists, which is a mid-career
sabbatical program for professional print and broadcast journalists. I
retired from that position and from the Stanford communication faculty
last September.
Before going to Stanford in 1985, I worked for 20 years as a
journalist, including 15 years here in Washington where I was the
bureau chief for The Des Moines Register.
I am one of three members of the independent committee that CNN
asked to investigate and write a report on the problems that CNN
experienced on Election Night 2000 in reporting the returns and the
outcome of the presidential election.
In our report, which was released on February 2, we concentrated on
the most obvious problem of Election Night--the two calls and the two
retractions made by CNN with regard to Florida.
But we believe that our findings and our recommendations, both
directed specifically to CNN, go farther than just the Florida case in
demonstrating more broadly what is wrong with the current system of
reporting election returns. We also believe that they apply in general
to the networks, not to CNN alone by any means.
The committee has access to the report, so permit me to summarize a
few points that strike me as the most important.
I take CNN at its word that it prefers being accurate over being
first in the reporting of state-by-state returns. That is the only
proper stance for an all-news network that wants to keep its journalism
at a high professional level and its integrity intact. The fact that
CNN set up a truly independent committee to evaluate its performance
shows that it cares about this, as does its response since our report
was issued.
But at the same time competition--intense competition--does exist
in the television news business. And the system created by CNN and the
other networks for Election Night allowed speed to gain the upper hand.
The networks believed that they could be both fast and accurate
but, on this Election Night at least, they were wrong.
The reasons that they were wrong were:
First, they all relied on just one source which they had jointly
funded, Voter News Service, to give them results from exit polls,
sample precinct returns, and county-by-county vote totals.
Second, VNS figures were not always correct. Exit polls were not
accurate enough, perhaps less accurate than in past elections due to
non-responses and due to the rise in absentee voting. Also, vote
tabulation errors were made.
Third, network ``decision teams,'' who analyzed VNS data and
decided when their network should call a state, were unable to, or at
least did not, adequately scrutinize that data. At least at the
decision team shared by CNN and CBS, there was a failure to consult
other available sources such as Associated Press vote tabulations and
the vote counts available on the web pages of state secretaries of
state.
This combination of factors led the television news networks to
call Florida for Vice President Gore (before polls had closed in the
western part of the state), retract that call, call Florida and the
election for Governor Bush, and then retract that call.
As a result, the networks suffered a grievous blow to their
reputations for delivering timely and accurate news. It turned out that
viewers could not, as CBS's Dan Rather had assured them, take the
networks' election calls ``to the bank.'' Instead, the networks found
themselves having to twice eat crow, as CNN's Jeff Greenfield said.
All of this for no good reason. Americans are much more interested
in having the election returns reported accurately than they are in
whether one network comes in a few minutes ahead of the other in
reporting the results. Very few people know at the time which network
is coming in first, and even fewer of them would remember today.
I am one who believes strongly in journalistic competition, and I
am also aware that if one network consistently came in far behind the
others in reporting election results, its audience share might suffer.
But what we're talking about really is a matter of a few minutes
difference here and there in reporting the winner of a state, and
viewers don't care about that.
Our committee concluded that there are serious doubts about the
validity of using exit polls to project the winners in each state and
that the presidential election is a sacred enough rite of American
democracy that nothing less than accurate reporting is acceptable.
As we said in our report, ``Exit polls, whether accurate or not,
are self-generated news. Their use by television networks to project
election results is an attempt to forecast what is not yet known--the
actual vote count--but which will be known within a few hours when the
votes are counted.''
We think the networks ought to slow down and get it right.
We recommended to CNN, and by implication to other networks, that:
The use of exit polling to project or call winners should be
stopped.
The calling of winners should be done from actual counted returns,
and a state should not be said to have gone for a particular candidate
until enough votes have been counted to make the outcome in that state
a certainty. The use of so-called ``sample precinct'' returns should be
limited to guidance in looking at and evaluating actual counted
returns.
No states should be called when any polls remain open in that
state.
Exit polling, which can provide very valuable information about the
electorate, should, if continued, be carefully studied to make it more
accurate.
Voter News Service should be fixed, if the partners decide to
continue its use, and other competing sources of information should be
drawn upon by the networks.
CNN's decision team should be strengthened and should adopt higher
standards, based on more information, before recommending a call to the
network.
The network should make a greater effort to inform viewers about
what is going on--that is, how it arrives at decisions to call states,
where its information comes from, what the uncertainties may be, and so
forth.
And finally CNN, and I would hope others, should simply take more
time to get it right, to show by its actions that it truly favors
accuracy over speed and that it does not wish to impinge improperly on
the election process.
CNN's representatives can explain for you their response to our
report. While it does not adopt everything we suggested, it goes a long
way toward meeting many of our most important findings and
recommendations and would put CNN on the road toward doing the kind of
job in the next election that I'm sure it wanted to do in this one.
Thank you.
Chairman Tauzin. Thank you very much, Mr. Risser.
Our next witnesses will be Ben Wattenberg, certifying
senior fellow of the American Enterprise Institute here in
Washington DC. Ben, if you will help me through this. You are
aware that the committee is holding an investigative hearing
and when doing so it has the practice of taking testimony under
oath.
Mr. Wattenberg. Yes, sir.
Chairman Tauzin. Do you have any objection to testifying
under oath?
Mr. Wattenberg. I do not.
Chairman Tauzin. The Chair then advises you that under the
rules of the House and the rules of the committee you are
entitled to be advised by counsel. Do you desire to be advised
by counsel during your testimony?
Mr. Wattenberg. I do not.
Chairman Tauzin. In that case, please raise your right hand
and I'll swear you in.
[Witness sworn.]
Chairman Tauzin. Sir, you are under oath and you may now
give a summary of your testimony.
TESTIMONY OF BEN J. WATTENBERG, SENIOR FELLOW, AMERICAN
ENTERPRISE INSTITUTE
Mr. Wattenberg. Thank you, Mr. Chairman and members of the
committee. I am Ben Wattenberg, I am a senior fellow at the
American Enterprise Institute. Over the years I have been
active in radio, newspaper, magazines, radio--excuse me,
television, book publishing as a reporter or columnist, an
editor, a publisher, and an author of a number of books about
politics and demographics. I wanted to begin my testimony by
reading the first paragraph of the preamble of our report,
which has been alluded to; but I thought it might be useful
because it pretty much sums up the feeling of the three of us.
And it goes this way: ``On Election Day 2000, television
news organizations staged a collective drag race on the crowded
highway of democracy, recklessly endangering the electorial
process, the political life of the country, and their own
credibility, all for reasons that may be conceptually flawed
and commercially questionable.''
And I would just like to dwell for a moment on that last
part, ``conceptually flawed and commercially questionable,''
because as we got into this study I know for myself I kept
asking myself what on earth are they doing and why are they
doing it? And as we all have heard of the motto ``no pain no
gain,'' what the networks were doing it seems to me was no gain
and plenty of pain. They weren't getting anything out of it.
They were competing with themselves to try to play beat-the-
clock in a way that was ultimately truly senseless in my
judgment.
Now I would like to present a thought to the committee
today that goes beyond our report and to the question and
beyond some of our answers as to what can be done about it,
that Joan Konner and Jim Risser have described clearly the
items we mention in the report. I have a further thought and
there's a short statement about it, which I would like to read
excerpts from.
Following the Election 2000 fiasco, the networks appeared
to have put in place procedures that will deal with the
problems of early release of State results, but the
Presidential election is a national election. There has been no
satisfactory agreement regarding the reporting of votes in
early time zones which may influence States with later polling
times. Unless such an agreement is reached on the so-called
East-West problem, the Nation may suffer through yet another
late election-night debacle in years to come.
A number of solutions have been examined, these include
uniform poll-closing times, 24-hour voting, full-weekend
voting, national holiday voting. All, to be sure, make some
sense but they are awful complex.
In the aftermath--so I want--I have come to try to think
this thing through as to how this second potential electorial
iceberg can be avoided. I think that CNN and other networks
have dealt with the State problem; they have not dealt with the
national problem, and they are keenly concerned, as they should
be, appropriately with first amendment problems. It seems to me
that as I understand it in the aftermath of Florida there will
be enough Federal legislation sending money to the States to
modernize their electorial systems.
Now, is there a way to use this vehicle to simply resolve
the East-West problem? And it would seem to me--to me it seems
so. Overall, States already hold votes. We were not offered the
results of absentee and early voting ballots until the polls
closed weeks or days later after those votes had been cast. The
vote of a voter who casts his ballot at 7 a.m. is not released,
typically, until 12 hours later at 7 p.m. when the poll close.
So my thought is that these new Federal grants, should they
come about, be made conditional: That the non-Western States
would not get grant money if they did not hold their vote
release until the close of West Coast polling times.
The networks seem to me duly chastened regarding the
promiscuous use of exit polls for projection purposes and would
then have no early raw votes to project from. And because the
votes would not be available in the first place, there would
seem to be no first amendment problem. It seems to me this
offers a simpler form of getting at this problem that many
people on this panel--and I'm sure that the network executives
and that our panel--have investigated. Thank you very much.
[The prepared statement of Ben J. Wattenberg follows:]
Prepared Statement of Ben J. Wattenberg, Senior Fellow, American
Enterprise Institute
By calling Florida wrong twice--first for Al Gore, then for George
W. Bush--television networks hit an avoidable iceberg that had been on
their radar screen for many years. This blunder contributed mightily to
the confusion and rancor that followed, which hasn't fully subsided.
Attention has been paid; that iceberg won't be hit again. But there is
another one looming--bigger, still ignored--that should be addressed by
Congress.
After the election, in an act of corporate courage, CNN
commissioned an independent panel to investigate what went wrong on
that infamous night. Along with two-time Pulitzer Prize-winner James
Risser and Joan Konner, former dean of the Columbia School of
Journalism, I was asked to serve on that panel. CNN President Tom
Johnson gave us the opportunity to question whomever we wished, to
obtain any documents we wanted and to write a report that would be made
public.
The report begins, ``On Election Day 2000, television news
organizations staged a collective drag race on the crowded highway of
democracy, recklessly endangering the electoral process, the political
life of the country, and their own credibility, all for reasons that
may be conceptually flawed and commercially questionable.''
CNN has examined our report and others; the material is on the CNN
Web site. They have responded with a list of reforms. These include not
using exit polls for projections in close races, not projecting a
winner in a state if the balloting shows that there is less than a 1-
percent margin and not projecting the winner in a state until all the
polls are closed within that state.
In short, the early projection of statewide results will be slowed
down appreciably. Such reforms, by CNN and the other networks, should
take care of any Florida-like situations in the future. But that leaves
the rest of the country. Let me go beyond our report.
Americans hold about a million separate elections in the course of
a four-year cycle. Only one, for president, is national and
continental, albeit conducted indirectly through separate winner-take-
all contests. In America, for practical purposes, that presidential
election includes four time zones.
The case has been made, passionately, that early reporting of
states in the East influences voting in the West. Democrats were
outraged in 1980 when the networks called the presidential race for
Ronald Reagan early in the evening. That announcement, said Democrats,
depressed Democratic turnout, negatively influencing some congressional
races as well as other ``down-ticket'' offices and referenda. In 2000,
it was the Republicans who were outraged, claiming that early state
calls for Gore depressed Republican turnout in the Florida Panhandle
(which is in a different time zone than the rest of the state), as well
as in western states.
Or consider the 2000 election if the electoral math had worked out
somewhat differently. Republicans might have claimed that closely
contested races in Oregon and Washington would have gone Republican if
the networks had not started talking about Gore's winning ways in the
East. (GOP Sen. Slade Gorton lost in Washington by only 2,229 votes.)
Clearly, as the networks now concede, an election jurisdiction with
two time zones, such as Florida, shouldn't be called until all polls
are closed. Why, then, should a national race be called before that
national jurisdiction has all its polls closed?
In the early 1980s, angry congressional hearings were held.
Unanimous resolutions were passed asking the networks to exercise
restraint for a couple of hours every four years during a delicate
moment in the democratic cycle. The networks refused. They had First
Amendment rights! No one was going to tell them how to cover an
election--certainly not politicians.
The academic judgments on the matter of early counting vary widely.
It is not known whether early election calling significantly influences
voting behavior while the polls are open. But surely there may be a
small effect. Florida taught Americans that small numbers of votes can
have enormous leverage and incendiary effect in a nation whose interest
groups have an all-purpose slogan: ``It's Not Fair.''
CNN and other networks have endorsed ``uniform poll closing''
legislation. This would jiggle ballot hours and daylight savings time
in order to get Americans voting together in real time. It is a complex
solution.
In the aftermath of Florida, there is an easier way, fully
respectful of First Amendment rights. It is quite apparent that there
will be a new law sending money to the states to modernize their
election systems. Hang the chads! Death to the dimpled ballots!
But such monies should be conditioned, as are most federal grants
to states. The states should get the funds only if they don't release
ballot counts until polls close on the West Coast, just as if this were
one large country with four time zones. That condition, coupled with
network restraints on early exit poll use, would avoid the next
iceberg.
Chairman Tauzin. Thank you, Mr. Wattenberg.
Our final member of this panel is Mr. Paul Biemer, the
Research Triangle Institute, in Research Triangle Park, North
Carolina, formerly introduced originally by the vice president
of our committee, Mr. Burr. He asked me, by the way, Mr.
Biemer, to ask you if you were named after the BMW or was the
BMW named after you.
Mr. Biemer. I wasn't named after a BMW.
Chairman Tauzin. Mr. Biemer, as is our custom of the
committee when holding an investigative hearing of the
committee, the practice has been to take the testimony under
oath. Do you have any objection to testifying under oath?
Mr. Biemer. No, I do not.
Chairman Tauzin. The Chair advises you under the rules of
the House and the rules of the committee you are entitled to be
advised by counsel. Do you wish to be advised by counsel during
your testimony today?
Mr. Biemer. No.
Chairman Tauzin. In that case, please raise your right hand
and I will swear you in.
[Witness sworn.]
Chairman Tauzin. You are under oath sir and if you will
please summarize your testimony.
TESTIMONY OF PAUL BIEMER, RESEARCH TRIANGLE INSTITUTE
Mr. Biemer. Thank you, Mr. Chairman, members of the
committee. My name is Paul Biemer. I am a senior----
Chairman Tauzin. I don't think your mike is on.
Mr. Biemer. Does this work?
Chairman Tauzin. We will switch mikes. That is one of the
reasons I'm asking the committee administration to help us
high-tech this committee room.
Mr. Wattenberg. Do I get it back later?
Mr. Biemer. We will share. Thank you, Mr. Chairman, members
of the committee. My name is Paul Biemer. I am a senior
statistician at the Research Triangle Institute. I hold a Ph.D.
In statistics and have 23 years of experience in server
methodology in statistics. The Research Triangle Institute is a
not-for-profit research organization located near Raleigh,
North Carolina.
In December 2000, the Voter News Service board of managers
contracted with RTI to review VNS's data collection procedures,
estimation methodology, and other operations. The purpose of
this review is to provide an independent scientific assessment
of the causes for the mistaken calls in the Florida
Presidential vote and to make recommendations for improving the
election-day forecasting methodology to avoid mistaken calls in
future elections.
RTI assembled a team of 16 senior statisticians and server
methodologists to conduct this review, including myself. None
of us had any significant prior knowledge of or connection with
VNS operations, nor any experience with election-day voting
process prior to this review. The scope of our review was
confined to six areas, precinct sampling methods, preelection
research methods, projection models and calculations, data
collection procedures, quality control procedures, and an audit
of the information in the VNS report dated September 8, 2000.
Our review found the VNS methodology to be generally
statistically sound and adheres to good statistical practices
in many areas. However, there were several areas where their
methodology can be improved. We found no evidence of political
biases in their procedures and operations. However, we did find
some evidence of potentially important statistical biases.
And our findings are as follows: VNS's methods for
estimating the total vote, including absentee and early voters,
are subject to potentially serious statistical bias. VNS can
make better use of the available data on past elections with
regard to absentee and early voters to reduce this bias. The
exit poll samples are too small for producing accurate State-
level results and for providing unbiased ratio estimates of the
candidate vote-shares. The exit poll nonresponse rate was 49
percent. We believe this is a potentially important source of
statistical bias in the model projections.
We found the quality control procedures to be adequate for
most situations. However, there are still opportunities for
important errors to enter the system and dramatically change
the election results. The quality control methods for ensuring
that accurate data is received from precinct and county
reporters appear to be inadequate, particularly for close
races. The measures of uncertainty reported on the decision
screens sometimes do not reflect potentially important sources
of error in the VNS system. Thus the true uncertainty in the
estimates may be understated. The information provided of the
decision screens is prone to misinterpretation and the rules
used for election decisionmaking are inappropriate, given the
continual flow of data into the process.
In summary, we believe that the errors that led to the Gore
call in Florida and then the late-night shift to Bush were the
products of a number of system errors that tended to work in
concert at various points in the evening toward favoring one
candidate over the other. Stricter quality controls and quality
standards could prevent this from occurring in future
elections. The measures of uncertainty provided on the decision
screens underestimate the true total error in the estimates.
Thus the risk that an election analyst will call an election
erroneously could be substantially higher than indicated by the
information on the decision screens. The complexity and the
amount of information provided on the screens increase the risk
of the misinterpretation of the election results.
Finally, we developed five key recommendations for VNS.
One, improve the methodology for estimating the effect of
absentee votes on estimates of the candidate vote differential;
two, improve the methodology for estimating the outstanding
votes for candidates needed to win an election; three, improve
the measures of uncertainty for key election estimators to more
fully reflect the total variation and statistical biases in the
process; four, improve the quality control systems to quickly
and reliably signal the occurrence of an error; and, five,
develop better guidelines and decision rules for deciding to
either make a call or wait for additional data.
[The prepared statement of Paul Biemer follows:]
Prepared Statement of Paul Biemer, Research Triangle Institute
In December 2000, the Voter News Service (VNS) Board of Managers
contracted with Research Triangle Institute, a not-for-profit research
organization in North Carolina, to review VNS' data collection
procedures, estimation methodology, and other operations. These VNS
operations provide data for the news media to project the outcome of
U.S. elections. The purpose of this review was to provide an
independent, scientific assessment of the causes for the mistaken calls
in the Florida presidential vote and to make recommendations for
improving the Election Day forecasting methodology in order to avoid
mistaken calls in future elections.
RTI assembled a team of six senior statisticians and survey
methodologists to conduct this review: Drs. Paul Biemer, Ralph Folsom,
Richard Kulka, Judith Lessler, and Babu Shah, and Mr. Mike Weeks. None
of these six individuals had any significant prior knowledge of, or
connection with, VNS operations nor any experience with the Election
Day voting process. This apparently was a key requirement of the VNS
Board for selecting RTI for the review and ensured an objective and
impartial assessment of VNS operations.
The scope of the review was confined to the following six areas:
(1) evaluation of precinct sampling methods; (2) evaluation of pre-
election research methods, particularly with regard to any evidence of
political bias; (3) evaluation of projection models and calculations;
(4) evaluation of data collection procedures, particularly with regard
to any evidence of political bias; (5) evaluation of quality control
procedures throughout the system; and (6) audit of information in the
report written by VNS for the VNS Members dated December 8, 2000. The
Executive Summary 1 highlights the major findings in each of
these areas and summarizes RTI's recommendations based upon the review.
The details of our evaluation and recommendations are contained in the
full RTI report1.
---------------------------------------------------------------------------
\1\ RTI's full report and the Executive Summary were provided to
the Subcommittee. These documents are included as part of this
statement by reference.
---------------------------------------------------------------------------
Our review was based upon essentially three sources of information:
(a) documentation provided by VNS (the list of materials RTI received
appears in Appendix A of the full report), (b) a two-day meeting with
VNS staff during which we discussed (1) through (6) above, and (c) ad
hoc interactions between RTI and VNS during the review for
clarification purposes. A major focus of the review was the Election
Day system as it performed for Florida. Although the VNS Board
requested that we assess the system in at least five other states, data
from state and local officials for the selected states were not
available in time for our review. Therefore, an analysis of the data
for states other than Florida was not possible.
Due to necessary constraints on resources, time, and the scope for
this review, we did not independently verify all of the information in
the primary documents and data sources that were available to us.
However, there were several key areas where VNS information was
independently verified, and these are noted in our report.
Unless otherwise noted, the term ``bias'' used in our report refers
to statistical bias, not political, media or other bias. Statistical
bias is the expected difference between the value of an estimator and
the population value it is intended to estimate when these differences
are averaged over many, hypothetical repetitions of the entire election
forecasting process. It is caused by random errors occurring in the
data collection, data processing, or estimation processes that are
either uncontrolled by, or beyond the control of, the designer.
Political bias refers to deliberate errors in estimates, which force
the election outcomes in the direction of one political party or
another in order to bring about a desired election result. In our
review, we found no evidence of a political bias; however, we did find
evidence of statistical biases in the estimates, as we discuss below.
1. Precinct Sampling Methods
Our investigation revealed that VNS' precinct sampling and
associated estimation methods were well designed for estimating the
Election Day vote and generally follow standard statistical practices.
The oversampling of precincts that are designated as high-percent-black
units, which was done to improve the precision of the estimates, is a
widely accepted technique for sampling heterogeneous populations. To
account for the higher percentage of blacks in the sample using this
method, the estimate for blacks is given a lower weight in the
calculation of the overall estimates, so that the resulting estimator
of the total vote is technically unbiased. We verified that this was
done appropriately in the exit poll estimates.
While VNS' sampling methods for estimating the Election Day vote
are well-conceived, the methods for estimating the total vote including
absentee and early voters are subject to potentially serious
statistical bias. We also found that the exit poll samples typified by
the 45 precincts drawn for Florida are too small for producing accurate
state-level results. A small sample size not only reduces the precision
of the exit poll results, it also increases the statistical bias in the
ratio estimates of the candidate vote shares.
In reviewing the VNS estimation formulas, we had questions about
two approximations VNS had employed in their standard error
calculations. Therefore, we recalculated the standard errors for the
Gore percent advantage in Florida based on official vote totals with
and without these approximations and found no important differences in
the standard errors of the candidate vote differentials. Thus, we
believe the standard errors for the exit poll and VPA estimates
appropriately account for sampling variation.
2. Pre-election Research Methods
Overall, VNS' pre-election methods for collecting data from sampled
precincts are generally sound, but the ways in which they are used both
to perform edit and consistency checks on the election night system and
in formulating decisions could be substantially improved. The VNS
decision processes would benefit from both better information on the
absentee/early vote in each county/state and better use of the
available data on past elections from the research. Further, in spite
of an excellent track record, the VNS Research Department could benefit
from a general review of new methods and technologies by which it could
implement some of its protocols more quickly and accurately.
3. Projection Models and Calculations
The exit poll and VPA (Voter Profile Analysis) precinct estimates
are based on a well-designed sample survey that provides an excellent
source of timely information on the voting preferences of voters, their
characteristics, and their opinions. The Core estimation process
provides timely information on race outcomes and also attempts to
incorporate measures of uncertainty in the estimates, despite the fact
that the samples are not random.
However, the measures of uncertainty presented on the decision
screens eliminate some potentially important sources of error in the
VNS system and, thus, the true uncertainty in the estimates may be
understated. In close elections, the risk that an election analyst will
call an election erroneously could be substantially higher than the
information on the decision screens indicates, even if the analyst
correctly interprets the information. In addition, the information
provided on the decision screens is prone to misinterpretation, and the
rules used for election decision-making are inappropriate, given the
continual flow of data into the process. A much-simplified screen that
uses a sequential-sampling-type decision rule would better control the
error in forecasting a winner future elections.
4. Data Collection Procedures
VNS' field staffing operations worked as planned for the exit polls
and the NETS county reporting system. The problems were more frequent
in the operations for the VPA sample and the NETS precinct reporting
system. Nationally, reports were received from 98% of the exit poll
precincts, 84% of the VPA precincts, 62% of the precincts in the NETS
precinct reporting pool, and 100% of the counties (or county
equivalents) in the NETS county reporting system. However, these rates
are similar to those in past presidential elections.
We found training materials and the training process for the exit
poll interviewers to be thorough and appropriate. Based on a sample of
interviewer debriefing questionnaires, it appears that the exit poll
data collection protocol was implemented reasonably well in non-
problematic voting places--e.g., those with only one exit, a good
interviewer location close to the exit, and a cooperative polling place
official. The follow-up telephone conversations we had with 12 exit
poll interviewers reinforced the data in the questionnaires.
While many of VNS' data collection procedures worked as intended,
we have some concerns about their adequacy to produce data of
sufficient precision to call very close elections like the 2000
presidential election in Florida. We believe the most serious data
collection problem VNS had in the 2000 election was noncoverage of the
absentee vote and early voters in critical states such as Florida.
Another problem area is exit poll nonresponse (refusals and
misses). According to VNS, the average state-level response rate in
2000 was 51%; this compares with 55% in 1996 and 60% in 1992. A
nonresponse rate of this magnitude is a potential source of statistical
bias in the model projections if the voters who respond have voting
characteristics that are significantly different from those of
nonrespondents. In our two-day meeting with VNS staff, we were told
that the exit polls more often overrepresented Democrats than they do
Republicans. This effect could be the result of a statistical bias due
to nonresponse.
Finally, another area of concern is the absence of any direct
quality control check on the interviewers' data collection activities.
5. Quality Control Procedures
One of the strengths of the exit poll survey quality control system
is the repetitive nature of the reporting process. Another is the use
of overlaid precinct estimates of exit poll survey bias. This
indicator, if used appropriately, can provide valuable information
regarding the accuracy of the exit poll results and can sound a warning
if there are major problems with the exit poll data. The NETS (New
Election Tabulation System) makes good use of the data from similar
past races to check the counts provided by the county reporters. In
addition, the multiple reporting procedures of the NETS provide a self-
correcting feature in the process that can be effective for correcting
previous erroneous reports.
Despite the numerous inspections, verifications, and edit checks
that have become part of the VNS Election Day system design, there are
still opportunities for important errors to enter the system and
dramatically change the election results. For example, the number of
overlaid precincts in Florida was too small at the time of the Gore
call to be a reliable indicator of the exit poll bias. In addition, the
quality control methods for ensuring that accurate data are received
from precinct and county reporters is inadequate, particularly for
close races, and can be improved.
6. Audit of Information in VNS Report
A strength of the VNS report (dated December 8, 2000) is its broad
scope, which provides a good indication of the difficulty in mounting
the extremely complex process that culminates in the collection,
analysis, and reporting of data from diverse and dynamic sources in a
single 24-hour period. The authors took a broad view of what could have
gone wrong and attempted to determine if there was an error in the
procedures that contributed to the errors in the calls. This report
should be useful to VNS Members for gaining an understanding of why
errors were made in the Florida calls and to formulate ideas as to how
the system needs to be changed.
However, the report needs more documentation. Although
considerable work was done to assemble information, there is frequently
not a clear description of how this information was compiled. The
procedures and data that were used to prepare the individual reports
included in the December 8 report are largely not discussed. Many
statements are made that reference results that were assembled by VNS
staff without showing the supporting data. The report also inadequately
describes the contribution that the modeling and the use of partial
samples made to the error in the early call for Gore. This lack of
explanation of the modeling and lack of documentation on information
available for analysis also limit the reader's ability to formulate
ideas as to how the system needs to be changed.
7. Summary
In summary, we believe that the errors that led to the Gore call in
Florida and then the late night shift to Bush were the product of a
number of system errors that tended to work in concert at various
points in the evening toward favoring one candidate and then the other.
The major sources of error were: (a) estimation of the early/absentee
vote, (b) exit poll ratio estimator bias, (c) end of night outstanding
vote needed estimation, and (d) county-level reports. Stricter quality
controls and quality standards and improved estimation methodology
could prevent these errors from occurring in future elections.
In addition, we believe the measures of uncertainty provided on the
decision screens underestimate the true total error in the estimates.
Thus, the risk that an election analyst will call an election
erroneously could be substantially higher than indicated by the
information on the decision screens. The complexity and amount of
information provided on the decision screens increase the risks of
misinterpretation of the election results. In addition, we believe the
rules used for election decision-making are inappropriate, given the
continual flow of data into the process. A much simplified screen
format that uses a sequential sampling type decision rule would better
control the error in predicting the outcomes of future elections.
Our key recommendations for VNS are as follows:
1. Improve the methodology for estimating the effect of absentee votes
on estimates of the candidate vote differential.
2. Improve the methodology for estimating the outstanding votes needed
by candidates to win an election.
3. Improve the measures of uncertainty for the key election estimators
to more fully reflect the total variation and statistical bias
in the measurement process.
4. Improve the quality control systems to quickly and reliably signal
the occurrence of error.
5. Integrate the ideas of sequential analysis in the election decision
process and develop better guidelines and decision rules for
deciding to either make a call or wait for additional data.
Chairman Tauzin. Thank you very much, Mr. Biemer.
The Chair recognizes himself for a round of questions. Mr.
Biemer, you are making some very strong recommendations for
changes in which VNS models its exit polling and its data and
presents that data to the networks. Is that correct?
Mr. Biemer. Yes.
Chairman Tauzin. And the report of the three of you to CNN
basically says don't use exit polling any more. Which is the
better option? Can we trust that exit polling can be remodeled
at VNS to a point of high confidence at the networks and with
the American public, or should the networks heed your advice
and simply stop relying upon exit polling to predict results?
Ms. Konner. I think our report doesn't say don't use exit
polls. It says don't use exit polls for projecting elections,
that exit polls are very valuable instruments for analyzing how
voters vote and that they should continue to be done.
Chairman Tauzin. So what you are saying is that exit
polling is a very good tool for analysis that follows an
election, but you're basically saying don't use exit polls to
make these predictions and these calls. Mr. Biemer is, on the
other hand, recommending improvements in the VNS system which
the networks rely upon now. And I'm asking, basically, which is
the better approach. Mr. Wattenberg.
Mr. Wattenberg. My sense of the matter is that as Joan
Konner says is that exit polling is extremely valuable as an
analytical tool, and becoming almost worthless as a predictive
tool.
Chairman Tauzin. Almost worthless.
Mr. Wattenberg. Yes, sir.
Chairman Tauzin. Why is it getting worse?
Mr. Wattenberg. I will tell you why. Because of the move to
absentee balloting, early balloting, nonresponse rates which,
as Mr. Biemer says, are now down below 50 percent, and we saw
in Florida that the exit poll was off by 7.6 percent. That is
an enormous error. That is not, as the VNS people would like to
say, well, it is a one in 200 chance. That is a 1 in 10,000,
one in a million sort of a shot. And the only way it can be
fixed, Mr. Chairman, as I understand it, is to use standard
polling, telephone polling in advance of the election to find
out how the absentee voters and early voters voted; but the
standard preelection polling is in a bigger ditch than election
polling. They are getting only 20 percent of response of every
hundred people called. They are getting only 20 percent of the
people responding.
Chairman Tauzin. So your conclusion is that exit polling
itself is becoming less and less reliable as a predictor rather
than increasingly more reliable.
Chairman Tauzin. Mr. Biemer, do you want to comment on
that?
Mr. Wattenberg. And so is polling generally.
Chairman Tauzin. And so is polling generally.
Mr. Biemer.
Mr. Biemer. I think that exit polls can be useful for
calling races, and I think that the main lesson here is that we
have to look at all of the data. We have to look at the
measures of uncertainty associated with what the exit polls are
telling us. And the first rule is to get those measures of
uncertainty right so that if we have a projection that is being
made by an exit poll, but measures of uncertainty are large, we
hold back and we don't make the call. It is like a margin of
error. If you use the margins of error and let them direct you,
whether it is close or not----
Chairman Tauzin. Let's talk about that. VNS does establish
what we call ``crit numbers,'' which are the minimum
statistical numbers that should be reached in these surveys in
order to justify a call for one candidate or another in that
State; is that right?
Mr. Biemer. Right.
Chairman Tauzin. Now, one of the questions we were focused
on as we convened this hearing is the one your committee
actually addressed and that is, was the rush to be first with
the call more important than the responsibility to be right?
You've concluded that there was too much of a rush to be first
and the responsibility to be right perhaps may have suffered as
a result. I want you to explain that to me. If VNS is setting
these crit numbers and VNS is giving these crit numbers to
every one of the networks commonly, how is there a contest?
These decision desks at the networks actually make decisions in
front of or behind those crit numbers? Do they make decisions
to call a State for example before the crit numbers are
reached?
Mr. Biemer. Can I address that?
Chairman Tauzin. Please do.
Mr. Biemer. I think in the case of the Gore call the crit
numbers, the margins of error, on the screen were misleading
because it did not show the total error.
Chairman Tauzin. But I have given all of that----
Mr. Biemer. But if----
Chairman Tauzin. Did they, in fact, call States before the
crit numbers were reached in some cases?
Mr. Biemer. I don't think so. I don't think they----
Chairman Tauzin. We think we have evidence that that
happened. We want to get to that.
Mr. Biemer. I don't know about other States. I don't think
it was the case in Florida.
Chairman Tauzin. We have also know in some States, even
though the crit numbers were released, the calls were delayed
significantly; and we are wondering how that happened, and I
wondered if the committee had looked into that and what was
your conclusions about that.
Mr. Risser. I don't think we felt we had the expertise to
look into it in great detail. But we did ask about it, and our
report includes an appendix by the CNN-CBS decision team giving
their answer as to why each of those States was called.
Chairman Tauzin. I think we will get into that later with
the witnesses, and I believe my time has expired. I simply want
to put one thing on the record though: What were your
instructions from CNN when you were commissioned to do this
report?
Mr. Risser. Our instructions were to find out what went
wrong on election night, why it went wrong.
Chairman Tauzin. And you had full freedom to do that.
Mr. Risser. Absolutely.
Chairman Tauzin. Thank you very much. The Chair yields to
Mr. Dingell.
Mr. Dingell. Thank you, Mr. Chairman. Welcome to the panel.
Mr. Biemer, about the Research Triangle Institute, can you tell
me whether your work there indicates to you that there was any
intentional or unintentional bias that favors either party,
Republicans or Democrats?
Mr. Biemer. We found no evidence of any intentional bias.
Mr. Dingell. Does the rest of the panel agree with that?
Ms. Konner. Yes.
Mr. Dingell. The reporter doesn't have a nod button. You
have to say yes, please.
Ms. Konner. That was our finding.
Mr. Risser. Yes.
Mr. Wattenberg. Yes.
Mr. Dingell. Now, it can go both ways, does it not, the
bias sort of built into the system to be triggered by events
outside? Isn't that the way it works? It's a bias that goes
either way.
Mr. Biemer. Yes, the biases can go either way. There are
systematic errors.
Mr. Dingell. What is the correction for that?
Mr. Biemer. A correction?
Mr. Dingell. Yeah, what do we do to correct that? More
money? Better polling? More assistance to the agency that did
the work for the network?
Mr. Biemer. If you're talking about the bias in the exit
poll.
Mr. Dingell. How do you break the bias out of the system?
Mr. Biemer. There's a bias in the exit poll that I think
was caused by the so-called ratio estimator, which is a
technical bias associated with the type of estimator. One way
in which one can alleviate that bias or eliminate that bias,
this estimator uses a past vote. In the case of Florida, it
used the, I think, the 1998 gubernatorial past vote as a way of
correlating this past vote with the current vote to try to make
a projection. If another vote, the 1994 Presidential election--
I'm sorry, 1996 Presidential election had been used, you would
have gotten a very different estimator and the call for Gore
would not have been made.
So part of this is not to rely on one past vote or maybe on
several past votes to try to get some idea of how much these
estimates are bouncing around. And that would slow down the
calls and prevent, I think, what happened with the Gore call in
Florida. So that is the technical bias.
Then there's another bias associated with the nonresponse
bias of the exit poll. And as I mentioned, the nonresponse rate
is around 49 percent. If the respondents--if the people who
respond to the poll and the people who do not respond to the
poll have essentially the same split for one candidate or
another, there's no bias associated with that. The problem
comes in when the people who do not respond tend to favor one
candidate more than the other more so than the people who do
respond. And there's very little you can do I think to try to
alleviate that bias unless you can somehow increase the
response rates in such a way that it doesn't exasperate the
differential candidate split in the nonresponse.
So I think more research looking at ways to increase the
response rate in the exit poll might be one answer to that,
eliminating that bias.
Mr. Dingell. Would you suggest that if we forget all of
this business of statistical correction which appears to be
difficult and just set a higher risk level that that would
probably reduce or terminate the risk?
Mr. Biemer. I think if one can widen the error, thinking in
terms of confidence level, if one widens the bounds of the
confidence level, that is to say, increase the measures of
uncertainty, that it makes the estimate look less precise than
they currently do on the decision screens, what the effect that
will have is to slow down the process of calling an election.
And I think that may be one thing to investigate, whether or
not we can obtain measures of uncertainty that better reflect
the total error, not just sampling error but the nonsampling
error as well. And that way there won't be any significant
difference between the candidate vote-share and it will be too
close to call for more elections which, I think, will, I think,
slow down the process of calling the elections. So that is one
possibility.
Mr. Dingell. Would you say the right precincts or the wrong
precincts were used for sampling?
Mr. Biemer. The precincts were drawn randomly and it looks
like that randomization was done properly.
Mr. Dingell. It was done properly.
Mr. Biemer. Yes.
Mr. Dingell. Should they have used a larger sample?
Mr. Biemer. I think a larger sample would certainly have
solved a number of problems.
Mr. Dingell. It would have eliminated the bias.
Mr. Biemer. It would certainly have affected the bias, the
technical bias in the ratio estimator, because that estimator
was based on 45, 44 precincts. If we double that, I think we
would have gotten a more accurate estimator; and if the bias
would have been reduced, the technical bias would have gone
down with the increase in sample size. It wouldn't have reduced
the nonresponse bias due to the exit poll, however, which was
not a big factor.
Mr. Dingell. How much would it cost to fix this thing? And
that would mean that more money would have to be paid by the
networks to get to VNS to get a better work product, would it
not?
Mr. Biemer. If you doubled the sample, the field work would
increase by a factor of two; and I imagine the budget would
increase by the same factor, or close to it.
Mr. Dingell. My time has expired, Mr. Chairman. Thank you.
Chairman Tauzin. I thank the gentleman. The Chair would
recognize the gentleman from Florida, Mr. Stearns.
Mr. Stearns. Thank you, Mr. Chairman. Mr. Biemer, I want to
direct my attention to you. I've got this report here that you
folks prepared, and I specifically want to go to page 28 and
just read--so I think you have in front of you--``since the
bias tends to consistently favor the Democratic candidate, such
an adjustment could be quite effective for reducing this bias.
In addition to the point estimate of the bias, an estimate of
variance of the bias estimate should also be given to user so
they can evaluate how reliable the available information on
exit poll bias really is.''
And then when you go into another section of your report on
page 38 in the second paragraph, it says: ``We are not
convinced that nonresponse of this magnitude can be safely
ignored. Further, in our 2-day meeting with VNS staff, we were
told that the exit poll have consistently overrepresented,
Democrats which could reflect a bias from nonresponse.''
So I just want to clarify, in your written testimony you
indicate that VNS advised you that exit polls more often
represent Democrats than they do Republicans and that this
effect could be the result of statistical bias due to
nonresponse. If there is this bias, how can you properly
account for this?
Mr. Biemer. What we suggested in here was that if there
were a State, for example, that had a consistent bias over a
number of elections that favored, say, Democrats and we knew
that that was, you know, at least 1 percentage point or 2
percentage points, one could require that the margin of victory
for that Democratic candidate be greater than that, say, twice
what we think that bias is before we would call an election. In
other words, we would hold back calling the election if there
was any chance that the Democratic bias associated with the
exit poll could be causing the numbers to sort of tilt toward
the Democratic candidate. So one way we could do it would be to
try to work that into the decision rules. Now that would only
be the case if we knew that the State over a number of
elections consistently, the exit poll consistently was favoring
the Democratic candidate. I don't know how often that happens--
--
Mr. Stearns. Let me take that statement. Consistently, are
you saying it is unintentional or intentional.
Mr. Biemer. It is unintentional----
Mr. Stearns. How do you distinguish between unintentional
and intentional? How do you do that quantitatively not
qualitatively? Can you do it?
Mr. Biemer. I think what we have to do is there are a set
of procedures that interviewers follow to try to get the exit
poll data and we look at those procedures and say okay are
those procedures favoring one candidate or another or are they
going to be nonpartisan-type procedures. We looked at those
procedures and we didn't see any evidence that they would favor
one candidate or the other. It doesn't mean that the
respondent, the Democratic respondents, may behave differently
in terms of how they might respond. But from what we could see,
it wasn't anything we could see that the interviewers were
doing; and the procedures, you know, seemed to follow the
statistical practices and what we would consider to be good
statistical practices and conservative methods.
Mr. Stearns. Isn't it true these procedures have been used
for years?
Mr. Biemer. Right.
Mr. Stearns. So in your report you're talking about an
unintentional, to use your words, bias to the Democratic Party.
You have pointed out at least four times in your report--in
fact on page 50 it says under ``potential areas for
improvement,'' it says in fact it is well documented that exit
polls tend to be biased toward the Democratic Party. The so-
called Democratic bias referred to in earlier sections--the
other two, three sections I mentioned, thus some verification
of the respondents selection process would seem warranted. So
here we have you saying unintentional bias toward the
Democratic Party has consisted for years, the procedures have
been there, no one has continued to ferret this out and stop
it. So based upon what you said in page 50, how do you intend
to correct this in the future if they have not corrected it in
the past for all these years?
Mr. Biemer. Well, what has been done that I know of in the
past has been to try to increase response rates by providing
incentives to a respondent.
Mr. Stearns. What's a response rate mean?
Mr. Biemer. A response rate, essentially you take the
number of interviews, divide that by the number of interviews
plus the noninterviews. The misses and the people that refused.
So it is essentially the rate at which the people approached
respond to the exit survey. It is around 50 percent for the
exit poll. One way to try to reduce the nonresponse bias is if
it were zero nonresponse rate we would have zero nonresponse
bias. So one idea would be to try to make the nonresponse rates
smaller. Their attempts to do that, however, have actually made
the problem worse because with some experimentation they have
done they have found that these methods to provide incentive
tend to affect the Democrats more than Republicans, so they've
been trying to--VNS has been trying to work on this to
eliminate this bias.
Mr. Stearns. I thank you, Mr. Chairman. I think it would be
important for you or someone on the committee to bring up that
chart again to show the overestimate for Gore in 34 States plus
the District of Columbia. And that was, I think, a very telling
chart on what we've been talking about.
Chairman Tauzin. I thank the gentleman. I am intrigued by
your statement of incentivizing responses. I'm not sure what
you meant by that. If you want to explain that.
Mr. Biemer. What we typically do in survey work when we
have a problem with a low response rate and we recognize that
participating in a survey is a burden is that we offer an
incentive.
Chairman Tauzin. What do you mean? Tip them, pay them?
Mr. Biemer. Yes, we might give them a gift. Usually, this
is with no strings attached. We give them a gift like a pen. We
might even give them money. We might give them a dollar bill,
two dollars.
Chairman Tauzin. Cigarettes? The Chair will yield to the
gentleman from Massachusetts, Mr. Markey.
Mr. Markey. Thank you, Mr. Chairman, very much. Mr. Biemer,
in your testimony your analysis--you point out that you have
uncovered other major sources of errors at VNS, including the
exit poll ratio estimator, the end-of-the-night standing-vote
needed and county-level reports. In addition, the RTI study
highlights that the quality-controlling methods for ensuring
that accurate records from precinct and county reporters are,
quote, ``inadequate particularly for close races.'' Obviously,
the news directors and news anchors did not know of these
deficiencies.
Mr. Biemer. I don't know if they did or not.
Mr. Markey. Did you find any evidence that they did?
Mr. Biemer. I think in discussions with VNS we were told
that they do provide the decision desk analysts with
information prior to election night about things about the
absentee votes.
Mr. Markey. Let me read the quote from your own study here
on page five. You say: ``Measures of uncertainty presented on
the decision screens eliminate some potentially important
sources of error in the VNS system and thus the true
uncertainty in the estimates may be understated.'' I continue
to quote: ``In close elections the risk that an election
analyst will call an election erroneously could be
substantially higher than the information on the decision
screens indicates even if the analyst correctly interprets the
information.''
In other words, VNS is at fault here, not the network news
divisions. They were sending up faulty information without
giving the actual degree of true uncertainty as pointed out
here that existed in the information as it was being presented.
Isn't that accurate?
Mr. Biemer. I don't know who's at fault, but I will say the
statement we made there about the decision screens being
misleading in terms of uncertainties, that is correct. I
believe they are to some extent, and this is true when you look
at what happened in the exit poll in Florida. The measures of
uncertainty did not include things like the potential bias due
to the ratio estimator, the absentee bias. Those things weren't
reflected.
Mr. Markey. That is what I'm saying. So even if the
information was accurately interpreted, it could have still led
to erroneous conclusions; is that correct?
Mr. Biemer. Yes.
Mr. Markey. Do you think the news directors would risk
their careers based on something that could have led to
erroneous conclusions, or do you think that VNS was the source
of the problem and it wasn't properly communicated on the
decision screens that there was in fact a much higher level of
risk that was entailed in making projections based upon these
numbers?
Mr. Biemer. One of things we did not get into is what the
addition desk analyst--what their approach was for calling
these elections. All we did was look into what procedures VNS
used to produce these estimates, to collect the data, and put
it on the decision screens. What happens after that we don't
know.
Mr. Markey. Exactly. The point I am making here is the news
anchors don't have Ph.D.s in statistics. VNS failed them--VNS
did not give them the information in a form that was usable
that would protect the reputations of the news divisions. Don't
you agree with that?
Mr. Biemer. I don't know if I agree or not because I don't
know what information they got other than what's on the
decision screens. I understand that there was other
information----
Mr. Markey. You say right here in your report--you say in
your report, sir, you say measures of uncertainty presented on
the decision screens sent to the networks eliminates some
potentially important sources of error in the VNS system and
thus the true uncertainty. So you're saying that they sent
erroneous information.
Mr. Biemer. If they're only using the information on the
decision screens, they are being misled.
Mr. Markey. That has to be said here.
Mr. Biemer. If they are using the information beyond the
decision screens along with the information on the decision
screens, they might be correcting some of their
misinterpretation. I don't know about that because I don't know
the process that the decision analysts use.
Mr. Markey. Did VNS mislead the news directors and the news
anchors? Yes or no.
Mr. Biemer. The information on the decision screen is
misleading. Now if you----
Mr. Markey. Would a news director be misled?
Mr. Biemer. Yes, they would.
Mr. Markey. If they depended on the information that VNS
sent them?
Mr. Biemer. Yes, they would?
Mr. Markey. Would anyone who had that information that you
say should have available to them, in your opinion, that made
the same decisions?
Mr. Biemer. I don't know what other information they had
available.
Mr. Markey. In other words, if VNS had accurately
communicated all of these deficiencies on the decision screen
to the network, the news directors and news anchors that night.
Mr. Biemer. If all this information that has come out later
was on the screens, no they wouldn't.
Mr. Markey. They would not have made it.
Mr. Biemer. Right.
Mr. Markey. Do you understand what we are trying to do
here, sir?
Mr. Biemer. Yes.
Mr. Markey. We are trying to put the Queen of Spades in
front of someone. I believe the Queen of Spades should sit
right in front of VNS. I think there was obviously a terrible
set of mistakes that were made at VNS that were then sent on to
all of these other people who depended upon them. I know this
because we pay for polsters as well as politicians. If a
polster makes a mistake even though he's a professional, I'm
going to make all of these expenditures based upon that
erroneous polling. So once VNS makes the mistake, everyone
who's dependent upon it is going to be prone to looking pretty
silly because they believe the numbers are accurate.
So what you're saying right now is that you agree that
these network news people had every reason to be mistaken once
the erroneous information was given to them. And I think that
is very helpful to us as we move forward. May I also say, Mr.
Chairman--and I beg your indulgence on this--I do believe this
should be a legislative hearing and not an oversight hearing. I
really don't think that we should be conducting this as though,
you know, this kind of a criminalization of news-gathering
here. I don't think there's any crime that's been committed. I
don't think there's any news--I think this should be a
legislative hearing and this should avoid kind of confusing
this with the tobacco hearings or the Firestone tire hearings.
There are clearly problems here; but as far as I'm
concerned, it's pretty clear that VNS was at fault and that
everyone else detrimentally relied upon this erroneous
information. And obviously in retrospect they did not want to
call the country wrong twice in one evening. There was great
risk they all ran depending upon this information.
Chairman Tauzin. The Chair thanks the gentleman. The Chair
acknowledges the gentleman is entitled to his opinion and I
accept your opinion. Let me just remind the gentleman, however,
we are in a fact-finding mode, and that is what this committee
goes into. And that's where we are today. Let me quickly get
something before I move on, that is, that we have evidence that
there was a new unified VNS system that would combine voter
tabulation and projection and analysis functions under
development since 1993, but that due to budget limitations the
project has not moved beyond the creation of a partially
unified data base.
Mr. Biemer. I don't know.
Chairman Tauzin. You don't know. We will find out a little
later. Who owns VNS?
Mr. Biemer. Who owns VNS?
Chairman Tauzin. Yes.
Mr. Biemer. I thought the board of members.
Chairman Tauzin. Mostly the networks. Right?
Mr. Biemer. Right, the networks.
Chairman Tauzin. Mr. Bilirakis is recognized.
Mr. Bilirakis. Thank you, Mr. Chairman. I guess much of the
discussion today has sort of revolved itself around the word of
bias, and statistical biases may be a nicer adjective for it.
But I submit that as long as we are alive there's going to be
bias. And that bias is there with the networks, just like I
have a bias, just like you all have a bias. It is going to be
there. And I'm not sure that anything this committee can do is
going to climb into the heart and the head and the mind of an
individual or corporation or whatever to try to keep a bias
from entering into it.
And so I personally think that we have got to be concerned
with a process which offers the opportunity for the bias. And I
guess it goes maybe more to what Mr. Wattenberg and some
others, like Ms. Konner said that they don't use exit polls at
all. Did you say that at all during the election period?
Ms. Konner. To project.
Mr. Bilirakis. To project. But you think we should use exit
polls during the election period for other purposes?
Ms. Konner. After the election.
Mr. Bilirakis. After the election, not during the election?
Ms. Konner. After the election. They are a very useful tool
for analyzing the vote.
Mr. Bilirakis. I can understand that, but after the
election. Mr. Risser do you agree with that?
Mr. Risser. Yes, I do.
Mr. Bilirakis. Mr. Wattenberg.
Mr. Wattenberg. I certainly agree that it is a very
valuable tool for students of politics. I would just like to
make one note about Mr. Markey's search for the Queen of Spades
and this distinction between the news directors and VNS. If I
were a network news director, I would have a little sign on my
desk that said ``VNS or Us.'' In other words, VNS is, am, are
the networks. They own it lock, stock, and barrel. They have
been advised, as I understand it--I mean, there's a paper trail
going back 20 years as to how we should do this process: Is it
a good process? Is it biased? When should we call? When
shouldn't we call? I think the whole thing, listening to this
discourse, is somewhat bizarre.
Mr. Bilirakis. Mr. Wattenberg.
Mr. Wattenberg. I'm sorry.
Mr. Bilirakis. You're doing this on my time. And it's okay.
Mr. Markey. Could you open the mike for me.
Mr. Bilirakis. I don't know that I have the time.
Mr. Wattenberg, you refer to the preamble and you read the
preamble to us of the report. So I ask you, how tough should we
as a Congress be? We all believe very strongly in the first
amendment; and we all want to be careful that we don't trample
all over it and that sort of thing. But how tough should we be
considering this is a real problem. After all, it can affect
and has affected our elections.
Mr. Brown and others have referred to some of the things
that have happened in Florida, et cetera, et cetera, that's,
you know, that's one opinion. But we know what exit polls can
do. We know what calling these elections early, if you will, in
an incorrect manner or even a correct manner can do insofar as
voting around the country is concerned. If these early Eastern
elections had been called even correctly, if you will, I think
we all agree it would have affected what took place in the West
in California and what not.
Ms. Konner. I don't think we would all agree to that.
Mr. Bilirakis. You wouldn't agree with that. With what?
Ms. Konner. That no matter what happened in the East it
would affect the West. I don't think there's any evidence for
that.
Mr. Bilirakis. Is that--Mr. Wattenberg.
Mr. Wattenberg. The evidence is that there is no evidence.
That no one knows what sort of effect----
Mr. Bilirakis. All right.
Mr. Wattenberg. [continuing] what sort of effect early
reporting has. In my own view, I think perhaps Joan and I may
or may not disagree on this, but that sets up a mode of
suspicion and rancor some of which is reflected.
Mr. Bilirakis. Mrs. Konner, what is wrong, then, with exit
polling which may be correct, which may be incorrect, things
that have took place, what is wrong with it all if it isn't
going to make--disenfranchise, if you will, people in certain
election areas in certain States, the Western part of Florida,
for instance, the Northwestern part of Florida, California, et
cetera, what is wrong with it then if it's not going to----
Ms. Konner. It's up to a voter to vote. It's up to a voter
to vote. The polls are open, the voter may vote. It's just that
simple. We are not disenfranchising, we being----
Mr. Bilirakis. You don't feel that they might be
influenced?
Ms. Konner. That is their responsibility whether or not to
be influenced. The voter. Exactly as the respondent in a
polling operation is responsible for responding or not
responding. As it was said, that the procedure was judged to be
in concert with the best polling procedures. What they cannot
control is the response from the person that's being asked the
question.
Mr. Bilirakis. What are we doing here today then? What are
we concerned about. What are you concerned about when you say
there should be not be an exit poll.
Chairman Tauzin. The gentleman's time has expired, but you
may answer if you would like.
Ms. Konner. We are concerned about using polls for
projections.
Mr. Bilirakis. If they don't cause any harm what's the
trouble? What's the problem?
Ms. Konner. Faulty information causes harm. It distorts the
view of reality. Our job is to give a picture of reality that
is based on fact. That is the role of the journalist.
Chairman Tauzin. The gentleman's time has expired. The
Chair recognizes the gentleman from Ohio Mr. Brown.
Mr. Brown. Thank you, Mr. Chairman. I was intrigued by Mr.
Waxman's opening statement when he read an article from, I
think, the L.A. Times citing a reporter who was surprised and
perhaps more than surprised about the election night about the
placement of certain people, if you will, in the election night
operation of Fox, that the first cousin of the Governor of
Florida, Governor of Texas, the first cousin worked in a
crucial position on the election night desk.
And that that article, I believe, went on to say from Mr.
Waxman's statement that that was the--that was the first
network that called the election and at 2 or 3 in the morning
for Governor Bush, that network. Then in the very highly
charged competitive atmosphere as we saw in the clips, as we've
heard from several up here, as we've heard from others and that
the highly charged competitive atmosphere that the other
networks soon followed suit and called the election for
Governor Bush. The article went on to say the burden that sort
of put, or the imprimatur that that put on the election on the
American public's view of who won the election, that and
immediately some very clever people came up with the ``Sore
Loserman'' signs, it looked immediately a take-off on the Gore-
Lieberman signs as you recall. It immediately made it really
look like--it gave some impetus to Republicans consistently
saying Gore is stealing the election, Gore is stealing the
election, Gore is stealing the election.
I'll start with you, Mr. Wattenberg. Do you recommend that
the networks adopt a policy prohibiting the hiring of a
Presidential candidate's close relatives, second cousins first
cousins, brothers, sisters? Should that be a statement that
this panel makes?
Mr. Wattenberg. No, sir. I think that's preposterous.
Mr. Brown. And why?
Mr. Wattenberg. Because you hire people based on their
merit and their qualifications. And I don't think anybody ought
to be kept out of a job because he's related to someone or not.
And the idea that the other five distinguished and
professionally--professional journalistic operations base their
call on what Fox did, which is one of the minor players in this
game, strikes me as equally preposterous. And I would be much
surprised if any of these network executives acknowledged any
sort of--they all were calling from the same screen. Somebody
had to call first. It was within 2 minutes they all started
calling.
Mr. Brown. You believe it was all coincidence that Fox,
where Governor Bush's first cousin worked and helped make a
decision having talked several times that night with, I
believe, both Governors both of--two of his first cousins, that
they went first and the other networks when afterwards. That's
total coincidence.
Mr. Wattenberg. Fox called Gore while the election was
still in the first call, Fox called Gore along with all the
other networks, and that was while the election was being held.
This was post election. I see no particular evidence.
Mr. Brown. You say it was a coincidence.
Mr. Wattenberg. I should hope. So I would imagine so. I
can't imagine it being anything else.
Mr. Brown. Do any of the other three of you think just for
appearances sake, I know that journalists generally believe
that the appearance of a conflict of interest is a problem. And
when, you know, the fact is the networks, no--reporters didn't
report very extensively that George Bush's first cousin was in
a decisionmaking position, if you will, for what is
acknowledged by almost everybody up here as the most
conservative, politically, of the major networks. Most of the
public doesn't know that. When you tell people that George
Bush's first cousin works for Fox and they called the election
first, I'm not a conspiracist either but I also think that the
appearance of that could be troubling to some people. And
journalists are always very aware of appearances.
Mr. Wattenberg. Sir, there are a lot of Americans who think
that the other five networks are too liberal. So there's bias
and there's bias.
Mr. Brown. Only because conservative columnists, yourself
excluded, of course, have been telling the American public for
years that the networks and the media in this country are so
liberal.
Mr. Wattenberg. Based on some pretty good data.
Mr. Brown. We could debate that. Anybody else on the panel
think it would be a good idea that there's a policy you don't
hire in decisionmaking positions, on election night, close
relatives of Presidential candidates? Ms. Konner.
Ms. Konner. Policy decision?
Mr. Risser. I don't think there should be any policy set by
Congress or an official body.
Mr. Brown. No. I don't either.
Mr. Risser. I think networks, when they look at who they're
going to hire to operate the news side of the business, they
look at their past, what their qualifications are; but I
certainly wouldn't recommend any kind of ban. There are lots of
people working in the news media who have been in government or
politics on one side or the other over the years and still can
do fine jobs in journalism.
Mr. Brown. Ms. Konner.
Ms. Konner. I don't think there should be a policy. I don't
know the qualifications of that particular individual. If there
is a perceived conflict of interest, I would think that there
would be some layers of insulation between that individual and
the person who was going to make the announcement on the air.
Mr. Brown. Mr. Biemer, do you have an opinion?
Mr. Biemer. Well, all I want to really point out is that
the, you know, when the Bush call was made there was only like
a .6 percentage point difference between the candidates. And
VNS did not make the Bush call, the networks did. But looking
at the data on the screen, I would not have made that call. So
I don't know what happened in terms of calling it for Bush. The
data didn't support the call as far as I'm concerned.
Mr. Brown. So it may not be out of the question that the
calls to Austin and to Tallahassee to wherever, I don't know
where the Florida Governor was that night, that those calls
could have possibly had an impact or the American public could
think they had an impact on Fox making a decision that VNS
really didn't suggest that they make?
Mr. Biemer. I guess anything is possible. I don't
understand why the call was made is all I would to say.
Chairman Tauzin. The gentleman's time has expired. The
Chair needs to report however to the committee that we do have
some factual information gathered by the investigatory staff on
this point. That there were, in fact, four members of that
decision desk, three of whom were Democrats, that Fox had a
unanimous decision requirement all four had to agree before a
call was made. And that other members of the desk were, in
fact, talking with Democratic operatives throughout the
election night just as calls were being made to the Bush
operatives that evening.
That's the facts we've uncovered. Judge as you may. The
other thing we have, we do have a VNS screen from that time
period at 2:10 in the morning which we will discuss when VNS is
here.
Mr. Brown. Can I ask for Ms. Konner's response.
Ms. Konner. That somebody should check those facts? As we
should check all facts. There are three Democratic operatives
plus one Republican operative.
Chairman Tauzin. No, Ms. Konner, let me repeat what our
investigative team discovered, that there were four members of
the desk at that network. And that the policy of the network,
according to the network, that was all four had to agree upon a
call. Three of the members of that four-man team were Democrats
and that calls were made to Democratic operatives that evening
just as they were made to Republican operatives. That's all I
know.
Mr. Brown. That was done by an internal not external
investigatory scheme if you will. I would add if we're going to
play tit for tat.
Chairman Tauzin. The gentleman----
Mr. Brown. Who is the CEO of this Fox news is also a former
political consultant. If we want to play that game. The point
is----
Chairman Tauzin. The gentleman doesn't have the time. I
will yield to the gentleman if he wants to continue this. But
the Chair is simply laying down facts discovered by our
committee investigative team--by the committee's investigative
team. If you dispute those facts, you can do so in your
inquiries.
The Chair now recognizes the gentleman from Pennsylvania
Mr. Greenwood.
Mr. Greenwood. I'm calling that the hen in the Fox house
conspiracy theory. It was a laugh line. And I also note that on
the question of whether these TV programs influence voters,
they did an exit poll and 100 percent of the voters coming out
of the polling booth said they were not dissuaded from voting
by television programs. That was also a laugh line.
Anyway, absentee ballots. Mr. Biemer, you indicated that in
your report that failure to accurately account for bias due to
under coverage of absentee ballots was a major source of error
in the VPA estimate. Can you quantify that?
Mr. Biemer. Yeah.
Mr. Greenwood. How much of the problem was due to absentee
ballot undercoverage?
Mr. Biemer. I think we have a table on page 10 that shows
that the absentee ballots accounted for about 20 percent of the
bias.
Mr. Greenwood. Okay. And do you have information as to for
how long VNS and/or their network owners were aware, have been
aware of this issue of absentee ballot under coverage?
Mr. Biemer. I think they were aware of that. And I think
they tried to project an absentee correction in one of the
models that they used on their decision screens. What happened
though is the projection was off.
Mr. Greenwood. Isn't it the case, and I think this is a
serious issue, because of the use of absentee ballots is on the
increase. I believe in some States you can vote by absentee
ballot for weeks in advance of the election. I think they're
going to become increasingly a higher percentage of the total
vote tally.
And in Florida, I understand, that Bush received 23.7
percent more absentee ballots than Vice President Gore; is that
correct?
Mr. Biemer. I think so. 23.7.
Mr. Greenwood. Isn't it the case that in order to--in order
to dispense with this source of error, to do it effectively,
you'd have to do telephone surveys in advance of the election.
Isn't that the most effective way to correct this? In other
words do some kind of telephone survey to get a statistically
accurate count of how many people had voted by absentee ballot
and how they had voted?
Mr. Biemer. That's the best way I can think of.
Mr. Greenwood. Is there reason to believe that the reason
that the networks--despite the fact that the networks had known
and VNS had known this is a source of--significant major source
of error, and that they've known it for awhile and that they've
known the way to correct it is with the telephone surveys, is
it apparent that they didn't do that because of the cost
involved?
Mr. Biemer. That would be my guess, it would be the cost
because it is pretty expensive.
Mr. Greenwood. Do any of the other panelists have
information with regard to this issue and for how long it has
been the fact that the networks have known that the absentee
ballot undercoverage is a major issue and to what extent they
chose to resolve it; ignore it?
Ms. Konner. I don't know.
Mr. Greenwood. Let me return to Mr. Biemer then. Would your
recommendation to the networks be, if they continue to use this
kind of a system, that they pretty much have to if they want to
be accurate, that they're going to have to spend the money to
do the telephone surveys in advance of the election in order to
accurately count the absentee ballots they would have to do
that in every State.
Mr. Biemer. I'm not sure it would be necessary in every
State because I think it becomes a real issue in States that
are going to have a very close election. I think there might be
other ways which are less expensive to try to project what the
absentee vote would be well enough so that in States where
there's a wider margin it's not going to come into play.
In Florida because essentially it was a tie between the two
candidates, a small error in their projection of the absentee
vote made a difference, a pretty important difference in the
way that they called the election. In other States where
there's a wider margin of victory there, the methods that
they're using maybe with some improvements, looking at sort of
trends over a number of elections to see which way, you know,
if absentee votes are on the increase, how much are they
increasing and be able to sort of project what they are for the
next election, do that a little bit better than they're
currently doing. I think there might be some ways there which
you can use that methodology. But it would not work in all
States.
Mr. Greenwood. Mr. Chairman, I yield back the balance of my
time.
Chairman Tauzin. The Chair thanks the gentleman. The Chair
recognizes the gentleman, Mr. Stupak, for questions.
Mr. Stupak. Thank you, Mr. Chairman. Ms. Konner, the
original allegation made by this committee was that the
networks had delayed or deliberately otherwise certain State
calls for George W. Bush that made the rest of the country
believe that Mr. Gore was sweeping the country and discouraging
people on the West Coast from voting. Did you find any evidence
of that?
Ms. Konner. No.
Mr. Stupak. Mr. Wattenberg, you provided us a copy of your
column in the Washington Times in which you discuss much about
Gore's, if I can quote, winning ways on the East Coast, that
Slade Gordon lost the Senatorial election in Washington State.
Washington, of course, is a State that has 50 percent absentee
vote. Well let me ask you, was that a conclusion reached in
review of your report for CNN or is this your own personal
view?
Mr. Wattenberg. My column is my own personal views.
Mr. Stupak. Okay. Was there anything in this report that
you did for CNN that would lead you to conclude that?
Mr. Wattenberg. That to----
Mr. Stupak. That Slade Gordon lost because of the so-called
winning ways of Al Gore on the East Coast.
Mr. Wattenberg. The point that I was making, Congressman,
was that he lost by just a couple of thousand vote in a medium-
size State. And if, in terms, of all the scholarly evidence
that we have about the impact of voting behavior in States that
vote after other States are called is that we don't know. There
are scholars who say it one way and scholars who say it the
other way. And my point is that is just opening up a whole new
ground of rancor and----
Mr. Stupak. The CNN study you did, anything in there that
would lead you to make that review or that decision. Or it was
just these other scholarly reports that you've read in the
past.
Mr. Wattenberg. We looked at the scholarly report, and I
think we jointly came to the conclusion that they're pretty
well all over the lot. And my own conclusion was that they all
seemed to end up with the idea that well, even if it is a
factor of the ones that were positive in this sense, that even
if it is a factor, it was a very small factor. And the whole
lesson of Florida is that every vote counts. And consequently,
the early release of these votes is just asking to take the
Titanic to another iceberg.
Mr. Stupak. But early announcement like this on the East
Coast would have less impact in Washington than other States on
the West Coast because Washington has such a high absentee
voter after votes were in before the announcement could be
made. So the impact upon the State of Washington of any of the
West Coat States would be minimal if anything.
Mr. Pickering. Is 2,000 votes minimal?
Mr. Wattenberg. Yes. That's my point. It might be minimal.
But if the margin of victory was only a hundred votes or ten
votes or one vote, you're opening up a system that leads you
directly to these kinds of hearings with everybody else saying
you know, you're not telling the truth and the election was
stolen.
Mr. Stupak. Did anyone else on the panel other than Mr.
Wattenberg reach that conclusion that Slade Gordon lost because
of----
Mr. Wattenberg. I don't say that, sir. I indicated what the
margin of difference was, using it as an example to show how
close certain elections can be.
Mr. Stupak. In Michigan, Mike Rogers beat Diane Byrum by 88
votes, and that was called before. Do you think that would have
influenced the Mike Rogers win over Diane Byrum?
Mr. Wattenberg. No.
Mr. Stupak. But yet Washington it's different.
Mr. Wattenberg. It's two time zones later. It's 2 hours
later. Certainly.
Mr. Stupak. Well, part of Michigan is on central time zone
too.
Mr. Wattenberg. Then it's 1 hour later.
Mr. Stupak. Could be 2 hours later. Did anyone else reach
that same conclusion? Ms. Konner.
Ms. Konner. All of the--what we read, this is secondhand
evidence, what we read that none of the so-called scholarly
studies were beyond criticism. That all of them had serious
flaws in the methods they use to study it. And there was no
conclusion that could be drawn from the studies that early
reporting had any impact on voting in the other time zones.
Mr. Risser. If I can just add, one of the problems or
questions is even if it does have an impact, it's not always
clear which way the impact would be. If I hear my candidate
lost, okay, maybe I don't go to the polls then; but maybe so
does the person who wanted to vote for the winner because they
think well he's already won. So you don't really know which way
the effect would go if there is any.
Mr. Stupak. Ms. Konner, in your testimony, you indicated
also that the networks pooled their resources in VNS to cut
costs and expanded their polling research. It's my
understanding that they actually reduced it from 2000 precincts
that CBS previously did plus the 2000 that NBC did which may
have some overlapping to some extent to about 1400 precincts.
So there is actually less polling instead of more. Is that fair
to say?
Ms. Konner. My understanding is that the service was
expanded when this collaboration was set in motion.
Mr. Stupak. Okay. So it was expanded. But was there some
overlapping then which would actually produce less precincts
being looked at.
Chairman Tauzin. The gentleman's time has expired.
Ms. Konner. That may be so in some areas, but I think
overall there was an expansion of service.
Mr. Stupak. Thank you, Mr. Chairman.
Chairman Tauzin. I thank the gentleman. The Chair
recognizes Mr. Deal, I believe, is next in order.
Mr. Deal. Thank you, Mr. Chairman. Ms. Konner, in your
testimony you cite the report. And one of the statements you
made is to follow up on what was just requested or commented
on, that the concept of VNS also effectively eliminates
competition in the market for the establishment of a second
system and it might also stifle journalistic enterprise. The
idea being that more than one source is always good in order to
check a result. It would seem to me that in light of this
result which is embarrassing obviously to the networks, that
instead of continuing to pool efforts that maybe the race
should be on to see who has the best sources rather than each
relying on interpreting the same source. Is that, in effect,
part of what the report is recommending?
Ms. Konner. That's one possible solution. Even so, let us
say CNN had its own source, it would be a single source. It
might be better to have two sources which they share in, two
independent parallel sources of polling. That's just--I mean
that's just hypothetical. It's up to the networks to determine
now they're going to solve this problem. But it is a problem to
have a single source.
Mr. Deal. Well----
Ms. Konner. If you had five independent polling operations,
but each network was only relying on its own, it might not be
an improvement. I don't know the answer to that question.
Mr. Deal. The other question of course, and I think
everyone has commented on it to some degree, you say we also
found no convincing evidence that calls made before polls are
closed within a State or in another State have an impact on
voter turn out. I suppose the corollary of that is true as
well, there is no evidence that it did not. Is that correct?
Ms. Konner. There is no evidence.
Mr. Deal. That's sort of like trying to call the election
almost, isn't it. There is no evidence either way.
Mr. Wattenberg. That's exactly the point.
Mr. Deal. That's all, Mr. Chairman.
Chairman Tauzin. I thank the gentleman. The Chair
recognizes Mr. Doyle. I'm sorry, Mr. Deutsch will be first.
Mr. Deutsch. Thank you, Mr. Chairman. Mr. Biemer, you know,
we've talked a lot about so far today the statistical bias and
you know, the error system. And I want to present at least
something which we know happened, which is that blacks in
America, when they think they're voting for someone, the chance
of their voting being counted as an actual vote is lower than
non-blacks in America. Would you accept that as a factual
statement?
Mr. Biemer. I'm not sure I'm clear on it.
Mr. Deutsch. That when someone goes to the poll, that your
exit poll, as a black African American and they are responding
to you that they voted for a particular candidate that
statistically the chance of their vote not counting ultimately
in the actual count of the ballot as opposed to the poll is
less than a non-African American.
Mr. Biemer. I don't know.
Mr. Deutsch. It's something you should look at because it's
factually accurate. There have been a number of statistical
analyses that have been done, I think, for to you take a look.
And I'm going to question Research Triangle Institute in your
whole premise because if you're looking at the bias, which you
have said exists, for you not to have looked at that question,
I question the professional competence of the report. I mean
think about what you just said.
Okay, there's factual basis that black Americans' votes
actually don't count. What I am proposing to you is you're just
missing the whole point. And, to some extent, I think the
committee and the hearing is missing the whole point.
The chairman has said the error is in the system. The error
is in the systems that the exit polling doesn't accurately
reflect the results, that there's a bias. Why is there a bias?
Why? I would present to you one of the reasons there is a bias
is that we have a factual reality that African Americans and
others, you know, people in different segments in our society,
but specifically African Americans' votes don't count as much
because of a system in terms of how votes are counted.
Mr. Biemer. Right. Are you saying then that the exit poll
would count their vote but yet it wouldn't be registered in the
official?
Mr. Deutsch. Absolutely.
Mr. Biemer. So there is a bias there. I think we did
mention that in our report. There's I think a potential--well
it wasn't necessarily with regard to African Americans. But I
think 3 percent of the vote in Florida was not counted. And we
considered that as a potential source of bias on the other
side, that is what we hold to be the gold standard is the
actual official vote.
Mr. Deutsch. Right. But again the point that I'm making is
that the chairman has mentioned, my Republican colleagues have
mentioned 34 percent--34 States are biased toward Gore. Why was
there that bias? Why? I am presenting to you one of the reasons
there is that bias is because African Americans votes were not
counted as much.
Let me talk to you a little bit about why. If you're really
doing research. And this committee is really talking about the
election. There have been other studies that have pointed out
that to get through the--Florida had the distinction of having
the most candidates for President of the 50 States and the
District of Columbia.
It's not a happy distinction in hindsight, but it's--very
few, I'm sure, of my colleagues are aware of it. We have the
largest number of candidates for President on the ballot of any
State in the country. And there's also been an analysis that it
took a 4th grade education to get through the average ballot,
not just the famous butterfly ballot in Palm Beach County but
in Duval County and Jacksonville, 26,000 people in Duval
County, predominantly African Americans. And you can go by
precinct because Florida is a Voting Rights Act State where
race is actually on your voter registration card. It's not
theoretically, it's not just knowing the neighborhoods. We
actually know who voted. We know that black, which people
voted. That and in Duval County, there was a two-page ballot.
So there is research done that says that ballots required a
fourth-grade education. What happens when you need a--if you
only have a first-grade education and have you a ballot that
needs a fourth-grade education? That would create a statistical
bias. Because those people can't count.
Mr. Biemer. That's not a bias of VNS system; that's a bias
in the official----
Mr. Deutsch. What I am presenting to you and what I am
presenting to my colleagues is what we really should be talking
about. You acknowledge statistically that there is a bias. What
I am saying to you is that that bias is really at the core of
what's wrong with America. And that if we want to be talking
about this election, that's what we ought to be talking about.
We ought to be talking about the fact that literally
hundreds of thousands of people's votes didn't count, why
didn't they count, how we can change that so that never happens
again. So that we don't have hundreds in fact millions of
people in this society who are functionally illiterate. If you
don't have a literacy--there's more than one way to have a
literacy test which effectively is what we've had in this
country.
Again I would present to you that you know, we're looking--
my colleagues are just totally missing it. They're totally
missing it. The people on the other side of this room are
totally missing both VNS's faults and the faults of what
happened in the election. I see my time is up, Mr. Chairman.
Chairman Tauzin. Thank the gentleman. The Chair recognizes
the gentleman from Oklahoma, Mr. Largent.
Mr. Largent. I'm not even sure I know what I'm missing, but
apparently I'm missing something. My friend from Florida maybe
he can explain to me what we're missing.
But I would like to ask, we have a panel of experts up
here. I'm just wondering is there any factual data that
reflects what he just said, that more African American votes
are discounted or not counted than any other percentage of our
population? Is there any factual data to that? Anybody on the
panel.
Ms. Konner. I think you have to ask him.
Mr. Wattenberg. I have seen some data just in the press
that indicates that an African American casting a ballot, that
ballot is somewhat less likely to end up being counted than for
a non-African American. But--and I don't know that for a fact.
I've seen allegations of that. But there is a statistical point
there to be made which I find very interesting, is if you say
that African Americans voted 92 percent Democratic and they
were underrepresented in Florida, say, where they represent 15
percent of the vote, so then you have to ask yourself even if
there was a lesser discounting of the remaining 85 percent, the
non-African American votes who were disproportionately for
Bush, doesn't that--don't those errors cancel each other out or
perhaps even tilt the other way? I don't know the answer to
that.
Mr. Deutsch. If the gentleman would yield for 1 second. I
would hope that the record can be open for the hearing. I'll
provide again these are newspaper accounts and just some you
know some quick and dirty stuff I've done myself.
If you cross reference it, the percentage of undervotes or
overvotes in precinct, there is a direct correlation between
percentage of black, you know, race in that precinct, I mean
throughout the State of Florida. You can look at it yourself. I
encourage you to look at it yourself. I encourage the staff and
committee to look at it yourself, and maybe we'll have another
hearing related to. But again just hold the record and,
Congressman Largent, I look forward to giving it to you
tomorrow.
Chairman Tauzin. We always allow the record open, and the
gentleman may indeed submit the material which will be
considered for entry.
Mr. Largent. I'm always open to receiving down and dirty
information.
Mr. Biemer, do you do any accounting for the minority vote,
particularly the black vote? Do you do--in exit polling do you
do any kind of adjustments or recalculations or different
calculations to account for a minority disparity?
Mr. Biemer. I know that in--within Florida, there was a
separate stratum for blacks that was oversampled to provide
more accurate information.
Mr. Largent. So explain what you just said. There is an
oversampling of the black vote in the State of Florida. What
does that mean?
Mr. Biemer. When I say oversampling, this is a statistical
term to mean that they're sampled at a higher rate.
Mr. Largent. Why?
Mr. Biemer. This is a very common way in all survey
statistics, it's called the ``stratified sample,'' a common way
of increasing the precision of the overall estimate. Because
from sample to sample if you were to just sort of repeat
samples in Florida, just sort of keep drawing samples, you
would find that a stratified sample would be more consistent
across the number of samples that you drew than if it was
unstratified. And when you have a population that is going to
be voting somewhat very differently than the rest of the
population, stratifying them improves the precision of the
estimates.
Now what you do to compensate for that is you weight down
those--after you oversample and you compute an estimate for
that stratum, you weight it down in the overall estimate so
that in the overall estimates they're getting the right weight.
They're being represented in the same proportion as they are
represented in the population. They're not getting more weight
in the estimate. They're getting the same weight in the
estimate as their numbers dictate in the population. But it's
just a way of increasing precision.
Mr. Largent. So essentially what you're saying in laymen's
terms is because the black vote would be at 92 percent
Democratic that you would overweight that sample. It doesn't
have anything to do with whether they're votes counted more or
less.
Mr. Biemer. No. It has more to do that they're going to
vote differently.
Mr. Largent. Okay. I understand. Thank you. Dr. Konner, I
wanted to ask you a question. You stated a moment ago about
that what your job is to give a view of reality based upon the
facts. Is that correct?
Ms. Konner. That's correct.
Mr. Largent. Talking about--if exit polling in your
opinion, and I think I heard this correctly, you said that you
did not know or are not sure that exit polling used for
projections affected the outcome of--in polling places where
the polls are still open; is that correct?
Ms. Konner. I said there was no evidence to show that any
early projections impacted upon the outcome of an election.
Mr. Largent. Why not? Why would there not--why wouldn't
that--as a political scientist, why would that not be a really
critical issue that some political scientist, some smart person
would want to know?
This is an issue that's been going on for 21 years since
1980. Why doesn't a political scientist want to know, do early
predictions, based on exit polling, which is not facts--I mean,
if you're talking about using exit polling and you're talking
about dealing with the facts and views of reality, then don't
talk about exit polling because that's not reality. That's a
guess. That's a thumbnail.
Ms. Konner. You have to address that to the political
scientist community. Seems like----
Mr. Largent. I thought you were a part of the political
scientist. You're in journalism. I'm sorry. That's maybe a more
appropriate question for another person. My time expired. Thank
you, Mr. Chairman.
Chairman Tauzin. The gentleman's time has expired. The
Chair recognizes Mr. Doyle of Pennsylvania.
Mr. Doyle. Thank you, Mr. Chairman. I want to ask the
panel, now CNN retained you three distinguished individuals to
tell them what happened in the election and why it happened and
how to make sure it doesn't happen again. And two of the
recommendations you made was to stop using exit polls and votes
from sample precincts to predict the winner of the States; is
that correct?
Now, it's my understanding that CNN and for that matter all
the other networks are not taking either one of those
recommendations. Is that correct?
Mr. Risser. Well, only partly I would say. CNN's response
is that it will no longer use exit polls for projections in
close races. And if they can't tell at the time the polls
close, clearly who a winner is in a State based on exit polls,
then they'll drop the use of exit polls and go to the use of
the actual counted votes and sample precincts.
Mr. Doyle. Have the other networks indicated, have they
also indicated that they're not accepting either of those
recommendations?
Mr. Risser. I don't know.
Mr. Doyle. Why do you suppose that is? Do you envision
they're worried that coverage would start at 11 at night and
would have winners projected at 4 in the morning or----
Ms. Konner. Again I think this should be addressed to the
network representatives themselves. But I think that it should
not be overlooked that, in fact, the percentage of projections
that were correct is very high.
Mr. Doyle. Let me--yes.
Mr. Wattenberg. I don't think Mr. Risser indicated that CNN
did not accept our recommendation. I mean, I think as I
understand it, they from our recommendations and from other
material they've gathered, they came to a roughly similar
conclusion that we did, which was that exit polls should not be
used for projection purposes.
We all agree, and everybody I know in the political
community thinks that exit polling is an extremely important
part of the democratic process if you don't squeeze it into the
demand for results inside of 75 minutes so you can beat
somebody else by 3 minutes.
Mr. Doyle. Let me ask you about actual vote counts too.
You've recommended that actual vote counts be used to make
these projections. What is the rule of thumb or have you set a
standard, when do you--when do you make that judgment when you
think that you've seen enough actual votes to actually call an
election, have you established some rule of thumb for that?
Ms. Konner. We didn't do it in the report officially, but I
would say a majority of votes. When you have the votes coming
in and you have the majority of votes that are able to be cast
and that's the--you can call the election.
Mr. Doyle. But----
Ms. Konner. The other thing is that there is a second
source available in vote counting. The AP has its own, it has
its own vote counting operation. And that is the second source
that's available.
Mr. Doyle. But, for instance, in my State of Pennsylvania,
the votes from Philadelphia and Pittsburgh come in rather
quickly and that tends to skew toward Democratic vote in our
State and then the middle of State and the part they call the
``T'' usually comes in very, very late at night. We've seen
elections where we've gone to bed at midnight, 1 in the morning
thinking the Democrat had won the election. And you wake up the
next morning and the ``T'' had come in and the Republican had
won the election.
It would have been impossible under that scenario in not
only Pennsylvania but many, many States to actually call an
election without using some sort of a voter projections. And
I'm just curious. If we're going to wait until there is the
actual majority of votes counted, in most likelihoods we're not
going to be able to call Presidential elections on the evening
that they occur but we'll know when we wake up in the morning
in most cases. I mean----
Mr. Risser. Excuse me, Congressman, before the days of exit
polls, what you would see if you watched television was an
account of what percentage of precincts were counted in your
State, let's say, and what the margin was for each candidate.
And the people on air and their advisors would know enough
about whether votes came early from one part of the State or
another to whether they could call it. And eventually they
would call it. And maybe they occasionally made mistakes. But
it was all based on actual vote counts and a determination that
at a certain time there was enough known to make a decision.
When you make it solely from exit polls, especially if you
do it at closing time, you're not doing it based on any vote
not even on one vote really. You're doing it on what people
leaving the polls told the exit poller they were going to do.
So it's a difference in what you're using.
Mr. Wattenberg. Congressman, my view is, and I think the
view of my co-authors of this report is, that what you
described that the public wouldn't know until the next morning
who won the Presidential race, that is not a big problem. I
mean----
Mr. Doyle. There's not necessarily a bad thing, but that
would be the result of that. The networks might think that's a
problem.
Mr. Wattenberg. In a non-close election, you'd know right
away any way. I mean Reagan in 1980 or 1984, that would be a
no-brainer. The point that Mr. Risser has pointed it out that
it's a terrific news story if you say we don't know, if you
don't know. In some ways, a better story than saying we do
know, certainly saying we do know when you don't know. This was
a very, very, very close election. That's a dynamite story.
We're signing off now; we're not going to know until tomorrow
morning.
Chairman Tauzin. The gentleman's time has expired. There is
no question that----
Mr. Doyle. Thank you, Mr. Chairman.
Chairman Tauzin. [continuing] that whatever happened on
election night, an awful lot of people tuned in. It was pretty
interesting. The gentleman from Kentucky, Mr. Whitfield, is
recognized.
Mr. Whitfield. Mr. Chairman, first of all, I would like to
enter into the record the report from the 1984 hearing, joint
hearing in Congress, regarding this same subject matter.
And in this report it says, that nonetheless while the
committee is not prepared to state with certainty how large the
decline in voter turnout was, the available studies and surveys
indicate great likelihood that there was a significant decline
due to the early network projections.
And I notice that there were five studies conducted between
1981 and 1984 on this subject--one of them had a network
affiliation and the other four did not have a network
affiliation. Yet all of them indicated that the media did have
an effect on voting behavior by early reporting. And then on
the bottom part of this, there are 5 or 6 studies that said
there is no effect on voting behavior. And every one of those
except one, there was a network affiliation. I just want to
point that out.
Chairman Tauzin. Without objection the report is submitted
into the record.
[The report appears at pg. 182.]
Mr. Whitfield. Now, it's my understanding that this panel
is the only independent review of an election-night coverage of
a network. That your panel is the only one that's completely
independent; is that correct?
Ms. Konner. I believe so.
Mr. Whitfield. And you've recommended: One, that exit
polling not be used for calling an election; two, sample
precincts not be used for calling an election; and three, that
actual returns only be used and then enough votes to be counted
so that the outcome is certain so that there is no question.
Now this touches on what the gentleman from Pennsylvania
says. When we have local elections, there are no projections.
The actual vote count is reported, and that's it. The winner
wins. Now, if you make this recommendation to CNN which you
have made and maybe to other networks as well, how would the
American public suffer by not knowing the results of an
election until all the votes are counted?
Ms. Konner. I don't know how to answer that question. I
don't think that they would suffer by not having their appetite
for an immediate result known.
Mr. Wattenberg. I don't think----
Mr. Whitfield. So what is the rationale for this rush to
judgment that we must call these elections as soon as possible
without having complete data.
Ms. Konner. I think you have to go to the basis for the
competition between/among the networks.
Mr. Whitfield. You said that the core mission of journalism
is to serve the public.
Ms. Konner. That's what we believe. That's what journalists
believe. I do believe those are the core values that the
journalists at CNN have.
However, there is an underlying issue always. Journalism is
part of a commercial business. There has always been a tension
between commercial values, business values, and journalistic
values. They come into conflict frequently in determining what
you broadcast and when. However, there is a balance that is
struck that has served the public very well in the past. There
are indications today, and this is not in the report and I'm
going beyond the report, that that balance has shifted----
Mr. Whitfield. Right.
Ms. Konner. [continuing] and that corporate and commercial
and business values are holding much more sway over the
determination of the actions and behavior of journalists.
Mr. Whitfield. You know, I think that's a very good point.
We oftentimes don't refocus on that, but every network is a
company that's in the business to make money. And there is a
conflict sometimes with journalistic standards. I think that's
what's happening here. I yield back my time.
Chairman Tauzin. The Chair thanks the gentleman. The
gentleman from Louisiana, Mr. John, is recognized.
Mr. John. I'll be very brief. I just wanted to maybe follow
up what the gentleman from Pennsylvania was talking about. I
think it was Ms. Konner that was talking generally in answering
these questions about the actual votes or maybe it was Mr.
Risser that was talking about some of the networks who had
folks that were hired to look into the actual votes that were
coming in, and where they were coming in from. And, of course,
making that call once you get over that threshold would be the
safest way to call an election.
In your report, you recommend or talk about sampling and
some of the biases which exist out there. Is there a
methodology, to your knowledge, of where you combine both of
those type of sampling where you're looking at the actual but
you're meshing it together with some of the exit polling to
verify or not verify what's going on? Did that happen election
night, to your knowledge, in investigating this? And would that
be a good or bad thing to look at a model that might have both
of those methodologies combined?
Ms. Konner. I think that's the way it is done now. It's
just that they do not wait for--they don't wait for the actual
vote count to make the call. But those different figures are
being balanced. And they are all being taken into consideration
at the point at which they make the calls or the projection.
Mr. John. But I think it was obvious from some of the
predictions that the actual--depending on where they are coming
from and which State. I mean, I offer the State of Louisiana as
an example. If you call an election before the city of New
Orleans comes in, then you're making a big mistake. So I would
think that as you look at these and some of the States and some
of the States were called by just looking at some of this exit
polling and not--couldn't possibly be looking at actual votes
because of the timeframe of the calling and the closing times.
Mr. Wattenberg. I think what you're describing is a
function that is performed by the decision desks at the various
networks and there are some eminently qualified people there.
And I don't believe anything that we have recommended would set
up rigid rules certainly in non-close elections where you could
not use the accrued wisdom of experts to say, well, New Orleans
isn't in yet or the ``T'' in Philadelphia isn't in yet. I mean
that's part of the game.
I think what we came away with is that there is an arms
race going on, a time race, an arms race going on within the
networks that they feel pulled. On the one hand, they want to
be accurate. On the other hand, they don't want to be left
behind; or they want to be first. And my own--you'll have to
ask them, I don't want to engage in amateur psychiatrics, but
they are sort of in a mode of stop me before I kill again.
That's why I suggested in these various proposals of
uniform voting that they're in a mode where they can't stop
because of the competition. So if you have an outside force
that says, you're not going to get those real ballots until
later or we're going to have uniform poll closing, and that
takes the burden away from the appropriate defense of the first
amendment, that they can go about doing their job as admirable
professional journalists, which they are indeed.
Mr. Biemer. Can I comment?
Mr. John. Sure.
Mr. Biemer. I don't know if you are aware that the exit
poll is just one component in the process. They have the exit
poll, then they use the precinct, sample precinct information,
then they use the county level information.
Mr. John. So they overlay each other.
Mr. Biemer. Throughout the night, they bounce around. For
example, early on they may use the exit poll. If it's too close
to call for the exit poll, then they will wait for more data to
come in from the precincts; or maybe they'll even wait for the
county data to come in. So they postpone, they postpone these
calls if it's too close to call until the actual vote counts
are done at the end of the evening.
So they're doing it now it's just that I think what
you're--what we're talking about here is possibly not using the
exit poll component. I don't agree with that. I think the exit
poll component can be valuable in States where there's a wide
margin to call a race early if that's what you want to do. But
they do rely on other sources, not just that.
Mr. John. That really answers my question because there are
different components that we have to look at. If indeed what
happened on November 7 was--your initial was looking at exit
polls and then you moved down the ladder and if it's still too
close to call, you wait for those actual numbers. If that's, in
fact, what they did, then maybe I can understand. But I'm sure
that just by the time of some of the closing of the polls these
actual numbers in some of these places weren't able to be
verified or matched against some of the exit polls.
Mr. Biemer. They do call on partial information even when
they're waiting for data from the county, they may call a State
before all the county data are in.
Chairman Tauzin. The gentleman's time has expired. The
Chair thanks the gentleman. The Chair yields to the gentleman
from Mississippi, Mr. Pickering.
Mr. Pickering. Mr. Chairman, I want to thank you for
holding this hearing. I'm not going to ask any questions, just
make a quick statement because I know we need to get to the
next panel. And everyone has been very patient as we try to get
to the facts of this case.
To the gentleman beside me, the gentleman from Florida, who
raised some points a while ago, let me, in the spirit I hope
that will get us to a positive outcome of these hearings. The
reality of the truth is that the counties where you had the
most undervotes, overvotes, and the questions related to the
African American community were in Democratic counties with
Democratic election commissioners, represented by Democratic
Congressmen, in districts designed by Democrats. Now that's the
truth.
The reality is we need to all come together as Republicans
and Democrats and do something to modernize and reform the
election process. But it doesn't do any good to try to say one
side understands and one side doesn't understand and that there
was somehow a conspiracy on our side on those areas that we had
no control over.
The point of this hearing is to get to the truth, to get
the facts out, and to, hopefully, have constructive reforms for
the future both in the way the networks and the press cover
elections and the way we conduct elections. That's a separate
issue. But since it was raised, I just wanted to make that
point, Mr. Chairman.
Chairman Tauzin. I thank the gentleman. The Chair
recognizes the gentlelady from California, Ms. Harman, for a
round of questions.
Ms. Harman. Thank you, Mr. Chairman. I just would like to
say how happy I am to be a voice from the bleachers on this
committee and to agree with the comments of the last speaker.
I hale from what at least one member of this committee
calls the left coast of California where we are vulnerable to
fires, floods, earthquakes, and early election calls. And I
would just like to tell a little story which is that in the
1980's I had the good fortune to practice law with a man named
Jim Corman. James C. Corman was a senior Member of the House
until he lost his election in 1980. And many say now, certainly
or even then, that the reason he lost his election in 1980 was
that then-President Jimmy Carter conceded on the East Coast at
8 p.m. after East Coast election results were known. That was 5
p.m. in California and no Democrats showed up at the polls
after that. And Jim Corman lost an election that most had
thought would be a sure win. It was, of course, my good fortune
since I got to practice law with him.
But since my election to the House beginning in 1992, I
have never been elected by a margin of more than 5 percent. In
fact, I have only gotten to 50 percent once. And it was not
this past election; so I am extremely sensitive to the impact
of early election calls on the Left Coast of California.
The solutions, it seems to me, should include this notion
of a uniform poll closing; and, in that regard, I want to ask
my friend, Ben Wattenberg, why he thinks his idea of
conditioning the receipt of Federal funds on the agreement not
to disclose election counts would be easier to implement than
would a uniform poll closing time.
Mr. Wattenberg. I am not against a uniform poll closing
time. It seems to me that in a disparate country of 50 States
that jealously guard their rights, they are going to want to
set their own rules; and if you can work out a uniform poll
closing thing which, as I understand it, involves probably
disadvantaging the West Coast in terms of the number of hours
when people are wide awake that they will be able to vote, so
there are some built-in problems.
I am not against it. It just seems to me that this other
idea, if even a few States--I mean if you say Ohio and
Pennsylvania said, yes, we will take your money for voting
machines or we are going to hold our vote, none of the
political gurus are going to be able to make a call in a close
race nationally. They are not going to be able to come close.
Even if one of those States is out of the--if the race is close
in Ohio or Pennsylvania, the actual votes are not called, I
don't see how they can make a call before the Left Coast
closes. So it seems to me to be a much simpler matter.
I would just like to make one other point, which is, in the
course of our interviews, we had a long discussion with Tom
Johnson, who is the President of CNN who, in an act of rare
corporate courage, I must say, commissioned this panel to tell
it with the bark off, and he used to be the publisher, as you
know, of The Los Angeles Times, which has a lot of information
at its fingertips. And his view anecdotally, not
scientifically, was just as you say, is that a lot of people in
California believe they do not vote because the election has
already been decided. So that is a pretty high cotton source,
in my judgment.
Ms. Harman. Well, Mr. Chairman, time is short. I would just
like to conclude by saying that there are human casualties from
things like this. There may be no precise statistical proof of
that, but as one potential casualty, I have an enormous
interest in this and hope we get it right. I do think that
uniform poll closing is a piece of the answer. It may be that
this other idea could work, but I worry about counting on
appropriations for anything and feel that that may be more
uncertain.
Thank you, Mr. Chairman. Again, I am very happy to be part
of this committee.
Chairman Tauzin. Thank you, Jane. And we are very happy to
have you, by the way. Thanks for joining us.
The Chair is now pleased to recognize Mr. Buyer for a round
of questions.
Mr. Buyer. I have listened to some of my colleagues and
their comments. I just want to share with you, I spent 8 days
in Florida during the recount. So when you are there on the
ground and you have spent those days in three different
counties, you learn a lot. You learn what the rhetoric is, and
then you find out what the reality is.
I sat and I looked at a lot of those nameless and faceless
ballots of which allegations are being made here today. You
can't tell the race or the gender of that individual voter, but
I do know what happened. A lot of those thousands of ballots
that I saw is that for whatever reason that voter walked in
there, you cannot believe how many people voted for multiple
candidates for President, and you can't believe how many voters
actually walked in and voted in this election, not just in
Florida, but in all States, and didn't choose to vote in the
Presidential column. They went to vote for a sheriff's race or
the State election. They didn't vote in the Presidential
column. So we end up coming up with different theories as to
why.
In my State of Indiana, we had about the same number of
percentage of undervote as the State of Florida. So I get
really uneasy here when I hear these allegations of people
saying, well, something happened with regard to the suppression
of the black vote or that type of thing.
I just want to share the reality of actually seeing it on
the ground in Florida. Of course, you know, there has been no
discussion here about the Gore-Lieberman ticket having an overt
act to disenfranchise absentee military voters. That was the
most disgusting thing I have seen in my entire tenure in
politics, but Steve, let's don't talk about that here, that is
kind of a side issue, I suppose.
Chairman Tauzin. Please.
Mr. Buyer. Let me make several observations. Ms. Konner,
you were so strong and emphatic in your remarks, that there was
no evidence, with regard to any influence that a reporter might
report something that would have an impact upon a listener.
From my opinion, I think that is absurd. I do however agree
with your analysis that there is stress upon the journalist of
today. Their actions and behavior are a result of the
commercialism which exists today in the news industry.
Sitting there in Indiana, you know, news is almost more
about entertainment than it is news, and so I agree with that
comment. But when I look at this chart over there about
published studies indicating a media affect on voting behavior,
it is clear that evidence exists of media news reporting
affecting voters. There might be evidence that perhaps you
don't agree with, but I would say that your testimony that
there is ``no evidence'' may not necessarily be accurate. It
does appear that some information exists supporting your
position that there is no evidence of media affecting voters.
However, the evidence you speak of is all supported by
published studies that have a network affiliation. Now, that is
called bias. That is called bias. That bothers me.
Another thing that sort of bothers me at the moment is for
us to, well, there is no statistical evidence of a bias or we
have found no evidence of a bias. I think we ought to be
straightforward with each other. There are biases, and one of
the biases that I think is pretty obvious is that some or all
networks beat VNS to making calls for either Bush or Gore in 21
States.
So when Mr. Markey tried to say that we are going to lay
all the fault is VNS, I say wait a second. The networks, didn't
even wait for VNS on many occasions to make calls. Michigan and
Pennsylvania are two examples. Michigan and Pennsylvania both
were called for Gore. Michigan was called 83 minutes before
VNS, and Pennsylvania was called 37 minutes before VNS. But the
call for Ohio turns out differently it was delayed. The
networks called Ohio for Bush within 4 minutes of VNS call,
which happened to be almost 2 hours after the polls closed in
Ohio.
Look at Virginia. Virginia is another example of a delayed
call for Bush. Fox, NBC, MSNBC, calls Virginia for Bush at poll
closing; at poll closing Bush is ahead by 7.6 percent. The crit
is very safe, yet they delayed making the call for 30 minutes.
There was no aggressive action in this case. Networks were so
aggressive to make the call in Michigan and Pennsylvania for
Gore, yet in other States like Alabama, and others there was a
delay. So it does appear that there was bias by the networks.
Chairman Tauzin. The gentleman's time has expired.
The Chair recognizes the gentleman from Texas, Mr. Green.
Mr. Green. Thank you, Mr. Chairman. I won't--we need to get
to our other panel, but one of the things I wanted to point out
to my colleague from Indiana was that the undervote is not the
concern.
When you look at the numbers in Florida, and I would dare
say that that is true around the country, that of the 2.7
million votes cast in the eight largest counties, the overvotes
for Vice President Gore was 46,000 votes compared to President
Bush, with 17,000. And I would say that maybe we need to do a
lot better job on voter education and voter assistance in
helping people to know that if they have a spoiled ballot, they
can go get another one, which is I know Texas law and I would
assume it would be in Florida. And that is what I hope this
committee is about to say, okay, what can we do to make sure
people are exercising their right to vote and it is being
counted correctly.
I know I heard earlier that the, for example, the closing
or the announcement of the calling of Florida before the
Panhandle had finished voting, and we looked it up and I guess
it was 10 minutes before the polls closed in the Panhandle of
Florida, and I know my own experience in Texas standing out in
front of the poll for 10 minutes, I don't see people watching
TV or listening to the radio while they are waiting the 10
minutes for it to close. So the argument that people in west
Florida had decided they were not going to vote because it was
already called, I think we out to get down to the real issues
and not necessarily say somebody standing in line in Pensacola
walked away because it had already been called.
Again, hopefully that is what our committee will do, Mr.
Chairman, and see what we can do to make sure about voter
education, but also to make sure we have some type of
uniformity in calling these elections. Thank you.
Chairman Tauzin. I thank the gentleman. Thanks for also
expediting, because we do need to get to the next panel as
quickly as we can.
Mr. Pitts.
Mr. Pitts. Thank you, Mr. Chairman.
Mr. Biemer or any other panelist who may wish to respond,
if you look at VNS's best model estimates immediately before
and after poll closing, 20 minutes before the poll closing, the
best model underestimated Bush's margin of victory in 23 States
and underestimated Gore's margin of victory in only 5 States.
The best model overestimated Bush's margin of victory in only 5
States, but overestimated Gore's margin of victory in 12
States.
Now, the underestimate for Bush was 4\1/2\ times Gore, and
the overestimate for Gore was 2\1/2\ times more than Bush.
Ten minutes after poll closing, the best model
underestimated Bush's margin of victory in 24 States, and
underestimated Gore's margin of victory in only 4 States. That
is six times underestimated. The best model overestimated Bush
margin of victory in 4 States, but it overestimated Gore's
margin of victory in 14 States. That is three and a half times.
Now, that may not be intentional, but why did that happen?
Can you explain that disparity?
Mr. Biemer. I am not able to answer that. We only looked at
Florida, and we really didn't do the analysis that you're
talking about there, so I would have to actually see the data
and delve into it. On the surface of it, you know, it doesn't
sound like that type of thing would happen purely by chance,
but that is if there were equal chances for both candidates to
be under and over. It sounds like that is not happening, but I
am not sure what the cause would be or what the data are saying
in that instance.
Chairman Tauzin. The Chair will keep the record open and
members will be permitted to submit that kind of information
and get responses. You, perhaps, would like to do that so that
the gentleman can review your data.
Mr. Pitts. We will provide the data.
Mr. Wattenberg, do you have any comment?
Mr. Wattenberg. My general view is that we asked the VNS
people about that, and the answers they gave us, I think, were
convincing, which was, in effect, that it was--that that was
the way the cookie crumbled, that there was nothing devious
about it or preprogrammed in it.
What I found was bizarre was the magnitude of the miss--of
the miscall by VNS in so many States. I mean I just have a
couple here. Both Alabama and North Carolina, the swing
difference was 16 percent off. Now, that is--that is why I come
to the conclusion, and I think our panel came to the
conclusion, is that the last thing in the world you want to do
is start making--using this for projection purposes. When you
can let the dust settle and use it for analytical purposes and
hopefully have a second or third source so you can see who is
the outlier, that is of enormous value, but when they are
missing calls by 16 percent with no adequate way of correcting
for that, that is where the scandal is, in my judgment.
Mr. Pitts. Any other panelists want to respond?
We will provide this information to you. If you have any
insight as to why this was always tilting toward Gore and in
such magnitude, we would be interested in your opinion.
Thank you, Mr. Chairman.
Chairman Tauzin. Thank you, Mr. Pitts. I believe Mr. Walden
is next.
Mr. Walden. Thank you, Mr. Chairman.
I have a question, back to Mr. Wattenberg, the point about
being off by 16 percent. What should they do to fix the problem
that got them there?
Mr. Wattenberg. I am not a statistician, but I have been
working with polls for many years; and this is an intractable
problem they have. It is built into the structure of low-
response rates and getting lower, it was the 60 percent, now it
is at 55, now it is at 51, whatever it is, exacerbated by the
growing tendency of States to have early voting and more
absentee voting, and combined, that sets up the exit poll for
wide margins of error, and the only way to fix it, they say, is
to go to standard telephone polling. But standard telephone
polling, as we saw from the results during the course of the
election, have--they have enormous new problems themselves.
I mean the telemarketers are driving people away from
responding, phone answering machines, there is a whole panoply
of problems in regular polling where, as I understand it, of
every 100 people they call, they are only getting 20 responses.
So the logistical problem of picking up the absentee ballots,
plus the general accuracy of those polls, which were all over
the lot during the election; they came out all right, but at
the same time, in the course of the election, the polls varied
enormously, tells me that this early calling is not only a bad
idea in terms of democratic theory, but it is becoming
practically less and less plausible.
Mr. Walden. Should we then be using the exit polling to
call elections?
Mr. Wattenberg. I don't think--our panel thinks not, thinks
that they are a great tool for analysis, but not a tool that
should be used to project races.
Mr. Walden. Let me project into another area as well. I
noticed in your report which, by the way, was quite
comprehensive and I think quite to the point, shall I say, well
written. But a question I have, you did not take a stand on
whether there is a trade issue involved with VNS with all of
the networks going to one source and then agreeing to this
thing, and I respect that. Do you think there is an ethical
issue, and do you think there is an ethical issue with CNN and
CBS, from a journalistic standpoint, not from standards or
ethical things that we look at here in the Congress sometimes,
but just from a journalistic ethic, is that a good way to
operate?
Ms. Konner. I think the question has to be asked legally,
not ethically. I don't think there is an ethical problem here,
and I don't think there is an ethical problem with CNN and CBS
sharing a decision desk. The question about VNS as a
collaborative effort among competing agencies has been raised
by thoughtful people in the legal profession, and I think they
are the ones that have to address it.
Mr. Walden. But as a professor of journalism, is this
something you would recommend to emerging journalism students
from competing organizations to team up to achieve a common
information source and outcome?
Mr. Risser. I personally don't think there is anything
wrong with them teaming--of teaming up and having the source
that they use; the problem is if it is the one source that they
use and if they either don't have competing sources or don't
check it against other available sources. I think that is
really where the problem came up this time.
Mr. Walden. Let me ask one final question then. In Oregon,
we now do all of our voting by mail; and as you know, in the
report, VNS, they did a telepoll. That is the only way, and of
course it ended up being two votes per precinct difference in
the Presidential race. Anyway, it was very close and one of the
last ones called.
In these States, as you see a great increase in absentee
ballots or vote by mail, what recommendation would you give to
the networks regarding how they treat, how they acquire
information? Should they be out calling and reporting along the
way? Because you can determine who has already voted. You can
go to the clerks and determine who has already cast their
ballot.
Do you think it is appropriate then for the networks or any
news-gathering organization to determine who those people are
who have voted, figure out how they voted and begin to report
that?
Ms. Konner. I think any factually accurate information that
is available can be reported.
Mr. Wattenberg. I think it is fine in Oregon; I don't think
it is fine in Florida because of the time zone difference.
Mr. Walden. Oregon has----
Mr. Wattenberg. Excuse me?
Mr. Walden. Could you elaborate on what you mean by the
time zone?
Chairman Tauzin. The gentleman's time has expired but go
ahead and answer.
Mr. Wattenberg. Well, if it is an East Coast State and you
are reporting real votes during the Election Day or at poll
closing time when American voters are still voting 3 hours
later, you are intruding in a sacred secular moment of
democracy; and we ought to devise some methods that make it
difficult, keeping in mind the first amendment, from keeping
the networks from intruding in that way.
Chairman Tauzin. The gentleman's time has expired.
Mr. Terry is recognized.
Mr. Terry. Thank you. Being last means we get to move to
the next panel, so I can feel the anticipation in the audience
that I am asking questions now.
Let me just ask two quick questions. First of all, when we
talk about this sacred moment in democracy, that is what we are
talking about here, is how can we encourage or discuss the
accuracy of the information from network media which the press,
the print media doesn't like to hear it, but that is where most
of us get our information any more. So I am curious, it seems
to me that we are splitting hairs when we talk about using
inherently flawed exit polling data for analysis versus
calling. I know in your field they are two distinct things;
from a listener if I am hearing Brokaw, you, or O'Reilly,
whoever, or Bernard Shaw talking about in their analysis, my
goodness, Bush is in a tight race in Alabama or North Carolina,
this is devastating to their campaign, what is the difference?
Mr. Wattenberg. Well, I think there are two big
differences. One is, in theory, at least, most of the
analytical examination of exit polling would come after the
polls close, so it is no longer influencing the vote, and
second it would deal in sort of broad categories to say, well,
people seem to be voting on the economy, or that is the No. 1
issue or no, as some of the polls have showed, it is more on
moral issues than the economy, blacks seem to be voting one
way, Jews seem to be voting another, this is enormously
important material for voters and for journalists to have, but
it doesn't--it is not intrusive on the process of voting.
Mr. Terry. All right. I accept that point. But if they are
still using it as a cloaked way to call an election by saying,
you know, in essence that this is too close to call, it
shouldn't be this close, that type of editorial language using
that statistic, I still think that is wrong. But I see your
point. I think we need to define more succinctly the proper
uses for the polling, exit polling data.
Mr. Biemer, let me ask you. You seem to be the point person
on statistics, and believe me, as a lawyer and as a politician,
the last thing I am good at is statistics in this realm, but is
there a way, of all of the suggestions you have made, of making
it at least a little bit more viable? Is there a margin of
error still to it that can be assessed to it that at least the
public will know, gees, even when they talk about this, as good
as the model has become, you are still talking about something
that inherently is going to be 4, 5, 6, 7 points off.
Mr. Biemer. Well, it is important to realize that the exit
poll does a pretty good job in estimating the election day
voting, despite the fact that, you know, you have this tendency
to overpressure the Democrats in some areas, and there is a
problem with the estimator, the ratio estimator in particular.
The second problem--the ratio estimator problem can be
fixed. We can either show more estimators on the screen and
actually show how the vote changes, depending upon which past
vote you are using to buildup your projection, or you can build
in the error in choosing a past vote into the measure of
uncertainty there, which would also allow people to know what
the uncertainty is and they would slow down the calls.
So the problems, the more difficult problem is of course
the nonresponse rate is going to be--is one that we I think can
address with research, try to increase response rate and reduce
the bias associated with that nonresponse, but the absentee
vote is not part of that exit poll right now. It is a separate
adjustment that has to be made to the exit poll, and we
shouldn't be throwing out the exit poll because we are not
getting some other component right, which is the absentee vote.
There are ways of getting that component right, and some of
those ways are expensive. And in close races, you may have to
use a more expensive approach rather than trying to use past
races to project what the absentee vote is. But the exit poll
itself is not bad for the election day vote estimate.
Mr. Terry. It is not bad, okay.
Mr. Biemer. It can be improved. I am saying it is not
hopeless. We can--I think we can fix the problems with it.
Mr. Terry. The point that I was trying to make is that no
matter how you improve it, it is still going to be flawed to
some extent.
Mr. Biemer. Well, it is an estimator. It is----
Mr. Terry. I am going to cut you off because I am curious
on how the executives are going to adopt the policy of using
this exit poll.
Chairman Tauzin. The gentleman's time has expired.
Let me thank you all very much. One final thought. I
remember back in the 1980's when Mr. Markey and I were
considering all of these same issues, and I think he and I came
to the conclusion that if Americans really didn't like exit
polls, perhaps they ought to adopt a simple strategy and that
is to lie about how they voted when they came out of the polls.
Do you think maybe they took our advice?
Thank you very much. We appreciate it.
We are going to welcome the next panel of witnesses, so I
would like to see if we could bring them up right now as a
group. I will introduce them all to you as we prepare to
receive their testimony. While we are doing so, let me thank
you very much for this long evening. I don't know if it is
appropriate, but perhaps I should wish you all a happy
Valentine's Day today. I realize this may not be the best way
to spend Valentine's Day together.
Ladies and gentlemen, the second panel that we are
assembling is representative indeed of the networks and the
news agencies of our country. Let me first introduce Mr. Roger
Ailes, the Chairman and Chief Executive Officer of Fox News in
New York, New York; Mr. Andrew Heyward, President of CBS News,
New York, New York; Mr. Tom Johnson, the Chairman and Chief
Executive Officer of CNN of Atlanta, Georgia; Mr. Andrew Lack,
President of NBC News of New York, New York; Mr. David Westin,
the President of ABC News, New York; and Mr. Ted Savaglio,
Director of Voter News Service of New York, who is accompanied
by Dr. Murray Edelman, Editorial Director of VNS; and finally,
Mr. Louis Boccardi, President and Chief Executive Officer of
the Associated Press of New York, New York.
This panel again has been asked to come and to give us
their thoughts on what went wrong that night, their thoughts on
the internal and external reviews that have been commented upon
by the previous panel, and also to let Americans know what they
and their networks, their news offices may suggest as
improvements to the process by which we are informed about the
most important election in our country's ongoing history.
In anticipation of that panel, let me announce to the
committee I have asked our investigators to be very careful in
interviewing and working with these witnesses, because indeed
we do respect the first amendment, because it is our tool as
well as yours. It is critical to this Nation that we always
respect the line between government and the first amendment.
The first amendment was, in fact, designed to protect citizens
from their government in their free speech, and so we are
deeply concerned that we tread very carefully here.
So it is with deep appreciation, frankly, that I thank you
for the cooperation you have given our investigators in
learning as much as we could about what went wrong on election
night in November 2000, and second, how much we deeply
appreciate your willingness to come and share with the American
public your own thoughts about how we can improve the situation
of reporting on national elections for the future.
I am going on a bit until we get all the pictures taken so
we can commence our hearing. Again gentlemen, thank you and
welcome.
We will begin as we did the previous panel. We will start
with Mr. Ted Savaglio, director of Voter News Service of New
York, New York. Mr. Savaglio, again, according to our rules of
our investigative hearings, it is our practice to take
testimony under oath. Do you have any objection to testifying
under oath?
Mr. Savaglio. No, sir.
Chairman Tauzin. The chairman advises you that under the
Rules of the House and the rules of the committee, you are
entitled to be advised by counsel. Do you desire to be advised
by counsel for your testimony today?
Mr. Savaglio. Yes, sir.
Chairman Tauzin. Yes, you do. Well, then I have to go to
another page. In that case, would you identify your counsel for
the record?
Mr. Penchina. Robert Penchina.
Chairman Tauzin. Would you say it again, please.
Mr. Penchina. Robert Penchina of the law firm of Clifford
Chance Rogers & Wells.
Chairman Tauzin. Counsel, you may move forward and sit at
the table with your client if you like. Counsel, will you be
giving testimony today? Counsel, will you be giving personal
testimony today?
Mr. Penchina. I will not.
Chairman Tauzin. In that case, let me ask you, Mr.
Savaglio, if you will raise your right hand so I may swear you
in.
[Witness sworn.]
TESTIMONY OF TED C. SAVAGLIO, DIRECTOR, VOTER NEWS SERVICE
Mr. Savaglio. Thank you, Mr. Chairman and members of the
committee. Last fall, we witnessed the closest election that
anyone could have imagined. Its closeness brought to light a
series of flaws in the election procedures and in reporting on
the electoral process to the American people. Those flaws
included errors made by the Voter News Service, of which I am
the executive director.
As professionals who have tabulated, analyzed, and reported
on thousands of elections, my staff and I have spent this post-
election period working to understand precisely how those
errors occurred and how to prevent them in the future.
The electoral process is a cornerstone of our Nation.
Reporting on the culmination of that process is a serious
responsibility. We owe the public an explanation on the
mistakes that have been made. I can assure you we feel that
responsibility keenly.
At the outset, there is one matter I would like to lay to
rest. In reporting to our members and our subscribers and
indirectly to the American public, we have one paramount
concern: Reporting and analyzing the results of the election
accurately and quickly as possible. The notion that some kind
of political bias enters into our work is simply without
foundation and I am pleased, Mr. Speaker, that it appears to be
common ground among us here today. I also appreciate the
assurances that you have given that this process will not
offend the free speech principles that we both must defend.
The Voter News Service was created in 1993 and is owned by
ABC, AP, CBS, CNN, Fox and NBC. They are among the leading news
organizations in the country, and they are committed to the
highest standards of journalism.
The purpose of the Voter News Service is to collect,
tabulate, and disseminate vote returns, exit poll data, and
projections of election contests. That information is
distributed to our six member organizations and to other
subscribing news outlets who conduct their own analysis and
interpretation and report it to the American people as they see
fit.
In addition to providing information to analyze election
results, VNS projects the outcome of contests to its members
and subscribers. Our projections are based on complex
statistical analysis that take many factors into account,
including, among other things, the actual vote in sample
precincts, tabulated vote at the county level, and the exit
poll. All of this data is reviewed and interpreted by VNS
analysts who add their own knowledge and experience before
making a decision. Projections are made by people, not by
computers. Since 1990, when the first joint polling and
projection effort began, we have been involved in nearly 900
elections around the Nation.
The methods that we use to project winners in those races
have only been wrong once before. In other words, we have been
right 99.8 percent of the time. Unfortunately, when you make a
mistake as glaring as calling Al Gore the winner in Florida,
the number of times you have been right seems less relevant.
The plain fact is that despite our best efforts, the Voter News
Service let down its members and subscribers and ultimately the
American public. We are determined never to let that happen
again.
On election night, our statistical models, based on our
exit poll and actual vote from a number of sample precincts,
showed that Vice President Gore was ahead in Florida. Our
decision team considered other variables and determined that
the data clearly justified making a call. The reality, however,
is that the race was a virtual tie.
Based on all that we have learned since then, the error in
Florida was due to a convergence of a number of factors to
which all polling and projections are subject, which, in this
case, all pointed in the direction of a Gore victory. None of
these factors alone would have caused the error but, taken
together, they did. Later, after the Gore calls in Florida had
been made and retracted, we discovered problems in the
tabulation of the actual Florida votes that led to the race
being called for President Bush. In one case, Volusia County,
VNS passed on incorrect numbers that were released by election
officials, and this went undiscovered until after the Bush
calls had been made. Moreover, we significantly
undererestimated the number of votes outstanding.
Based on this experience and following the recommendations
in the Research Triangle Institute study and in other reports,
we are actively pursuing a number of improvements, including
the following: using larger samples for the exit polls;
developing new procedures to account for the effects of the
growing absentee and early vote; rewriting the VNS projection
and statistical models; working to improve exit poll accuracy
and response; completing work on the integration of the
Associated Press's tabulated vote as a second source of
information; and developing more sophisticated quality controls
in the tabulated vote system and in the rest of our systems;
and finally, upgrading and modernizing the VNS technical
capabilities and infrastructure.
We are taking these steps because, as journalists, we are
deeply committed to the integrity and accuracy of our
reporting. We are determined to do everything humanly possible
to make sure that these mistakes will never be made again.
Thank you.
[The prepared statement of Ted C. Savaglio follows:]
Prepared Statement of Ted C. Savaglio, Executive Director, Voter News
Service
Thank you, Chairman Tauzin, Congressman Dingell, and Members of the
Committee.
Last fall, we witnessed a Presidential election that was closer
than anyone could have imagined. Its closeness brought to light a
series of flaws in election procedures and in reporting on the
electoral process to the American people. Those flaws included errors
made by the Voter News Service, of which I am Executive Director.
As dedicated professionals who have tabulated, analyzed and
reported on thousands of elections, my staff and I have spent this
post-election period working to understand precisely how those errors
occurred and how to prevent them in the future.
The electoral process is a cornerstone of our nation. Reporting on
the culmination of that process is a serious responsibility, and we owe
the public an explanation of the mistakes we made last November. As
individuals who have devoted much of our lives to educating the
American people about elections, I assure you that we feel that sense
of accountability very keenly.
At the outset, there is one matter that I would like to lay to
rest. In reporting to our member news organizations and, indirectly, to
the American public on Election Night, we have one paramount concern:
reporting and analyzing the results of the election as accurately and
quickly as possible. The notion that some kind of political bias
enters into our work is, quite simply, without any foundation, and I am
pleased, Mr. Chairman, that that appears to be common ground among
those of us here today. I also appreciate the assurances you have given
that this process will not offend the free speech principles that we
both must clearly defend.
The Voter News Service (VNS) was created in 1993 and is owned by
ABC News, The Associated Press, CBS News, CNN, Fox News, and NBC News.
They are among the leading news organizations in the world and are
committed to the highest standards of journalism.
The purpose of the Voter News Service is to collect, tabulate, and
disseminate vote returns, exit poll data, and projections of
presidential primaries and national and statewide election contests.
That information is distributed to our six member organizations and to
other subscribing news outlets. These news organizations take the data
provided by VNS, conduct their own analysis and interpretation, and
report it to the American people as they see fit.
Our organization makes possible the timely reporting and in-depth
interpretation and analysis of election results that the American
people have come to expect and rely upon. On Election Day 2000, our
work involved more than 40,000 people who staffed nearly 28,000
individual precincts and went to some 4,600 counties to obtain the
information that we needed.
In national elections, the final VNS National Vote totals, which
are verified with official state canvases, become a record of the
election that is widely published. The exit polls that we conduct are
widely recognized as a critical source of information for understanding
an election. They are used by students, scholars, officials, and
journalists throughout the world. They tell the public, among other
things: who voted and why, what issues mattered most to the voters,
which candidates' policy positions were most effective, and which
candidates' qualities attracted voters most.
In addition to providing information to analyze election results,
on Election Night VNS projects the outcome of contests to its members
and subscribers. VNS' projections are based on complex statistical
analyses that take many factors into account including, among other
things: the actual vote in sample precincts, tabulated vote at the
county level, and the exit poll. All of this data is reviewed and
interpreted by VNS analysts who add their own knowledge and experience
before making a decision that it is possible to project the outcome in
a given race. Projections are made by people--not by computers.
Since 1990, when the first joint polling and projection effort
began, we have been involved in nearly 900 election contests across the
nation. The methods that we use to project winners in those races have
only been wrong once before. In other words, we have been right 99.8
percent of the time.
Despite our strong record of accuracy, we constantly strive to
eliminate all errors. In between elections, our staff evaluates the
performance of our methods and models and considers how they might be
improved. The models are based on the accepted statistical theory of
sampling, the principles of which have not changed. Nevertheless, over
the years, we have made improvements in the models and procedures. For
example, we routinely research our sample precincts prior to an
election, in order to take into account changes in precinct boundaries
and demographic composition. This year, our decision screens for
California and Washington were revised in light of a significant
increase in absentee voting which has been evident in those states. In
this regard, since 1996, we have conducted telephone polls of absentee
voters in states where a high percentage of absentee ballots are cast,
and we continue to work to improve the methods for polling absentee
voters.
Unfortunately, when you make an error as glaring as calling Al Gore
the winner in Florida at 7:52 p.m. on November 7th, the number of times
that you have been right seems less relevant. The plain fact is that,
despite our best efforts, the Voter News Service let down its members,
subscribers, and ultimately the American people, on Election Night
2000. We are determined never to let that happen again.
Toward this end, VNS has conducted an intensive internal
investigation of what went wrong on Election Night--an inquiry that is
still continuing. In addition, our members commissioned an independent
review by the prestigious Research Triangle Institute. Several of our
members have also conducted investigations of their own.
On Election Night our statistical models, based on our exit polls
and actual vote from a number of sample precincts, showed Vice
President Gore ahead--decisively it seemed--in Florida. Our decision
team considered other variables, including absentee vote beyond that
which already was accounted for in the models, and determined that the
data clearly justified making a call, which we did shortly before 8:00
pm. The reality, however, is that the race was a virtual tie.
Based on all that we have learned since then, the error in Florida
was due to the convergence of a number of the anomalies to which all
polling and projections are subject, which in this case all pointed in
the direction of a Gore victory. None of these factors alone would have
caused the error, but, taken together, they did. These factors include:
the exit poll, showing Gore ahead;
the fact that the actual vote from the first sample precincts
reporting indicated that the exit poll was actually
understating the Gore vote;
the fact that the model selected the 1998 gubernatorial
election in Florida, rather than the 1996 Presidential race, as
a basis for statistical comparison, when comparisons based on
the latter would have prevented us from making the call; and
a larger than expected absentee vote.
Later, after the Gore call in Florida had been retracted, we
discovered problems in the tabulation of actual Florida votes that led
to the race being called in favor of President Bush. In one case,
Volusia County, VNS passed on incorrect numbers that were released by
election officials, and this went undiscovered until after the Bush
calls had been made. Moreover, we significantly underestimated the
number of votes outstanding.
Based on this experience, and following the recommendations in the
RTI and other reports, we are actively pursuing numerous improvements,
including the following:
using larger sample sizes for exit polls;
developing new procedures to account for the effects of a
growing absentee and early vote, including more extensive
telephone polling of absentee voters;
rewriting of the VNS projection and statistical models;
working to improve the exit poll accuracy and response rate;
completing work on the integration of The Associated Press's
tabulated vote as a second source of information;
developing more sophisticated quality control in the tabulated
vote system and the rest of the VNS systems; and
upgrading and modernizing the VNS technical capabilities and
infrastructure.
We are taking these steps because, as journalists, we are deeply
committed to the integrity and accuracy of our reporting. We are
determined to do everything humanly possible to make sure that these
mistakes will never be made again.
Our intention is to take all that we have learned and use it to
improve our Election Night procedures--and thereby return to the
American people the confidence that they will receive timely and
accurate Election Night information.
Thank you, Mr. Chairman and Members of the Committee.
Chairman Tauzin. Thank you very much, Mr. Savaglio.
Next we will welcome Mr. Louis Boccardi, the president and
chief executive officer of the Associated Press. Mr. Boccardi,
are you also aware that the committee is holding an
investigative hearing and when doing so we have the practice of
taking testimony under oath?
Mr. Boccardi. Yes.
Chairman Tauzin. Do you have any objection to testifying
under oath?
Mr. Boccardi. If it is necessary to be sworn to speak, I
have no objection; but I don't think it is necessary.
Chairman Tauzin. The Chair advises you under the Rules of
the House, you are entitled to advice by counsel. Do you desire
to be advised by counsel during your testimony today?
Mr. Boccardi. My counsel is here. I do not anticipate that
he will testify.
Chairman Tauzin. Okay. In that case, would you please raise
your right hand, and I will swear you in.
[Witness sworn.]
Chairman Tauzin. Sir, you are sworn in, and we may receive
your testimony.
TESTIMONY OF LOUIS D. BOCCARDI, PRESIDENT AND CHIEF EXECUTIVE
OFFICER, ASSOCIATED PRESS
Mr. Boccardi. Mr. Chairman and members of the committee,
good afternoon. Previous witnesses and those who are going to
follow me have spoken and will speak in some detail about
November 7 and 8. Before I say something about that, I wanted
to take just a couple of minutes very briefly to make a
different point. We all accept the seriousness of what
happened; that is beyond question. But I first want to place on
the record a deep concern about the nature and scope of the
committee's inquiry into decisions made by journalists in the
course of gathering and reporting the news.
The chairman has said in correspondence with executives of
Voter News Service and the networks that there are potential
first amendment issues raised by what you are doing. We agree
with that assessment; there certainly are. AP has serious
doubts that the committee and its staff, no matter how
sensitive they may be, can avoid crossing the line between
appropriate government concern with the electoral process
itself and, on the other hand, inappropriate government
involvement with the reporting on that process by a free press.
To put it more plainly, we believe that such an official
government inquiry into essentially editorial matters,
summoning the people who sit here, is inconsistent with the
first amendment values that are fundamental to our society. I
say that with conviction, but without disrespect to the
important role, important, but I think a critically different
role, than that of the media that is played by the various
branches of government.
I respect you. As a citizen, I benefit from you. But your
job is different from mine, and a hearing such as this confuses
the two.
We agree that there were serious shortcomings, call them
terrible mistakes, I do, in the election reporting from Florida
on November 7 and 8. These mistakes cannot be allowed to happen
again. But fixing them is a job for the Nation's editors, not
for its legislators. What we report and when we report it are
matters between us and the audience we try to serve; they are
not matters between us and our Congressmen.
The written statement I have submitted to you reflects what
we believe to be the limits of an appropriate public account to
a government body of how AP did its work last November 7 and
through the morning of November 8. It is an account we have
given in stories and speeches and interviews. I will not take
your time to repeat it in the few minutes you allow my this
afternoon.
The AP is a mutual cooperative that collects and
distributes reports to its member news organizations. Those
members in turn disseminate that news to their readers, their
viewers and listeners. Like newspapers, AP is free of
government licensing. Our member editors and publishers and
broadcasters hold us strictly accountable for honoring a
bedrock of impartiality while vigorously defending the rights
of the media to collect and report the news free of outside
interference both in the United States and overseas. And
overseas, some of our people have paid the ultimate price for
this commitment: their lives.
We have covered every Presidential campaign since 1848, the
year we were founded. AP editors staffed their newly opened
office in New York to report that Zachary Taylor had defeated
Lewis Cass and Martin Van Buren and to become the Nation's 12th
president. We meet with you this afternoon to talk about the
election of our 43rd.
Reporting the names of election winners promptly has always
required substantial effort on our part, because as the
committee members know, the official vote canvas can take days
or even weeks to complete and to announce. To produce
unofficial, but accurate results so the public can promptly
know who won, AP collects returns at the local level, tabulates
them with the greatest care, and reports them. In this way, we
are able to provide timely results not only of national and
statewide election contests, but also State legislative races.
We collect results on 6,000 elections in a quadrennial or
biennial year. That number includes the 500 or so for which
Voter News Service also has done tabulations.In terms of races
covered, we are the largest and, forgive me, we think the most
reliable collector of returns in the country.
Our standards for deciding when to call a projected--to
declare a projected election winner have not changed
substantially. They are not secret. We have recited them
publicly before. Statewide returns from VNS and from AP's own
vote collection network are monitored in each State bureau by
individuals well versed in State political demographics and the
dynamics of individual contests. In the case of Federal
elections, analysts in Washington become engaged. We have given
a public explanation several times of our work last November. I
have restated it in what I gave the committee before today.
We made one erroneous projection on election night, the
early call of Florida for Gore. It was based on flawed data and
analysis from VNS, but we take full responsibility for what we
did. The committee also knows from its review of several
publicly released studies by VNS and its members and from its
conversations with the managers of VNS, the search for the
origins of the erroneous early Florida projection is focused on
certain statistical assumptions about the make-up and behavior
of the voters that turned out to be incorrect. No point in my
reciting them again here.
In regard to the late Bush call, as the committee is aware,
AP did not join in the early morning victory declaration that
the networks made. It was our independent editorial judgment,
based on our own vote counting and what we saw from VNS and the
input of our analysts, that the race was too close to call, as
indeed it turned out to be. It would be right to surmise that
the pressure on AP at that moment was enormous.
In conclusion, Mr. Chairman and members of the committee,
AP agrees with much that has been said here all day and will be
said after I finish by the network news chiefs. We agree, no
projection should be made until all polls in the State are
closed. We agree the Florida mistake seriously damaged the news
media in the eyes of the public we serve. We agree that VNS
must be intensively reviewed to eliminate technical and any
other weaknesses, administrative or anything else. What is
broken must be fixed. What is broken, I might add, includes
many aspects of the election system outside the purview and
capacities of the media. But we feel deeply the distinction
that must be maintained between the editorial process and
legislative inquiry, and I worry that a proceeding such as this
blurs that distinction. Thank you very much.
[The prepared statement of Louis D. Boccardi follows:]
Prepared Statement of Louis D. Boccardi, President and Chief Executive
Officer, The Associated Press
The Associated Press first wants to place on the record its deep
concern about the nature and scope of the Committee's inquiry into
decisions made by journalists in the course of gathering and reporting
the news.
Chairman Tauzin has stated in correspondence with executives of
Voter News Service and the networks that there are ``potential First
Amendment issues raised by the nature of this inquiry.'' We agree with
the Chairman's assessment. There certainly are.
AP has serious doubts that the Committee and its staff, no matter
how sensitive they may be, can avoid crossing the line between
appropriate government concern with the electoral process itself and,
on the other hand, inappropriate government involvement with the
reporting on that process by a free press.
To put it more plainly, we believe that such an official government
inquiry into essentially editorial matters is inconsistent with the
First Amendment values that are fundamental to our society. That is
said with conviction, but without disrespect to the important role--
important but critically different from that of the media--played by
both legislative and executive branches of government.
I respect you. As a citizen, I benefit from what you do. But your
job is different from mine, and a hearing such as this one confuses the
two.
We agree that there were serious shortcomings--call them terrible
mistakes--in the election reporting of November 7 and 8 and that these
mistakes cannot be allowed to happen again.
But fixing them is a job for the nation's editors and news
directors, not its legislators.
What we report and when we report it are matters between us and the
audience we try to serve, not matters between us and our Congressman.
The statement that follows reflects what we believe to be the
limits of an appropriate public account to a government body of how AP
did its work last November 7 and through the morning of November 8. It
is an account we have given in stories, speeches and interviews.
The Associated Press is a mutual news cooperative that collects and
distributes reports to its member news organizations to be incorporated
into the news products disseminated by AP's members to their readers,
viewers and listeners. Like newspapers, AP is free of government
licensing.
AP traces its origins to 1848 when a group of New York newspaper
publishers agreed to share the cost of collecting overseas dispatches
as they arrived at Halifax, Nova Scotia, at that time the first
landfall for transatlantic shipping. The plan worked well and almost
immediately developed into a news service in which reports on major
events were delivered to all members.
Newspapers of the 19th Century were often fiercely partisan in
their editorial policies. The only way for their cooperative news
service to satisfy all of them was to provide reports that were
strictly factual and impartial. That was the core of AP's mission at
the outset, and remains so to this day.
Article I of AP's by laws reads in substantial part as follows:
``The union for a common aim and purpose of representatives
of all shades of thought and opinion--political, social,
economic, religious--is assurance the news gathered and
distributed by The Associated Press shall be as objective and
complete as human endeavor can make it.''
AP member editors and publishers continue to hold their cooperative
strictly accountable for honoring that bedrock policy, and for
vigorously defending the rights of the media to collect and report the
news free of outside interference, both in the United States and
overseas. Overseas, some have paid the ultimate price for this
commitment--their lives.
AP has covered every presidential campaign since 1848, the year of
its founding. AP editors staffed their newly opened office in New York
around the clock for the first time to report that Zachary Taylor had
defeated Lewis Cass and Martin Van Buren to become the nation's 12th
president. We concern ourselves today with the election of the nation's
43rd.
Reporting the names of election winners promptly has always
required substantial effort on AP's part, because as the Committee
members know, the official vote canvass takes days, sometimes weeks, to
complete and announce.
To produce unofficial but accurate results so that the public can
promptly know who won, AP collects returns at the local level,
tabulates them with the greatest care, and reports the totals. To do
this, we hire and train special data reporters and post them at county
election offices where officials assemble the totals reported from each
precinct.
As county totals are updated throughout the night with additional
precinct reports, the AP data reporters--about 5,000 of them
altogether--are instructed to relay those totals to AP, where they are
added to the growing collection of results from across the state and
nation.
In this way, AP is able to provide timely results not only of
national and statewide election contests but also of state legislative
races and a limited number of high-interest local elections.
Altogether, AP collects totals for about 6,000 elections in a biennial
or quadrennial year. That number includes the approximately 500
elections for which the Voter News Service, and before VNS the News
Election Service (``NES''), have also tabulated results. In terms of
races covered, AP is the largest and, we believe, most reliable
collector of returns in the country.
NES was a consortium formed by AP, UPI, NBC, CBS and ABC in 1964 to
share the cost of tabulating national and statewide votes. NES created
a collection network much like AP's, and AP's separate network served
as a backup to NES, as it now does to VNS, in addition to producing
results for the more than 5,000 races not covered by NES and VNS.
Neither NES nor AP conducted exit polls. AP does not do them now.
The broadcast networks have used them for many years, however, and in
1990 they combined their exit polling operations into an organization
known as Voter Research Service (``VRS''). In 1993, NES and VRS merged
into the present Voter News Service, of which AP is a one-sixth
partner.
VNS conducts exit polls at selected precincts on election day,
tabulates actual votes after the polls close, and through computerized
statistical analysis of both sets of data produces running forecasts
throughout election night of the final results in each race.
It is the quality and impact of those forecasts, of course, which
have become a principal focus since the night of November 7.
AP's procedures for deciding when to declare a projected election
winner have not changed substantially in decades. They are not secret.
We have recited them publicly before. Statewide returns from VNS and
from AP's own vote collection network are monitored in each state
bureau by individuals well versed in state political demographics and
in the dynamics of individual contests. In the case of federal
elections, analysts in Washington become engaged.
As already noted, AP has never conducted exit polling on its own,
and exit poll results only became available to the AP staff as a
resource with the formation of VNS.
Exit polls have proven useful to AP reporters and editors in that
they may provide advance notice either that the actual results appear
to be consistent with expectations from pre-election polls and our own
evaluation, or that a surprise may be in store. We consider exit polls
a highly valuable part of our understanding, and our audience's
understanding, of what the voters are saying.
Valuable, though as we know, not infallible.
We have given a public explanation several times of our work last
November. I restate that widely available account here.
AP reported at 7:53 p.m. on November 7 that we had concluded from
exit poll and early reports of actual returns from some Florida
counties that Vice President Gore would be the winner in Florida. It
was the only erroneous projection AP made that night.
From their familiarity with the campaign in Florida and with pre-
election poll results made public by the candidates and others, AP
editors expected a very close election. They were therefore surprised
and skeptical shortly after 7 p.m. Eastern time when the VNS exit
polling data were indicating that Mr. Gore could win by a margin of
more than 6 percent.
Ordinarily, that wide a projected lead would suggest a sure
outcome, but because it ran counter to expectations no projection was
made. As the exit poll data were supplemented in the VNS system with
the first actual returns from sample precincts, however, the Gore
margin held up. That is why AP made the decision to report its
conclusion that Florida would end the night in the Gore column.
There has been widespread discussion of the wisdom of projecting a
winner at a time when 5 percent of Florida's polling places would still
be open for several more minutes. AP's policy has been to comply with
the 1985 agreement between Congress and the networks that projections
would be withheld for any state until a substantial majority of polling
places for that state closed.
In future elections AP will forecast no final results for any state
until all its polling places in all time zones are closed. I believe
the networks have taken a similar stance.
It would be wrong not to add at this point, however, that if voters
were actually discouraged by media projections from casting ballots on
November 7--and we have seen no credible evidence to show that many
were--their number is eclipsed by the tens of thousands of voters in
Florida and the millions nationwide who were disenfranchised by voting
machine breakdowns, confusing ballots, lost votes, and a host of other
consequences of official error, disorganization and incompetence in
administration of the elections.
In fact, problems in Florida's official vote counting apparatus
were a part of the media's troubles on election night. AP's projection
of a Gore victory in Florida was withdrawn a little over two hours
later at 10 p.m. It might have been retracted sooner, but for a
keypunch error by officials in Florida's Duval County that inflated the
Gore Florida total by 40,000 votes. Until that error was found and
fixed, the Gore victory projections continued to flow from the VNS
computer.
The Committee already knows from its review of several publicly
released studies by VNS and its members and from its conversations with
managers of VNS that the search for the origins of the erroneous early
Florida projection is focused on certain statistical assumptions about
the makeup and behavior of Florida voters that turned out to be
incorrect. AP's knowledge of the details comes from the same studies to
which the Committee has access, so it would serve no purpose to repeat
them here.
The owners of VNS are determining what must be done to eliminate
the technical and organizational weaknesses that made the early
misdirection in Florida possible.
As the Committee is aware, AP did not join in the early morning
projection of a Bush victory.
As AP has publicly reconstructed events with the help of the
reports the Committee has seen, the VNS computer system was indicating
at about 2 a.m. November 8 that only 180,000 votes remained to be
counted. In fact, because turnout had been higher than the VNS
forecast, there were still twice that many votes outstanding. Since we
knew a high proportion of the uncounted ballots were in heavily
Democratic precincts, Vice President Gore still had a much better
chance of overtaking President Bush than it appeared from the VNS
reports, despite an apparent margin of 50,000 votes.
It wasn't really 50,000 votes, however. As the Committee is also
aware, there had been another official error. Because of a defect in a
data storage device in a Volusia County election computer, President
Bush's statewide lead was overstated by 20,000 votes, further
bolstering the impression that Mr. Gore had no chance to catch up.
The correction of the Volusia County error did not appear in either
the AP or the VNS tabulations until after the networks had committed
themselves, declaring President Bush the winner at about 2:15 a.m.
AP was not yet ready to follow suit. Even before the Volusia County
correction was made, and even allowing for the inaccurate VNS estimate
of the number of remaining votes, AP believed Mr. Gore retained a slim
chance of overtaking President Bush. That judgment was based on the
collective wisdom of AP reporters and editors in both Miami and
Washington.
The margin separating the two candidates had descended in the AP
tabulation from over 100,000 at 1 a.m. on November 8 to about 45,000
shortly after 2 a.m. In the next 15 minutes, it plummeted to less than
16,000. AP continued to report that it was too close to call.
Because the hour was late, already past the deadlines of many
newspapers, and because the networks were reporting that President Bush
had won, editors around the country wondered why they had received no
such declaration from AP. Many of them called our bureaus or our
general editing desk in New York to demand an explanation.
As another hour passed and the pressure increased, however,
President Bush's lead continued to shrink. Shortly after 3 a.m., it
stood at just over 6,300. At 3:11 a.m. AP transmitted a note to editors
and broadcasters on its news wires advising them that with 6,000 votes
separating the candidates' totals and with votes outstanding in heavily
Democratic Broward and Palm Beach, the outcome remained uncertain.
Just minutes later, the gap closed to roughly 2,000, and there it
remained.
AP is proud of its century and a half of election result tabulation
and political reporting. We believe our experience and our commitment
to accuracy and fairness have produced an extremely valuable service
for our membership and for the American public.
Chairman Tauzin. Thank you very much. I might add that we
all share a mutual respect, I hope we understand that, for our
different roles and we appreciate your concerns and we have
tried to be very sensitive to them throughout the process.
Let me add one fact. I don't believe that AP was invited to
testify in the 1980's. But I know the networks were, and they
did participate in hearings in the 1980's. And we did have
these conversations in the 1980's as a precedent to the ones we
are having today.
Mr. Boccardi. We were not involved.
Chairman Tauzin. I don't believe you were involved; that is
correct.
Let me also add, and I know this is not a rule of our
committee, but you all may respectfully decline to answer any
question if you think we are intruding. We will always give you
that right, and I hope you will use it lightly. Obviously we
are here to find the truth and the facts, but you always have
that right in respect for the different roles that we play.
Let me turn to our next witness, Mr. David Westin,
president of ABC News in New York. David, if you will join me
in the process.
David, under the committee practice for investigative
hearings, it is the practice to take testimony under oath. Do
you have any objection to testifying under oath?
Mr. Westin. No, sir.
Chairman Tauzin. The Chair advises you that under the Rules
of the House, the rules of the committee, you are entitled to
advice by counsel. Do you desire to be advised by counsel
during your testimony today?
Mr. Westin. No, thank you.
Chairman Tauzin. Then if you will raise your right hand, I
will swear you in.
[Witness sworn.]
Chairman Tauzin. You are properly sworn under oath, and you
may give your testimony, sir.
TESTIMONY OF DAVID WESTIN, PRESIDENT, ABC NEWS
Mr. Westin. Thank you. I appear before the Chair and the
members of the committee today to talk about ABC News coverage
of the election in November of 2000. In addition to a copy of
my remarks, I also would like to submit for the record a copy
of the February 8 statement that we submitted to committee
staff earlier in which we go through in detail.
Chairman Tauzin. Without objection, it is admitted into the
record.
Mr. Westin. Thank you very much. Those of us at ABC News
are very proud of the job that we have done over many years in
covering elections. From the days of Frank Reynolds and Howard
K. Smith through the days of David Brinkley, to the team of
journalists headed by Peter Jennings today, we have done
everything within our power to make sure that we report
elections to our audience in an accurate and timely fashion.
There is literally nothing that we do that is more important to
us. I am pleased that by and large, there have not been
exceptions. We have succeeded in our mission of being accurate
and timely.
But let me start right at the outset and say in November,
we failed twice in the projections we made for President in
Florida. Those were serious mistakes. We take them seriously.
ABC News is responsible for those and as the head of ABC News I
am responsible for them. But ultimately, the American people
will hold us responsible for what happened.
We all are mindful as well, as the Chair has said more than
once today, that in our system of government, it is ultimately
the marketplace of ideas protected by the United States
Constitution that will correct those mistakes for us. They
provide the check on us, not the government.
The morning after the election on November 8, early in the
morning, we began our investigation to find out as much as we
possibly could about what happened and what went wrong. To
answer the question earlier from the Chair, that investigation
was conducted by internal people from ABC News who are
responsible for our standards and practices, by in-house
counsel from ABC, Inc., which is our corporate parent, and
ultimately by the outside law firm of Verner, Liipfert located
here in Washington. So that is who participated in the
investigation.
From the investigation we learned a number of things, many
of which, frankly, have been reported here. Early on, 2 weeks
after the election, we came out with a statement in which we
addressed a number of the issues that have been raised here,
both the commitment to do whatever it takes to get VNS fixed;
to address things like absentee ballots; to address things like
errors that arise in exit polls; also to support a change,
frankly, in ABC News policy so that we will no longer project
the winner of a State until all polls, every one of them have
closed within the State, which is, as you indicated earlier,
Mr. Chairman, a change from what we talked about in 1985.
We also in that early stage said it was critically
important as we go forward that we are much clearer and more
emphatic about what a projection is and what it isn't. It is
not reporting the ultimate certified result of a race; it is a
statistical estimate which always has a margin of error in it,
and we need to do a better job of explaining that to all of our
viewers.
We are committed to making whatever changes need to be
made. At the same time, I have to say with all of the problems
we have talked about today and I have heard about with VNS and
with exit polls and with absentee ballots, all of which are
terribly important and need to be addressed, we could have
served our viewers much better if on election night we had been
clearer about what was going on when we were making a
projection. We have tried to do that over the years, but we
have fallen back in part, frankly, because of the success of
the process over many years. I think it bears reminding that
for many, many years this process has generated accurate
projections. And even on November 7, ABC News made some 100
projections of races; and we were wrong in one instance, a very
serious one and we take it very seriously, but we do need to
put it into perspective.
In addition, in reflecting on our system and how it works,
we concluded that we could have served our viewers better if we
had done a better job of insulating the key people, our
professional analysts who are looking at the data as they come
in and looking at the statistical models, had insulated them
better from the competitive pressures that inevitably arise. I
can tell you that if you sit there and four of your competitors
have projected the next President of the United States and you
haven't, there is a lot of competitive pressure, no matter what
anybody tells you.
You will notice, I haven't really talked about VNS and
there is a reason for that. We are a co-owner of VNS, we are on
the board of VNS. We are responsible, ultimately, for the
accuracy of the output of VNS, just as if ABC News people were
out there conducting all the exit polls and gathering all the
raw vote data from the counties and from the precincts, and I
do not want to shrink from that responsibility. I have to be
direct and honest with you about it. Having said that and
having reviewed what we have reviewed at this point, I must
also tell you that I continue to believe that properly
corrected, and there are a number of corrections we can talk
about some more, properly corrected, the VNS approach supported
by our decision desk and our professional analysts remains the
best and most accurate way of doing timely reporting of the
election, which is our only goal, ultimately.
Now, in conclusion, let me address for a moment the
possibility of effect on voter behavior of early projections,
because there has been a lot of discussion about that today.
From what I know, and I am not a professional in this area,
some academics have looked at it, it is inconclusive, it is
complicated. Nevertheless, as I said earlier, we are committed
to try to help address this in two ways.
No. 1, not to project the race in any State until all of
the polls, not just the substantial majority, have closed in
that State.
But No. 2, we wholeheartedly support the efforts of certain
members of this committee to adopt a uniform poll closing time.
We think that is the right approach and speaking
personally, I would be perfectly happy with a world in which we
don't make any projections, in which there is a uniform time
for closing and there is a system of voting in this country
that is so reliable and so instantaneous that ABC News, in
future elections, can simply get on the air and say, these are
the actual bona fide, certified results. I would be very
pleased with that. But until we can get to that time in this
country, ABC News, I can assure you, will remain committed to
doing our very best to reporting accurately and in a timely
fashion on all elections. Thank you.
[The prepared statement of David Westin follows:]
Prepared Statement of David Westin, President, ABC News
Mr. Chairman and Members of the Committee: I am here today to
discuss ABC News' reporting on the presidential election of 2000. I am
submitting with a copy of these remarks the complete ABC News statement
concerning its election reporting last year.
Those of us at ABC News are proud of our record of covering
elections. From the days of Howard K. Smith and Harry Reasoner, through
David Brinkley, to our present group of journalists headed by Peter
Jennings, we have always done everything in our power to bring an
accurate and timely account of the nation's elections to our audience.
There is nothing we do at ABC News that is more important than covering
elections.
With remarkably few exceptions, we have succeeded in our mission of
accurate and timely reporting of elections. But last year there was an
important exception: the projections we made of the presidential race
in Florida. The fact that this was one of the closest presidential
races in history is no excuse; the fact that we correctly projected the
results of almost 100 other races that night is no excuse; the fact
that others had similar difficulties is no excuse. Ours was a serious
mistake for which we are responsible and for which the American people
rightly hold us accountable. At the same time, all of us are mindful
that in our system it is the marketplace of ideas created under our
Constitution--not the government--that provides the check when we in
the press make a mistake.
The morning after the election, at ABC News we began an
investigation to determine everything we could about what went wrong in
our Florida projections. Although we continue to believe that our
system for making projections is fundamentally sound, we have found
several things that we must correct and several things that we must do
differently or better. We are committed to doing everything we can to
correct the problems we have identified.
In brief, our review shows the following problems:
1. We did not accurately anticipate the absentee balloting in Florida.
2. We did not adequately protect against the error that comes from some
people not responding to exit polls.
3. People made mistakes in putting actual Florida vote tallies into the
system.
4. The second ABC News projection in Florida was based on a
dramatically mistaken estimate of the remaining, uncounted
vote.
5. All election projections are based on a comparison of the current
election with the past election that appears to be most
similar. In Florida, the computer model chose the wrong past
election for comparison.
In our written statement of February 8, we set out the various
steps we will take to avoid each of these problems in the future. But
even with all of the problems that affected projections in Florida on
election night, ABC News could have served its audience better if it
had been clearer and more emphatic about the nature of election
projections and their basis. No matter how reliable, projections are
just that--statistical estimates based on past experience and the data
available. They are not the same as reporting the actual results of the
election, and they are always subject to some margin of error (as was
shown so dramatically in Florida). In addition, ABC News could have
protected itself against error by resisting the inevitable competitive
pressures that came with knowing that other news organizations were
projecting the next President of the United States while we were not.
Please notice that in my discussion I have not referred to the
Voter News Service. As an owner, we are responsible for the accuracy of
what VNS did in 2000--as responsible as if ABC News had itself
conducted all of the exit polls and gathered all of the actual vote
tallies across the country. With all of the criticism of VNS, based on
everything we now know, we continue to believe that an improved VNS
remains the best way for us to gather accurate, timely information
about individual races on election night.
Finally, there has been much discussion over the years of the
possible effect of election night reporting on whether voters go to the
polls. I know that there has been some academic work on this subject,
that it is a complex area, and that the academic work is largely
inconclusive. Nevertheless, ABC News from now on will not project the
results of a race in a state where any polls remain open--even if it is
a minority of the polls.
In addition, some--including some on this Committee--have advocated
adopting a uniform time across the country for all polls to close. ABC
News wholeheartedly supports such proposals. Indeed, I personally would
welcome the day when, shortly after all the polls closed at a set time,
we could simply report the actual results of all of the elections
because they had been quickly tabulated by a reliable, instantaneous
system. But until that day arrives, ABC News will remain committed to
providing its audience with the most accurate and most timely reporting
of elections possible.
Thank you.
Statement of ABC News Concerning the 2000 Election Projections
On November 7, 2000, the United States set out to elect a new
President. Leading up to Election Day, all indications were that the
race would be very close. As it turned out, it was the closest race in
recent history (in terms of electoral votes) and was not resolved for
over a month, when the Supreme Court of the United States ended the
legal challenges to the vote count in the State of Florida.
During the evening hours of November 7 and the early morning hours
of November 8, ABC News projected the winner of the presidential race
in 49 states and the District of Columbia.\1\ In each of these but one,
ABC News' projections were correct. But in what turned out to be the
key state of Florida, ABC News made two projections, one of them
mistaken and the other premature.
---------------------------------------------------------------------------
\1\ ABC News never projected a winner in the presidential race in
Oregon, which was resolved several days later, after the last of the
absentee ballots were counted. In addition to the projections in the
race for President, ABC News made 45 projections in races for the
Senate and for Governor, all of which were correct. ABC News never
projected a winner in the key Senate race in Washington, which was
ultimately resolved several days later as well.
---------------------------------------------------------------------------
In this statement, we discuss the practices and procedures followed
by ABC News in making its election projections, the reasons for the
flawed projections in Florida, and the steps we are taking to prevent a
recurrence of the mistakes we made on Election Night. This statement
follows a comprehensive review conducted by ABC News, ABC's in-house
counsel, and ABC's outside counsel. Among other things, we: (1)
reviewed transcripts and videotapes of ABC News' election coverage; (2)
interviewed members of the ABC News team responsible for making
election projections and producing ABC News' election coverage; (3)
reviewed archival copies of computer screens containing some of the
data and statistical models provided to ABC News on Election Day by
Voter News Service (``VNS''); and (4) reviewed post-election reports by
VNS and others analyzing the Election Night projections.
a. abc news' practices and procedures in projecting election results
1. Background
For many years, ABC News has included in its election coverage
projections of likely winners in individual races. When done properly,
such projections provide our viewers with highly reliable and timely
insights into the election, including the likely outcome of races in
the various states.
These projections go beyond mere reporting of actual vote tallies
as they come in. They involve the interpretation of sophisticated
statistical models that evaluate the exit poll and actual vote data in
various ways. Although grounded in mathematics and science, projections
of likely outcomes always depend in critical part on the informed
judgments of knowledgeable analysts. To make these judgments, ABC News
relies on teams of experts, including political scientists and
statisticians, with the experience and acumen necessary to interpret
the data properly.
2. The Role of the Voter News Service
In 2000, as in previous years, ABC News relied heavily on data and
statistical modeling from VNS in making its election projections. VNS
was established in 1994 through the merger of two predecessor
organizations. The first, the News Election Service, was founded in
1964 and collected raw vote data. The second, Voter Research and
Surveys, was formed in 1989 and did three things: (1) polled voters as
they exited the polling place; (2) reported the exit poll results; and
(3) used statistical models to help project race results based on the
exit polls and on vote data. Since 1994, VNS has performed the
functions of both of these predecessor organizations.
VNS was founded in 1994 by ABC, NBC, CBS, CNN, and the Associated
Press; in 1997, Fox became a full member. The members collectively
share the costs of operating VNS and govern it through a Board of
Directors consisting of one voting representative from each member. VNS
data and analysis are provided to all of the members equally. In
addition, many other news organizations subscribe to VNS, paying fees
that partially defray operating expenses in exchange for projections
and some VNS data. By sharing the costs among the members and
subscribers, VNS collectively provides a better service to all than any
individual member or subscriber could afford on its own. As one of the
owners of VNS, with a substantial role in its management and
supervision, ABC News (with the other members) is ultimately
responsible for the reliability and accuracy of VNS' product.
Each election, VNS collects survey responses provided to VNS
personnel by voters as they leave polling places (``exit polls''),
actual vote tallies from selected precincts, and the vote count from
all reporting counties, cities, or towns nationwide. VNS loads this
material into a central computer that feeds information to VNS members
and subscribers. VNS also provides members with statistical analyses
under various VNS models, which use different methods to extrapolate
from the data received by VNS at that juncture. Some models include
pre-election poll data; some include exit poll data; some models use
geographical distinctions within a state; some reflect political
distinctions within a state; others rely on actual vote tallies as
reported by precincts or by counties. VNS models also estimate the
remaining vote outstanding for precincts not yet reporting, and provide
a check on the accuracy of exit poll data by comparing that data to
actual vote tallies as they come in.
Before polls close and actual vote tallies become available, the
VNS models rely upon exit poll data provided by VNS, together with
various pre-election poll data. As actual vote tallies become available
from precincts, these data are included in the models, replacing exit
poll data. When county data is available, it is included in several
other models.
Early in the afternoon of Election Day, VNS begins providing its
members with preliminary results of the first exit polls taken. As the
day progresses, VNS updates these results, including later exit poll
data and actual vote tabulations. All of these data are provided to VNS
members so that they may evaluate for themselves whether and when to
project the results in a race. VNS also makes its own projections for
the benefit of its members and subscribers.
Prior to the election of 2000, those most closely involved with VNS
can remember only one instance in which a VNS projection ultimately
proved wrong. In the 1996 New Hampshire Senate race, exit poll data
available at the time the polls closed showed the Democratic candidate
to have such a commanding lead that VNS members (including ABC News)
projected the outcome of the race. When the actual vote tallies began
to come in, it became apparent that the exit poll data were seriously
flawed and the projections were retracted. After the 1996 error, VNS
conducted its own study and retained outside experts to examine what
happened and how to avoid similar mistakes in the future. Despite
intensive study, the experts could find no definitive solution to the
problem that could be applied across states and elections. In 2000, ABC
News included as part of its team two of the experts most knowledgeable
about the 1996 New Hampshire race, so that they could brief the other
ABC News experts responsible for making projections.
3. ABC News' Decision Desk
On Election Day, ABC News employs a separate unit of professionals
charged with deciding whether and when to make projections. This group
is generally referred to collectively as the ABC News ``Decision
Desk.'' It is headed by an expert journalist who is experienced both in
covering previous elections and in the statistical analysis of
elections.
The Decision Desk on November 7 consisted in principal part of four
decision teams. Each decision team included two individuals chosen for
their experience in covering past elections and/or their background in
statistics or political science. The 50 states and the District of
Columbia were divided among three teams, each responsible for making
projections of statewide elections (for President, for Senate, and for
Governor) in certain states. The fourth team was assigned to follow
certain races in the House of Representatives that ABC News had
identified in advance as having particular importance in the election.
In order for ABC News to make a projection in any race, both members of
the decision team assigned to that race had to agree that the
projection was justified.
The ABC News Decision Desk also included a team of two individuals
reporting directly to the head of the Desk and responsible for
monitoring the overall presidential race and the battle for control of
the Senate. ABC News also retained two experts who were posted at VNS
headquarters as a liaison between the ABC News Decision Desk and VNS.
The Decision Desk independently studies the data as they are
reported by VNS, analyzes the results indicated by the VNS application
of the statistical models to these data, and decides whether and when
it is appropriate for ABC News to project an outcome in each race.
Sometimes this projection comes after VNS has made its own projection;
often it comes before VNS does so.
If the margin reflected in exit polls is sufficiently large to
justify making a projection based on these data alone, ABC News may be
able to project results at the time polls close. In determining whether
it can make an accurate projection based on exit polls and other
research at the time of poll closing, ABC News relies on the following
considerations:
1. the extent to which all of the exit poll data from the several exit
polls taken throughout the day are complete and available;
2. the size of the lead for a particular candidate indicated in the
exit poll models;
3. a comparison of the margin indicated in the exit polls with a
statistical calculation of the margin of error;
4. a comparison of the results generated by the different VNS
statistical models;
5. prior estimates of the race in the particular state;
6. special factors that might affect the reliability of exit poll
results, such as the size of past absentee balloting in the
state and the distance restrictions placed on VNS personnel
conducting local exit polls;
7. past experience in projecting results in the particular state or in
states having similar characteristics; and
8. any special messages from VNS concerning problems or irregularities
that may have arisen during the collection of exit poll data.
If exit poll results do not demonstrate a sufficiently decisive
lead for one candidate after assessment of these factors, then ABC News
considers the following factors as actual vote tallies arrive, in
addition to exit poll data, pre-election estimates, and the patterns of
prior elections:
1. the size of the lead and the margin of error indicated in the
various models as actual votes from the precincts are
substituted for exit poll results;
2. the size of the lead and the statistical margin of error indicated
in the models analyzing actual vote tallies in selected
precincts and in overall county results;
3. the degree to which the exit poll data for the state differed from
actual vote tallies in precincts where exit poll data are
available;
4. the number, percentage, and location of precincts from which actual
vote tallies have been received;
5. the number, percentage and location of counties from which actual
vote tallies have been received;
6. the likely outstanding absentee vote not yet counted;
7. the likely other vote not yet counted; and
8. the percentage of the remaining vote outstanding that the trailing
candidate would have to garner in order to prevail.
In each case, the question is whether ABC News in its independent
journalistic judgment can conclude that the data make it appropriate to
project the outcome in a given race--that is, that the results are
decisive enough to make a projection with great confidence that it will
be right.
In addition to these steps, ABC News makes an effort to say
explicitly on the air that it is reporting a projection, not an
accomplished fact. And, since 1985, ABC News has followed the policy of
not projecting any statewide race until the polls in the state have
closed. In the few states with multiple poll closing times, ABC News
has not projected the results of races until the substantial majority
of the polls have closed.
b. abc news' election projections in 2000
On the morning of November 8, ABC News initiated an investigation
to determine the cause of its erroneous and premature projections in
the Florida presidential race and the measures it should take to avoid
similar problems in the future. In addition, VNS conducted its own
internal investigation and--at the urging of ABC News and other
members--commissioned a thorough review of its actions by outside
experts. These investigations lead us to the following primary findings
and conclusions about the problems in Florida and some areas for
improvement.
1. The Flawed Florida Projections
a. ABC News' Projection for Vice President Gore--The first of ABC
News' Florida presidential projections in the 2000 election came
shortly before 8:00 p.m. EST on November 7. VNS projected that Mr. Gore
would win the presidential race in Florida at 7:52 p.m., after some 90%
of the polls in state had closed (but before polls in the panhandle of
the state were to close at 8:00 p.m.). At 7:55, ABC News Radio reported
that ABC News projected that Mr. Gore would win in Florida. At that
time, the ABC Television Network was in commercial and local broadcast
time. ABC News did not project Mr. Gore to win in Florida on the ABC
Television Network until shortly after 8:00 p.m., after all polls in
the state had closed.
A review of the computer data preserved by VNS that night \2\
indicates that, as of 6:40 p.m., exit poll data and statistical models
revealed a lead for Mr. Gore in Florida, but not by a sufficient margin
to warrant a projection. As a result, ABC News did not project a winner
in the Florida race for President when 90% of the Florida polls closed
at 7:00 p.m.
---------------------------------------------------------------------------
\2\ VNS automatically archived the primary data (or ``decision
screens'') it sent to members at only 10 and 40 minutes past each hour
on Election Day. Minute-by-minute screens reflecting then-available VNS
data were not retained.
---------------------------------------------------------------------------
By 7:40 p.m., VNS was reporting actual vote tallies from eight of
the sample precincts. These actual vote tallies significantly reduced
the margin of error for the projections under the various statistical
models, and therefore indicated a greater confidence in the accuracy of
the projections. Based on these indications, and in accord with the
factors listed above, the ABC News Decision Desk projected that Mr.
Gore would win in Florida shortly before 8:00 p.m. At 8:10 p.m. and
8:40 p.m., the numbers continued to show a solid lead for Mr. Gore in
Florida. In fact, at that time, a comparison of the actual vote tallies
with the exit poll data in those eight precincts suggested that the VNS
exit poll models had actually understated the extent of Mr. Gore's lead
in the state.
At 9:10 p.m., the statistical models continued to indicate
sufficient strength to justify the projection for Mr. Gore--even with
36% of the county vote tallies reported. But the precinct models showed
some weakening. At about 9:40 p.m., VNS began to send messages to its
members that called into question the accuracy of the Florida
projection. And by 10:10 p.m., the models incorporating real vote
tallies were mixed, with some continuing to project Mr. Gore the winner
with sufficient confidence to warrant the projection, but the county
vote model pointing to Mr. Bush, although without sufficient certainty
of accuracy to warrant a projection in his favor. At approximately
10:00 p.m., ABC News withdrew its projection that Mr. Gore would
prevail in Florida.
Based on subsequent reviews, it is now apparent that there were
three principal causes of the flawed projection for Mr. Gore in
Florida. First, the exit poll data actually overstated--rather than
understated--Mr. Gore's lead. Although the risk of such error normally
is checked by the comparison of the exit poll data in the first several
precincts with actual vote tallies in these precincts, in this
particular case the first precincts reporting actual vote were atypical
of the larger sample.
Second, the first Florida projection was based in part on a
mistaken estimation of the absentee vote in Florida. Although VNS
anticipated a significant absentee vote in the state and expected that
the absentee vote would favor Mr. Bush, it turned out that there was a
much larger vote than anticipated, and that vote was slightly more
favorable to Governor Bush than predicted.
Finally, the projection for Mr. Gore resulted in part from the VNS
system failing to choose the most appropriate past race to use in its
models. It chose the 1998 Governor's race, rather than the 1996
presidential race. As it turned out, the latter was more similar in
pattern and result than the former. Absent any of these three VNS
errors--the mistaken check on the exit poll data, the mistaken
estimation of the absentee vote, and the choice of the wrong past race
to use in its models--ABC News likely would not have found the data
sufficient to make the first projection in the Florida presidential
race.
b. ABC News' Projection for Governor Bush--From 10:10 p.m. until
1:40 a.m. EST, the data and statistical models continued to be mixed,
with all but one model (the one that included actual vote count from
the counties) indicating that Mr. Gore would prevail. By 2:10 a.m.,
however, the data provided by VNS indicated that a full 96% of all
precincts had reported and that, given the projections of the remaining
outstanding vote, Mr. Gore would have had to win over 63% of the
remaining vote to prevail. Based on these data and these estimates, ABC
News projected at 2:20 a.m. that Mr. Bush would prevail in the Florida
presidential race.
It now appears that two primary factors precipitated the second ABC
News projection for President in Florida. First, raw vote data coming
into VNS from Volusia County significantly overstated Mr. Bush's totals
and significantly understated Mr. Gore's totals. Normally such
variations in a single county would not be significant, but with the
race in Florida as close as it turned out to be, this variation alone
led ABC News to have more confidence in its projection than was
warranted.
Second, the VNS model projected significantly fewer outstanding
votes at 2:10 a.m. than in fact was the case, leading the VNS model to
underestimate the outstanding vote and thereby to overestimate the
percentage of the vote that Mr. Gore would have to receive to prevail.
Once again, these mistakes in the data and the models led ABC News to
make a flawed projection.
2. Projections in Other States
As noted above, ABC News made 49 other projections of state races
for President on November 7-8, and each one was correct. Overall,
leaving Florida aside, ABC News correctly projected that Mr. Bush would
prevail in 29 states and that Mr. Gore would prevail in 19 states and
the District of Columbia. Of these, ABC News projected the winner at
the time of poll closing in 28 states, including 16 states for Mr. Bush
and 11 states and the District of Columbia for Mr. Gore. Of the 21
states in which ABC News waited to project a presidential winner until
after polls had closed (other than Florida), ABC News projected 13 for
Mr. Bush and 8 for Mr. Gore.
Some have questioned not the accuracy, but the timing of some of
our projections. In particular, they have questioned the delay in the
projection of some of the states won by Mr. Bush. Based on our review
and the analyses described above, the timing of each of these
projections is fully explainable by the data available at various times
of the night, the application of uniform statistical models to those
data, and prior experience with the states involved. There is no basis
whatsoever for concluding that there was any intentional bias on the
part of anyone who took part in the projection process at ABC News.
In determining whether and when to make a projection, there are a
variety of factors ABC News considers, as set forth above.
Significantly, however, the ultimate actual margin of victory in a
state in no way indicates the speed with which a projection could have
been made with a sufficient assurance of accuracy. Rather, the question
is whether at any given time the margin shown by the exit poll (and,
when available, actual vote tallies) is sufficiently larger than the
statistical margin of error that ABC News can make a projection with a
high level of confidence that it will be accurate. This statistical
margin of error can be affected by a number of things, including the
number of precincts sampled, the relative homogeneity of the state in
question, and the percentage of actual vote available at the time of a
projection.
This last factor is extremely important, as the ability of VNS and
ABC News to assess the accuracy of close exit poll data and the
certainty of a projection often depends on how quickly a given state
reports actual vote data. In some states, precincts and counties report
vote totals relatively quickly, and may facilitate an earlier
projection; in others, vote totals arrive more slowly, and may delay
the projection of the race.
A careful review of ABC News projections in states other than
Florida shows that in each case--whether for Mr. Gore or for Mr. Bush--
the projection was not only accurate, but was made at a reasonable
time, given the available data and the need for great confidence in its
accuracy.
Some have also raised concerns that, before all polls have closed
in a state with multiple poll closing times, (like Florida) projections
of race could influence voting behavior in other parts of the state.
There have also been claims that projecting winners in presidential
races in states where polls have closed may affect voting behavior in
other states where polls remain open. These questions have been
researched extensively for many years, with no clear answer.
c. areas for improvement
Based on our experience in making the Florida projections, our
post-election review of ABC News' practices and procedures in general,
and our review of other reports about the events of Election Night, we
have identified the following primary areas that warrant further
attention and improvement.
1. Improve the Accuracy and Reliability of VNS' Data and Statistical
Analyses
VNS has a long and enviable track record of accuracy in projecting
races over the years. In the November election, however, VNS fell
short.
In the days since the election, VNS' performance has been analyzed
extensively by internal and external investigations. Most recently, the
VNS Board of Directors commissioned and released a detailed report by
an independent consultant, the Research Triangle Institute (``RTI'') of
North Carolina, containing comprehensive findings and recommendations.
On the basis of the RTI report and our own review, we believe VNS
must do a better job in future elections in the following principal
areas:
First, it must account more accurately for the size and likely
outcome of absentee ballots. In 2000, VNS made special efforts to
account for absentee ballots in Washington, Oregon, and California by
conducting telephone polling targeted at these ballots. In Florida,
however (as noted above), it relied upon a rough estimate of the size
and the outcome of the ballots. In the future, one or both of two
things must happen: (1) VNS should make similar, targeted efforts in
all states where there is reason to believe that absentee balloting may
affect the accuracy of projections; and/or (2) ABC News will need to
take into account the increased risk of inaccuracy from VNS' not having
made such efforts and, as a result, be more conservative in its
projections in those states.
Second, VNS must do a better job of quality control. For example,
in 2000, there was a plan to include in the VNS computer models process
data from the separate Associated Press reporting of actual vote
tallies. This would have provided a check on the accuracy of VNS' data.
ABC News believed on Election Night that this had been done; as it
turned out, it was not. In the future, the AP (or other, similar
sources of data) should be included in election projections--either by
VNS itself or by ABC News if VNS is unable to include such data.
Third, VNS must make adjustments in its statistical models to
ensure that the best past election in a state is selected for
comparison with exit poll and raw vote data. VNS itself has suggested
that one solution may be to consult more than a single past election.
Fourth, VNS must devote further study to the causes of
discrepancies between exit poll data and actual vote returns within the
same precinct. In any given race, such discrepancies can favor either
candidate, but on average over time they have been shown to favor
Democratic candidates somewhat more than Republican. Experts believe
these discrepancies may result in part from the refusal of some people
to respond to exit poll surveys as they leave the polls. However,
studies have not shown a clear direct relationship between overall
response rate and exit poll accuracy. Moreover, the size of the
discrepancy and whether it favors the Republican or the Democrat is
highly variable from precinct to precinct, state to state, and year to
year, and experts have yet to develop an overall statistical solution.
Nevertheless, we must work to develop reliable ways either to reduce
the causes of the discrepancies or to compensate for them.
Fifth, VNS must improve its system to provide better correction on
inaccuracies in exit poll data. As discussed above, it was previously
thought that the comparison of actual vote totals in six selected
precincts against exit poll results in those same precincts would give
some reasonably accurate indication of the reliability of the exit poll
data in all the precincts. The experience in 2000 with the exit poll
data in Florida demonstrates that this is not the case.
Finally, VNS must do a better job estimating the outstanding vote.
This is a crucial piece of information in any close election.
Based on its review, ABC News at this point believes that with
improvements such as those outlined above VNS can remain a highly
reliable means for analyzing and reporting election results. Any
system, however, no matter how sophisticated and how reliable, is
inherently fallible. ABC News will remain open, therefore, to any
reasonable alternative sources of information that can help to improve
the accuracy of its election projections.
2. Further Insulate ABC News' Decision Desk from Competitive Pressure
In the past, and in 2000, the ABC News Decision Desk has been
located separately from the remainder of ABC News editorial operations
on Election Night. Communications between the Decision Desk and those
responsible for ABC News election coverage have been structured through
a single senior producer located in the control room.
Until now, however, ABC News has not sought to restrict access of
members of its Decision Desk to the reporting of other news
organizations, including competing television news organizations. It
was thought that the knowledge of what other credible news
organizations were and were not projecting could be helpful to the Desk
in determining when it was appropriate to make a projection.
Competition, in news reporting as in other enterprises, can be a
good thing. It can spur us to work harder, do better, be faster. But
competition that encourages a journalist to report a story prematurely
is bad. In the particular instance of the Decision Desk, it is most
important that the individuals making projections do so based on two
things: the data provided to them (from exit polls, from actual vote
tallies, and from statistical models), and their own experience and
judgment. They should not be distracted or influenced by the decisions
of other news organizations.
3. Improve the Manner in which ABC News Reports and Discusses its
Projections
Although valid statistical methods applied to raw vote tallies and
exit poll data are the best and most accurate means of projecting
election outcomes, they also pose risks if not properly used and
explained. First and foremost, they are projections--not actual results
of elections. As such, there is always some margin for error, some
chance that they will ultimately be proved wrong.
ABC News for some time has attempted to describe its predictions of
the likely winners in election contests as ``projections.'' We try to
avoid saying on air that we are ``calling'' a race with the implication
that the election is truly over simply because we feel sufficiently
confident in the statistics to make a prediction. This policy was
generally followed on November 7-8 in the initial announcements of
projections in the presidential races in individual states.
In reviewing the transcript of ABC News' Election Night coverage,
however, it appears that ABC News did not make sufficiently clear to
our audience the nature of projections in election races in several
respects. First, after having initially described the prediction as a
``projection,'' ABC News journalists on the air sometimes later
referred to the candidate as having ``won'' the race in a particular
state. Second, we were not always as careful in our use of language in
making projections in races for the House of representatives, the
Senate, and Governors. Third, ABC News did not explain adequately to
our viewers what a projection means. We did not make it clear that, as
with any statistical projection, there is a margin of error. We would
not be reporting the projection unless, according to our analysis, the
margin of error is sufficiently small. But we need to do a better job
of pointing out that our projection can be wrong.
d. conclusion
In light of our review, ABC News will implement the following
changes and clarifications to its practices and procedures in making
election projections. Many of these measures were announced on November
22, 2000.
1. ABC News will project the winner in a race in a given state only
after the last scheduled poll closing time in that state.
2. ABC News continues to support a uniform national poll closing time.
We also support efforts to reform balloting processes to enable
faster and more reliable official tabulation and reporting of
vote totals.
3. ABC News will continue to make projections only if they are
justified by ABC News' independent analysis of the data and the
relevant statistical models.
4. In making and discussing projections, ABC News will explain to its
viewers that they are informed, statistically based estimates
of the probable results of elections. Projections are not
reports of the actual, final results of elections.
5. ABC News will take all reasonable steps to insulate those involved
directly in making projections from the pressures of
competition from other news organizations.
6. ABC News will ensure that voting data from the Associated Press is
fully incorporated into its projections, providing a check on
inaccurate information. ABC News will remain open to additional
sources of information on election night, including national
exit polling conducted by organizations other than VNS.
7. ABC News will support a continuing comprehensive review of and
improvement in the operation of VNS, including improvements in
the collection of data, reporting of data, and application of
statistical models to those data. Further, ABC News will
provide its share of resources to ensure that these
improvements and upgrades are made as quickly as possible.
Chairman Tauzin. Andrew Heyward, president of CBS News in
York. Mr. Heyward, if you will help me through the process
again.
As you are aware, the committee is holding an investigative
hearing. When doing so, we have the practice of taking
testimony under oath. Do you have any objection to testifying
under oath?
Mr. Heyward. Given that it is your practice, Mr. Chairman,
it is okay with me.
Chairman Tauzin. Thank you, sir.
The chairman advises you that under the Rules of the House
and the rules of the committee, you are entitled to be advised
by counsel.
Mr. Heyward. No, thanks.
Chairman Tauzin. Do you desire--you do not. In any case
then, would you please raise your right hand and we will swear
you in.
[Witness sworn.]
TESTIMONY OF ANDREW HEYWARD, PRESIDENT, CBS NEWS
Mr. Heyward. Thanks very much Chairman Tauzin, Congressman
Dingell, members of the committee. I am Andrew Heyward,
president of CBS News; and I appreciate the opportunity to turn
on the microphone and to provide my comments on this subject,
news coverage of Election Night 2000.
CBS News and the other network news operations made very,
very serious mistakes that night, and they are mistakes that
all of us at the table and certainly I deeply regret. Our
Florida flip-flops were deeply embarrassing to us; and more
importantly, damaging to our most important asset, which is the
hard-won credibility we fought for over the years with our
viewers and listeners and Internet users. It is evident, in
retrospect, we should not have called Florida for either
candidate. Our method of projecting winners, one that, as you
have heard, has produced only six bad calls in over 2000 races
since the 1960's, failed us this time; and as a well-known
candidate would say, failed us big time in the very State that
held the key to this election.
That is why everyone at this table has acknowledged the
problems; and I think moved very quickly to address them, not
in response to outside pressure or to criticism, but at our own
initiative. The American people who are our viewers and
listeners deserve nothing less than this.
On November 14, CBS News appointed a distinguished three-
person panel, including a well-known outside expert, Kathleen
Hall Jamieson, who is Dean of the Annenberg School of
Communication at the University of Pennsylvania, to investigate
what went wrong and to recommend a set of steps for future
election coverage. On January 4, the panel issued an exhaustive
87-page report, and we made it available to you and every
American citizen on our Web site, CBS News.com. We have also
entered its recommendations into the record of these
proceedings.
The recommendations are far reaching and thoughtful, and
CBS News intends to adopt all of them. Very briefly, we will
strengthen the checks and balances on the CBS News decision
desk which is the entity responsible for analyzing exit poll
data and vote data on election night; we will beef up our news-
gathering resources on the ground in key States with
particularly close races and will toughen the criteria for
projecting winners in very close races. We will develop and
consult multiple sources for vote tallies. We will explain to
the audience very clearly how exit polls work and exactly how a
particular projection is made. We will clarify our language and
our graphics to distinguish more clearly between projections
and final results. We will also work with our network
colleagues to address problems with the Voter News Service;
and, if necessary, we will develop alternatives to VNS.
Finally, we will not project a winner in a State until all the
polls have closed there.
Now, I believe that these changes and similar ones that
have been announced by our competitors will go a very long way
toward ensuring the credibility that draws a vast national
audience to election night coverage on television. Having said
that, I think it is equally important to point out that I don't
accept all of the criticisms that have been leveled at the
networks.
The notion that the pattern of State by State calls
reflected bias against President Bush, for example, has been
rejected by every single outside expert who examined each of
the networks, even those experts, and you heard from them
today, who are the most highly critical of us. I was glad to
hear you say again today, Mr. Chairman, that the committee's
investigators found no evidence of intentionally misleading or
biased reporting.
This election also revived a decades-old debate about
calling races in States before all of the polls have closed
there. Our report, like the findings of the other networks,
rejects the argument that the first call in Florida, which
occurred about 10 minutes before the final 5 percent of the
State polls closed in the Panhandle, had any measurable effect
on voters. Nevertheless, as I mentioned earlier, given the
widespread perception that network projections do affect voter
behavior, CBS News has decided that in future elections we will
not project the winner in a State until all the polls have
closed there.
There is a simple way to resolve this issue once and for
all. And we have heard a lot about it today. One that CBS News
has advocated since the 1960's, and that is a uniform national
poll closing. We applaud the news that you, Mr. Chairman, along
with Congressmen Dingell and Markey and Stearns and several
other committee members are sponsoring such a bill.
Finally, I think it is important, and this is really
important to me, not to confuse news coverage of the election
with the election itself. It took the Nation, not the networks,
the Nation 5 weeks to pick a President. Let's assume for a
second that we had gotten Florida right and never projected a
winner there. The country would still have undergone its 5-week
marathon, and there would still have been debate about the
outcome and how it was reached.
We come here today voluntarily, out of our sense of duty to
this respected body and to the American people that all of us
here serve. I want to state for the record that I am very
grateful to you, Mr. Chairman, for your written assurances that
the hearing is not out to prove a point or make a political
statement and that this committee will, in all respects,
continue to be mindful of our first amendment rights and
protections in this matter. The Constitution does protect us
against unwarranted interference from government; but we, like
you, are accountable to the most important constituency in
America, the citizens of this great Nation. So please accept
that it is in that spirit that I am here to answer your
questions today. Thank you very much, Mr. Chairman.
[The prepared statement of Andrew Heyward follows:]
Prepared Statement of Andrew Heyward, President, CBS News
Good morning, Chairman Tauzin, Congressman Dingell and members of
the Committee. I am Andrew Heyward, President of CBS News. I appreciate
this opportunity to provide my comments on the very important subject
before us today: news coverage of Election Night 2000.
The election of 2000 will long be remembered as a test of our
democratic system--a test that the nation passed with distinction.
Arguably, our democracy emerged from the five-week ``Election Night''
even stronger than it was before. Whatever your views about its
outcome, the nation was able to resolve a complex and contentious
election peacefully, and we now have a new President of all the people.
I believe the network news divisions will also draw strength from
this unique experience and emerge as even better public citizens than
we were before November 7, 2000. That is because we have, since that
night, worked hard to confront what went wrong and to chart a series of
reforms that we believe will ensure, within the limits of what is
possible, that our mistakes are not repeated.
CBS News and the other network news operations made serious
mistakes that long, confusing night--mistakes I deeply regret.
Projecting Vice President Gore as the winner in Florida . . . then
retracting that projection . . . then projecting and retracting a
similar call for then-Governor Bush was not only embarrassing to say
the least, but it was damaging to our most important asset--our hard-
won credibility with our viewers, listeners, and Internet users. These
citizens have every right to expect accurate information above all
else. We would quickly lose our audience and soon our entire business
if people could not rely on the truth of what we say.
It is evident, in retrospect, that we should not have called
Florida for either candidate. Our method of projecting winners--one
that had produced only six bad calls in more than 2000 races since the
1960's--failed us this time, and in the very state that held the key to
the election. That's why everyone at this table has acknowledged the
magnitude of the mistakes, analyzed the problems we encountered, and
moved quickly to address them ``not in response to outside pressure or
criticism, but at our own initiative. The American people who are our
viewers and listeners deserve nothing less.
On November 14, CBS News appointed a distinguished three-person
panel, including a well-known outside expert, Kathleen Hall Jamieson,
Dean of the Annenberg School for Communication at the University of
Pennsylvania, to investigate what went wrong and recommend a set of
steps for future election coverage. On January 4, the panel issued an
exhaustive 87-page report, and we made it available to the Committee
and to every American citizen on our cbsnews.com website. It can be
read in its entirety at http:///CBSNews.com/htdocs/c2k/pdf/
REPFINAL.pdf. The report consists of an internal CBS News review of
what happened Election Night, an analysis of the Election Night
broadcast by Dr. Jamieson, and a historical perspective by Dr. Kathleen
Frankovic, CBS News Director of Surveys and a well-known political
scientist. The chairperson of the panel was Linda Mason, Vice
President, Public Affairs, a respected 34-year veteran of CBS News.
The report analyzes the mistaken calls in great detail and makes
recommendations for how CBS News should improve its coverage of
elections in the future. The recommendations are far-reaching and
thoughtful, and CBS News intends to adopt all of them. Here they are,
taken in their entirety from the panel's report:
recommendations
Changing How CBS News Calls Races
As an added precaution, assign a member of CBS News senior
management to head the Decision Desk. The goal is to provide a larger
and more authoritative context for each call. This person, who would
report to the president of CBS News on Election Night, would have
significant training in the decision process, with extensive knowledge
of the data screens and how they work. He or she would monitor the
editorial flow (in this case, the Florida breaking news) and integrate
it with the Decision Desk's activities. This senior manager would also
have to be able to withstand the competitive pressure if others made a
call and he or she argued that more facts were needed before CBS News
also made the call. CBS News has to be ready to be second or even last,
and can make a virtue of its patience and determination to be accurate,
even if it takes longer.
Move the Decision Desk into the Election Night studio.
This will promote constant contact between the news gatherers and the
analysts. The consultants who work at the various correspondent desks
on Election Night could also contribute to this dialogue. If a story is
breaking, as it was in Florida this year, there will be constant
interaction, instead of the Decision Desk functioning in a vacuum, as
it did this time in an office three floors from the studio.
Assign a correspondent to the Decision Desk. He or she can
dissect close races in detail, with precise descriptions of what went
into a call or why one has not been made. For example, he or she could
explain that one call was made using only exit polls, another using
exit polls and tabulated data, another not made at all because the exit
polls did not match historical patterns, and so on.
Identify the closest races and toughen the criteria for
making those calls. CBS News should insist on a critical mass of both
exit-poll and tabulated data before making a call in those close races;
similarly, a call should be withheld in those states until the level of
certainty meets an even higher standard than usual for calling a race.
Such precautions might have prevented the bad calls in Florida.
Develop a new category of ``leaning'' to describe some
races. These are races in which one candidate has a solid lead but CBS
News is not yet ready to make a call. This category could also be
displayed graphically and integrated into CBS News' overall projections
for the night. We should be willing to trade the illusion of certainty
for genuine credibility.
Check multiple sources for vote tallies. Make certain that
members of the CBS News Decision Desk compare VNS numbers to those in
the AP reports and on the Web sites of the Secretaries of State and, if
there is a discrepancy, find out why. It would have rung an alarm if
CBS News analysts had consulted those sources on Election Night 2000.
Strengthen our information gathering in close states. We
must unilaterally strengthen our operation by placing local political
experts in appropriate state election locations to help us obtain
actual vote numbers quickly and to assess the situation on the ground
in places where it appears that the race will be close. We should
conduct more pre-election telephone polls in closely contested states
to deal with the growing number of absentee voters, and to achieve a
better grasp of unique circumstances in each state. There will usually
be no more than 10 or 12 states in this category.
``Fuller'' Disclosure
Tell the viewers how calls are made, as often as possible.
We must explain regularly throughout the early hours of the broadcast
how the exit poll is conducted and what it shows, so that the audience
knows we are not consulting a crystal ball. The process should be less
mysterious, more open: it will be informative and interesting for the
audience to understand more of how we come to our conclusions. An
explanation of how the exit poll is conducted should also be posted on
the CBS News Web site.
Label calls appropriately. We should use the words
``projected'' or ``estimated'' early and often, and make the word
``estimate'' much larger on the CBS News graphics. We need to remind
the audience repeatedly that these are just predictions until the votes
are actually counted. We should stress this language, with
explanations, on the CBS News Web site.
Tell viewers why calls are not made. We must clearly
distinguish between races that are too close to call and races for
which there is simply not yet enough information.
The Future of VNS
Invest more in VNS to address its problems or form a new
consortium to build an alternative service. VNS, in a preliminary
review, cites its own imperfections: problems with the sample, with the
equipment, with the software and with quality control. If the decision
is to fix VNS, CBS News will have to spend more to address these
issues, as will the other VNS members. The alternative is to develop a
new service to perform the functions of VNS. This decision should be
made after members receive the final results of the review by the
outside group that is studying VNS.
If the decision is to fix VNS, CBS News should recommend
reorganizing the board. To date, the VNS board has been made up
primarily of polling or election-unit personnel from each network. We
suggest that the board be composed of a vice president from each
organization and that it focus on broad-based policy rather than on
day-to-day management.
Poll Closing
Change the policy for calling with multiple poll closings.
We recommend that CBS News not make a call in any state until all the
polls have closed in that state; this is a new policy. However, in
states with multiple poll closings where less than 5 percent of the
voting-age population remains after the first poll closing, or in
states that report early results themselves, we recommend using the new
``leaning'' characterization if appropriate. Under this recommendation,
for example, races in Texas, Kansas and Michigan--states where the
voting-age population remaining after the first polls close is very
small--could be described as ``leaning'' if one candidate has a solid
lead. A Florida race could also be described as ``leaning'' under this
formula because the states itself releases early results, even while
polls are still open in the Panhandle.
Support a uniform poll-closing bill in Congress. As CBS
News has done since 1964, we continue to urge the adoption of a uniform
poll-closing time. This reform would completely eliminate the
possibility of voters being influenced by reported results elsewhere in
the country; all results would be reported at the same time, as the
polls close across the nation.
Encourage turnout. During the broadcast, the anchor should
repeatedly urge people to vote, as Dan Rather did on Election Night
2000.
I believe that these changes, and similar ones announced by our
competitors, will go a long way towards ensuring the credibility that
draws a vast national audience to Election Night coverage on
television. We have performed this important role in the political life
of our nation since the advent of broadcast journalism. Nothing is more
important to us than meeting that responsibility with accuracy and
fairness.
On a personal note, I am very proud of the process by which CBS
News has publicly exposed the flaws in its Election Night performance
and procedures. I believe we have met this challenge with a candor and
thoroughness that few corporations have ever displayed in acknowledging
and addressing their own problems.
Having said that, I think it is equally important to point out that
I do not accept all the criticisms that have been leveled at our
industry since the election.
The notion that the pattern of state-by-state calls reflected bias
against President Bush, for example, has been rejected by the outside
experts who examined every network, even those who were the most highly
critical of us.
I can tell you from personal experience that the analysts and news
executives who interpret Election Night data are worrying about one
thing above all others: getting it right. Every projection is unique,
based on a complex and ever-shifting set of data. And as we have all
seen, the networks' calls are exposed for the world to see--and the
penalty for mistakes is severe. It is simply not credible that politics
plays any role whatsoever in the analysts' recommendations or in what
makes it onto the screen. I was glad to hear Chairman Tauzin say last
Thursday that the Committee's investigators ``found no evidence of
intentionally misleading or biased reporting.''
This election also revived the decades-old debate about calling
races in states before all the polls have closed there. Our report,
like the findings of other networks, rejects the argument that the
first calls in Florida--which came just 10 minutes before the final 5%
of state polls closed in the Panhandle--had any measurable effect on
voters.
However, as I mentioned earlier, given the widespread perception
that network projections do affect voter behavior, CBS News has decided
that in future elections we will not project a winner until all polls
have closed in a given state.
In this regard, it is worth reminding the Committee that there is a
simple way to resolve this issue once and for all--one that CBS News
has advocated since the mid-1960's. This is a uniform national poll
closing. We applaud the news that you, Mr. Chairman, Congressmen
Dingell and Markey and several other Committee members are sponsoring
such a bill.
Finally, I think it's important not to confuse news coverage of the
election with the election itself. This election was unique in the
experience of every living American. It took the nation--not the
networks, the nation--five weeks to pick a President. Let's assume we
had gotten Florida right and never projected a winner there: the
country would still have undergone its five-week marathon, and there
would still have been debate about the outcome and how it was reached.
Here is another excerpt from the CBS Election Night report, from
the section entitled ``Lessons Learned.''
``The election of 2000 revealed to the American people what
had been a dirty little secret known only to politicians: even
when elections are conducted with the best of intentions, they
are approximations, prone to human error, mechanical error,
confusion and disorganization . . . Across the country, not
every vote cast is counted. In fact, according to the Committee
for the Study of the American Electorate, for every 100 million
voters, nearly two million ballots will not be counted for
various reasons . . .
Against this background, it is important to consider how many
factors were beyond the control of the news organizations covering this
election, factors that affected each organization's ability to make
some crucial calls correctly. There was human error: election workers
improperly entered votes into the computer; precinct workers
incorrectly copied or misread ballot tallies because of poor
penmanship; voters made mistakes marking butterfly ballots; and ballots
were lost. There was machine error as well: punch cards were not read;
a memory disk malfunctioned in Volusia County; and there were other
mechanical problems.
VNS could not or did not correct for these factors. Hindsight is
always 20/20, and it is easy to observe in retrospect that VNS most
certainly should have done so. Instead, it relied on, among other
things, models and methods that had been very dependable in the past
but that came up short in this extraordinary election. In the Florida
exit polls, people reported how they had voted, assuming that their
votes were being counted. Some may not have been. VNS also did not
accurately factor in the absentee balloting. The unique circumstances
of the Florida election exposed problems at VNS that must now be
corrected.
But the ultimate responsibility for the calls we made lies with us
at CBS News. It was we at CBS News who analyzed the data from VNS and
decided when to make a call. And it is here where there are the
greatest lessons to be learned. We hope we have incorporated all of
those lessons in our recommendations for future election coverage.
CBS News will continue to strive for perfection, realizing that, as
was made all too clear by this long election, perfection in any human
endeavor is difficult to achieve and impossible to guarantee. What we
can guarantee is this: that, just as we have learned from our mistakes
in the past, we will learn from the mistakes made during this election
and adopt new policies and procedures that will guard against similar
mistakes being made in the future; that we will continue to reach for
the truth in all we do, and report to the public without fear, favor or
bias the events as they occur, no matter how complex or difficult the
story might be.''
We come here today voluntarily, out of our sense of duty to this
respected body, and to the American people we all serve. I want to
state for the record that I am very grateful for Chairman Tauzin's
written assurances that this hearing is, ``not out to prove a point or
to make a political statement''--and that this Committee ``will--in all
respects--continue to be mindful'' of [our] ``1st Amendment rights and
protections in this matter.''
I believe the network news divisions play a vital role in our
democratic process by informing citizens about the issues that shape
their lives. In effect, we get ``elected'' every time someone turns on
one of our programs or logs on to one of our websites. The Constitution
protects us against unwarranted interference from government, but we,
like you, are accountable to the most important constituency in
America: the citizens of this great nation.
Chairman Tauzin. Thank you very much. I deeply appreciate
that attitude. That is what this is indeed all about. As I was
asked this morning a little bit what this was about, I
identified it as a platform for all of us to come and tell the
American public what we are going to try to do to make it
better and that is essentially it. I thank you for that.
Our next witness will be Mr. Tom Johnson, Chairman and
Chief Executive Officer of CNN in Atlanta, Georgia.
Mr. Johnson, if you will help me through this again. You
are aware that the committee is holding an investigative
hearing and when doing so, it is the practice to take testimony
under oath. Do you have any objection to testifying under oath?
Mr. Johnson. I do not.
Chairman Tauzin. The Chair then advises you under our rules
and the rules of the committee you are entitled to be advised
by counsel.
Mr. Johnson. I am.
Chairman Tauzin. In that case, would you raise your right
hand, sir?
Oh, you have your counsel?
Mr. Johnson. I do. Floyd Abrams.
Chairman Tauzin. I didn't hear that. Do you intend to
testify, counsel?
Mr. Abrams. No, I will not.
Chairman Tauzin. Then will you please raise your hand, sir.
[Witness sworn.]
Chairman Tauzin. Counsel, you are entitled to sit at the
table if you would like to.
Mr. Abrams. No, I am fine.
TESTIMONY OF TOM JOHNSON, CHAIRMAN AND CHIEF EXECUTIVE OFFICER,
CNN
Mr. Johnson. Mr. Chairman, members of the committee, I
appear before you today to outline the decisions CNN has
reached about the changes we will make in future election night
coverage.
At the outset, I would like to say that CNN is acutely
aware of our responsibilities to the American public and of the
responsibility of Congress to enact appropriate legislation
relating to the electoral process. At the same time, as you
have heard earlier today, there are sensitive first amendment
issues raised by any hearing relating to editorial decisions by
journalists; and I trust the committee will, as it has said,
bear those in mind.
Looking back on campaign 2000 coverage, I am very proud of
the hundreds of CNN journalists who devoted their efforts to
informing the American public about the issues of this last
election. However, CNN did make major mistakes, both in its
initial projection of Vice President Gore as the winner in
Florida and then prematurely projecting that then Governor
George Bush had won in Florida. As a result, I appointed a
totally independent panel to advise us on what went wrong, why
it happened, and what should be done to prevent a recurrence in
the future. You heard from that outside panel earlier today.
As a result of our full review and to ensure complete
reliability in the future, CNN has announced a number of
decisions last week. The first relates to CNN's future
connection with VNS. We will remain with VNS if and only if
significant changes are made. The errors that plagued election
night 2000 must never be repeated. Among the action steps: a
revision of VNS's projection system and statistical models.
These then will be reviewed by outside experts. Additional
research into methods for better estimating the increasing
number of absentee and early voters, as well as to better
analyze the nonresponse rates and possible statistical bias
which we have discussed in the exit polls themselves.
Beyond the efforts to improve VNS, CNN has decided that it
will have a second source for the data used to make projections
for the close races. CNN will fund a back-up sample key
precinct vote-reporting system in the States that are expected
to have the most competitive races. We will then cross-check
the information we received from this second source against the
VNS data, making any projections based on the data more
reliable. Also, CNN will insist that the Associated Press
tabulation system, which has been very reliable, be better
integrated into the VNS election night data collection system.
CNN also will limit our reliance on exit polls. We will only
use exit polls survey data to project a winner when the data
indicate clearly that one of the candidates has a large margin
at closing time.
Twenty-six States were called by CNN using VNS exit poll
data. All of those were correct calls. Despite that, CNN will
raise significantly criteria for future exit poll projections.
If a race cannot be called at poll closing, CNN will only
project a winner in that State using actual vote data from the
statewide vote tabulations and sample precincts.
One of the lessons learned from election night is that when
a race is extremely close, the reported vote might be
incorrect. Therefore, even if it is reported and all of the
outstanding ballots of the State are counted, CNN will not
project a winner if the balloting shows there is a less than 1
percent margin between the two candidates. These new policies
certainly will slow down our election night projections. Had
these standards been in place this past year, we would have
delayed at least 30 minutes in 10 States, and Florida never
would have been called early in the evening for Vice President
Gore or in the early morning hours for Governor Bush.
As you have heard from others today, CNN also has decided
they will no longer project the winner in any State until all
the polls are closed within that State.
As a result of our review CNN will also change our language
regarding projections. CNN anchors will avoid saying that a
State is too close to call if in fact we just don't have enough
data available to call that State, and CNN's anchors will
describe more specifically in each State the basis for our
projections, whether it's exit polls, sample precincts or the
reported votes. With these and the other proposed changes, I am
confident the errors of the past will not be repeated.
I do want to respond to one question that has been raised,
and I think it has been emphasized earlier today our reporting
about election night was not biased. CNN's selection of which
States to project winners at particular times in the evening
was not biased. And on the matter of legislation, and I say
this as much as a former Californian as a person today living
in Georgia, I strongly urge the Congress to adopt a nationwide
uniform poll closing act. If such an act were adopted, CNN
would not make any projections until all the polls are closed
nationwide.
To close, I assure this committee that CNN will go the last
mile to fix the problems which have been identified. As I have
told our staff, and I know that we all understand it, we would
rather be right than first.
Thank you, Mr. Chairman.
[The prepared statement of Tom Johnson follows:]
Prepared Statement of Tom Johnson, Chairman and CEO, CNN News Group
Mr. Chairman and Members of the Committee: I am Tom Johnson,
Chairman and CEO of the CNN News Group. I appear before you today to
outline the decisions CNN has reached about the changes it will make in
its future election night coverage.
Let me start with our election night coverage. I want you to know
that I am very proud of the hundreds of CNN journalists who devoted all
their efforts to informing the American public about the issues in the
last Presidential election, the electoral contest itself, the events on
election day, and the extraordinary 35-day period afterwards in which
the Florida vote was at issue. All of our efforts were dedicated to
ensure that the American public was better informed than ever before.
To a significant degree, I believe we did just that.
At the same time, however, CNN, as did the other networks, erred
both in its initial projection of Vice President Gore as the winner in
Florida and then in prematurely projecting that President George W.
Bush had prevailed in Florida. As a result, I appointed a totally
independent panel of experts to advise us on what had gone wrong at
CNN, why it had happened, and what should be done to prevent against a
recurrence in future elections. Three widely respected scholars
accepted CNN's request to work on this project: Jim Risser, two-time
Pulitzer Prize winner and former director of the Knight Fellowship
program at Stanford University; Joan Konner, former dean and current
professor of the Columbia Graduate School of Journalism; and Ben
Wattenberg, senior fellow at the American Enterprise Institute. Last
week, the panel completed its report and CNN released it publicly.
In addition to the panel's review, CNN, along with the other
networks, funded a separate report by the Research Triangle Institute
(RTI) of Raleigh, N.C. on the operations of the Voter News Service
(VNS), an organization created and jointly funded by ABC, CBS, CNN,
FOX, NBC and the AP. VNS provided, and all the networks used, Election
Day survey data for projections and analysis and also provided a
nationwide vote tabulation reporting system. As has become clear, some
of the data VNS generated was erroneous and an examination of VNS's
role was required and was initiated by the VNS board.
As a result of our review of events surrounding the erroneous
Florida calls, CNN has made a number of decisions about the future. The
first relates to CNN's future connection with VNS itself. CNN has
determined to remain with VNS if, and only if, significant changes are
made to ensure that the errors that plagued election night coverage in
2000 do not recur. The steps necessary to accomplish this include, but
are not limited to:
Implementation of the bulk of recommendations contained in the
RTI report regarding VNS.
Rewriting of VNS's projection system and statistical models
that will then be reviewed by outside experts.
Additional research into methods for better estimating the
increasing number of absentee and early voters, as well as to
better analyze non-response rates and statistical bias in exit
polls.
Initiation of a major upgrade and modernization of VNS's
technical capabilities and infrastructure as outlined in the
Battelle Practice Report, commissioned by VNS members in August
2000 and completed this month.
To accomplish these goals and to implement the bulk of the
recommendations made in the RTI Report, additional financial
contributions will be required of VNS members. CNN is prepared to pay
its fair share. Further, if it remains a member of VNS, CNN will urge
the addition of one or more outside respected academics, journalists or
research professionals to the board to provide independent perspective
regarding VNS operations.
CNN will, as well, appoint its Executive Vice President for News
Standards to take an active role in its involvement with VNS. Let me
reiterate, if VNS is not reformed, we will pull out and will support a
potential successor organization should VNS fail to meet CNN's
requirements.
Beyond the efforts to improve VNS, CNN has decided that it should
have a second source for the data used to make projections in the
closer races on election night. Therefore, CNN will fund a sample key
precinct vote-reporting system in the states expected to have the
closest races. This will ensure that our network can cross-check the
information it receives from its new second source against the VNS
data, making any projections based on that data more reliable. In
addition, CNN will insist that the Associated Press tabulation system
be better integrated into the VNS election night data collection
system.
In addition, CNN has decided to limit its reliance on exit polls.
CNN will only use exit poll survey data to project a winner when the
data indicate one of the candidates has a large margin at the time that
the polls close in that state.
If the race is too close to call, CNN will then only project a
winner in that state using actual vote data from the statewide vote
tabulations and key precincts.
If these standards had been in place on election night 2000, CNN's
projection of a winner would have been delayed by at least 30 minutes
in 10 states and Florida would never have been called early in the
evening for Vice President Gore. As for calls made at poll closing in
the 2000 election, 26 states were called by CNN, using VNS exit poll
data, when the polls closed in those states. All were correct calls,
and none was close. Despite that, CNN will raise significantly the
criteria for these exit poll projections above what was used on
election night.
CNN also has decided not to project a winner in a state, even if it
is reported that all the outstanding ballots have been accounted for,
if the balloting shows that there is less than a 1% margin between the
candidates. This would have prevented the too-early call of Florida for
President Bush.
CNN also has decided that it will no longer project the winner in
any state until all the polls are closed within that state.
We believe it is important that our viewers know more about how we
gather information and how we deal with the information we have
gathered. Accordingly, CNN will draw back the curtain for our viewers
on the exit poll and projection process. In the days leading up to the
election, CNN will produce a number of reports on its election night
reporting. We will provide other ``behind the scenes'' reporting to
show viewers how the projection process works. In addition, on Election
Day, CNN will assign a number of correspondents to report on how the
key precinct and exit poll workers do their jobs.
In addition, CNN will assign its own reporters to the control room,
the CNN Decision Desk, and to VNS (or its successor) during the
broadcast to provide ``behind the scenes'' reporting that will let
viewers see how the projection process works.
As a result of our review of our coverage on election night, we
have decided that CNN will change its language and graphics on election
night regarding projections. Until all the votes have been counted in
each state, CNN will no longer call anyone the winner of a state. No
longer will CNN anchors say, ``CNN calls Al Gore the winner in
Michigan,'' but will instead try to explain better what is the basis of
the projection. For example, anchors will say, ``CNN projects that,
based primarily on exit poll estimates, Al Gore will win Michigan,'' or
will say, ``Based on the results of voting from key precincts and an
evaluation of the returns so far from around the state, CNN estimates
George Bush will win the state of Virginia.''
In addition, CNN will be careful with its language regarding the
reasons for why a call is not being made. It will instruct its anchors
to be specific. ``It's too close to call in Georgia,'' has a different
meaning from ``We don't have enough information yet in Georgia to make
a call there.'' The former means it is a close race. The latter simply
means we need more information to make any type of characterization.
CNN's onscreen graphics will reflect better what it knows and
precisely what it means to say. On election night, CNN's full-screen
graphics included the words ``CNN Estimate,'' but the words were not as
prominent as they should have been. CNN's new policy will position the
words ``CNN Estimate'' as the title on the full-screen graphic.
In very close races, CNN's reporting will inform viewers at the
earliest appropriate time regarding a state's mandatory recount
provisions.
With these changes in effect, we believe that we can address the
future confident that the errors of the past will not recur. We will do
all in our power to ensure that they do not.
I want to add a further observation of my own about the past and
the future. We are, as I have said, proud of the CNN journalists who
worked so hard on election night to inform our public. And I want to
respond to one question that has been raised about our coverage in the
clearest, most unambiguous way. CNN's reporting about election night
2000 was not biased. CNN's journalists were not biased, CNN's coverage
was not biased, CNN's selection of which states to project winners, at
particular times in the evening, was not biased. I strongly believe
that the same is true, as well, of VNS.
Finally, I do want to address the matter of legislation. I strongly
urge the Congress to adopt a nationwide Uniform Poll Closing Act. If
such an act were adopted, CNN would not make any projections until all
the polls were closed nationwide, just as it will not make any
projections in any state until all the votes in that state are cast. To
the extent that the committee is concerned that the announcement of who
the likely, or even certain, winner is in one state could affect voting
patterns in another, the adoption of legislation providing for a single
time at which all voting ceases in the country would assure that even
the potential of any such impact was avoided.
Because the 2000 election was so close, it exposed problems with
the mechanics of voting, that have been there for some time, but were
not noticed. In the same way, it exposed problems with the operation of
VNS. I say this not as an excuse, but to put the situation in
perspective. As with the steps that are being taken with election
reform, we are going to take every measure to make sure our own
problems are corrected.
Chairman Tauzin. Thank you very much, sir. Our next
panelist will be Roger Ailes, Chairman and Chief Executive
Office of Fox News, New York, New York. Mr. Ailes, if you will
help me through this process. Mr. Ailes, you are aware that our
practice is to take testimony under oath when we are doing an
investigative hearing. Do you have any objection to testifying
under oath?
Mr. Ailes. It wouldn't make any difference. I plan to tell
the truth either way.
Chairman Tauzin. Very good. The Chair then advises you that
under the rules of the House and rules of the committee you're
entitled to be advised by counsel. Do you desire to be advised
by counsel here?
Mr. Ailes. My counsel is here, Diane Brandy, in case I need
some documents or whatever. She does not plan to testify.
Chairman Tauzin. In that case will you please raise your
right hand?
[Witness sworn.]
Chairman Tauzin. You are now under oath. You may give your
testimony.
TESTIMONY OF ROGER AILES, CHAIRMAN AND CHIEF EXECUTIVE OFFICER,
FOX NEWS
Mr. Ailes. Thank you, Mr. Chairman, Mr. Dingell, who I
guess is not here, members of the committee. I appreciate the
opportunity to tell what happened on election night 2000. This
was Fox News' first cycle with VNS, 1998 and 2000, and our
first Presidential election. In a sense both the networks and
Congress have similar tasks. We've had to come up with
solutions for our reporting on election night and Congress is
considering election reform and campaign finance reform. These
are very complicated issues and take years to sort out, but in
just a few weeks we've come up with what we collectively
believe are solutions to the problems that we had on election
night 2000.
Like you, I am concerned about the mistakes we made. Our
viewers depend on us for reliable and accurate information,
especially during important events like a Presidential
election. If we lose their confidence we have lost everything.
For this reason at my direction Fox News performed an in-depth
review of events that transpired on election night. As everyone
knows, Voter News Service, a consortium with a good track
record, gave out bad numbers that night. In the closest race in
history the wheels apparently came off a rattle trap computer
system which we relied on and paid millions for.
As Fox relied on those numbers, we gave our audience bad
information. Our lengthy and critical self-examination shows
that we let our viewers down. I apologize for making those bad
projections that night. It will not happen again. We were the
first network to announce that we will never call a State again
until all the polls are closed in that State. We are working on
internal safeguards, including placing more of our own
personnel in key precincts, in order to gather information and
report results. If we stay with VNS, we will spend more money
to help fix the computer models and help report the elections.
Fox News favors Mr. Markey, Mr. Dingell, Chairman Tauzin's
legislation for universal closing times across the Nation.
Now I would like to say a few words about this hearing.
From the beginning Fox News cooperated fully with this
committee to find solutions. While I honor and respect this
committee's role in searching for legislative solutions, Mr.
Chairman, I am deeply disappointed that this is being handled
as an investigative and not a legislative, fact finding matter.
I am further disappointed that this committee views its role as
adversarial, requiring us to take an oath as if we have
something to hide. We do not. With or without a swearing in
photo op, we will hide nothing. I know all of these gentlemen
at this table personally. I have worked with some. We are all
competitors and in some cases we don't agree on issues and in
other cases we are not even that fond of each other. However,
we all understand the importance of our respective journalistic
enterprises and journalistic integrity. And they as well as I
will tell the truth whether we are under oath or not.
Everything our organizations did on election night was done
under the protection of the first amendment, and that may
become more relevant as these discussions and questions
continue.
A final personal note: There seems to be a bipartisan
agreement that we should slow down our competitive spirit and
thereby slow down the election, and I can almost assure you
that if you put us through another day like this the next
results may not be known for 3 weeks.
Thank you, Mr. Chairman.
[The prepared statement of Roger Ailes follows:]
Prepared Statement of Roger Ailes, Chairman and CEO, Fox News Network
Let me begin by stating that Fox News, along with all the other
television networks, made errors on election night which cannot be
repeated, the biggest of which occurred in Florida.
Fox News acknowledges here that it failed the American public on
Election Night and takes full responsibility for this failure.
These errors have led to much self-examination of the processes we
used on election night, how the Voter News Service operated on election
night, and our membership in the Voter News Service.
Through our self-examination and investigation we have determined
that there was no intentional political favoritism in play on election
night on the part of Fox News.
In hindsight we made a significant error in relying on VNS data
alone, although that was the only data available. Obviously, it would
have been better to have at least one other source of data, but up
until now economic considerations have made this unfeasible. We look at
VNS in much the same way the networks combine resources for pool
cameras, the Associated Press, etc.
As you may know, the Fox News Channel launched on October 7, 1996.
From the moment we launched we intended to compete with the big
established television news networks. In order to cover elections in a
competitive manner, we believed we would have to join VNS. But it was
not an easy decision for us. First and foremost, membership in VNS was
(and is) very expensive, especially for what was, at the time, a
fledgling television network. But, after many internal discussions of
both editorial and financial natures, we decided to join. I understood
VNS had a good, solid record of calling races until the 2000 elections.
For example, 99% of the calls which VNS made over the last two election
cycles have been accurate; 100% of VNS' calls in 1998 were accurate.
Now, however, we feel the purpose, intent, processes and models of
VNS must be carefully examined in a formal manner and we are willing to
spend more money as a VNS Member to make this examination happen.
Let me assure you that Fox News operates in the interest of the
public and attempts at all times to conduct itself with that fact in
mind.
Since election night, the issue of voter suppression has been
written about and discussed.
Would it have made any difference in voter turnout if the
television networks waited until all polls in the state of Florida, and
in every other state for that matter, had closed before declaring a
winner. When Fox News called Florida for Al Gore at 7:52 pm, there were
eight minutes remaining for citizens in the Florida panhandle to vote.
Well, I don't know the answer to that question, but to remove all
doubt it is a simple enough remedy for a television network to wait
until all polls in a given state have closed before declaring a winner
in that state.
Shortly after election night, Fox News became the first network to
announce that going forward it would not call a winner in a given state
until all polls in that state were closed.
I would now like to discuss briefly Fox News' relationship with VNS
and our Decision Desk Team.
John Ellis headed our Decision Desk team. He was joined by John
Gorman, Arnon Mishkin and Cynthia Talkov. All four members of the
Decision Desk had to agree on a call before it was recommended to John
Moody, our Vice President of News Editorial. Mr. Moody then made the
final decision regarding whether or not to make a call on the air on
election night. The Decision Desk team would unanimously recommend a
call and Mr. Moody would either accept it, in which case it got to air,
or he would question it in which case the Decision Desk team would walk
him through the numbers until he was comfortable making the call on the
air. Mr. Moody was also responsible for the retractions we made on the
air.
I would like to say a few words about John Ellis because I am sure
you are all familiar with him and his family connections. Mr. Ellis is
the first cousin of President George W. Bush and Governor Jeb Bush. We
at Fox News do not discriminate against people because of their family
connections. I am more than happy to give you examples of offspring of
famous politicians who are employed at Fox News.
As for Mr. Ellis, he has almost 23 years of experience in calling
elections. I won't go through his entire resume with you, but I will
highlight the fact that he worked in NBC's election unit for over 10
years, including when George Herbert Walker Bush ran for President in
1988. I have personally known Mr. Ellis for almost 20 years. He is a
consummate professional. Much ado has been made about a column Mr.
Ellis wrote for The Boston Globe in July 1999 where he stated in effect
that his loyalty to then Governor George W Bush would prevent him from
writing further columns about politics.
I am aware that Mr. Ellis was speaking to then Governor George W
Bush and Jeb Bush on election night. Obviously, through his family
connections, Mr. Ellis has very good sources. I do not see this as a
fault or shortcoming of Mr. Ellis. Quite the contrary, I see this as a
good journalist talking to his very high level sources on election
night. Our investigation of election night 2000 found not one shred of
evidence that Mr. Ellis revealed information to either or both of the
Bush brothers which he should not have, or that he acted improperly or
broke any rules or policies of either Fox News or VNS. By the way, Mr.
Gorman and Mr. Mishkin were speaking to high level Democratic sources
throughout the evening.
To be clear, there was no information which John Ellis could have
given to anyone nor was there any unilateral decision which Mr. Ellis
could have made which would have affected the outcome of the election.
I understand that you may ask me specific questions about the
decision screens which were used that night and about other data. I
depended entirely on our Decision Desk. I trusted them (and still trust
them) to do a professional job and I believe they did the best job that
night they possibly could have.
I elected not to study these decision screens and other data solely
in preparation for this hearing since I have never looked at them
before in my life. So any questions you ask me about data will be a
blind alley.
Conclusion:
We must all question what purpose early calls serve, apart from
bragging rights. We must especially question the purpose of early calls
since we all rely on one source for our information.
You know I have toyed with the idea of not covering the next
election until the day after when all the votes have been tallied and
the winners are certain beyond all doubt. I have even thrown this idea
out to some members of my senior staff. They looked at me skeptically,
as if they weren't sure whether or not I was kidding. I am, after all,
the head of a news organization and this idea seems completely contrary
to all I should be about.
But in my heart I do believe that democracy was harmed by my
network and others on November 7, 2000. I do believe that the great
profession of journalism took many steps backward.
Here is what Fox News recommends going forward.
No announcing of winners in any state until all polls in that
state have closed.
Fox News favors uniform poll opening and closing times across
the entire country, perhaps opening at 11 am Eastern time and
closing at 11 pm Eastern time (i.e. opening at 8 am Pacific
time and closing at 8 pm Pacific time). Perhaps voting takes
place on a Saturday or perhaps a new federal holiday is
declared for voting.
Fox News intends to produce several news packages around the
time of the next elections to explain various elements of the
election process, including how the Electoral College was
established, what its purpose is, explaining how
``projections'' are made, explaining any built-in biases in the
systems that are used to interpret data and call elections.
Going forward, Fox News will inform its viewers of the margin
of error in each call it makes, however small.
Fox News is considering more internal safeguards for the next
election including placing more of its own personnel in key
precincts in key states to gather information and report on
results.
Fox News recognizes that having all networks relying on one
source of data, VNS, is problematic.
Accordingly, the purpose and processes of VNS must be
carefully reexamined in a formal manner. The models must be
rebuilt. The systems must be modernized. Fox News is willing to
spend more money as a Member of VNS to go through this
reexamination process to ensure that we get it right the next
time.
Thank you for the opportunity to address you.
Chairman Tauzin. Thank you, Mr. Ailes. I would like to
report to the committee that we held quite a number of
investigative hearings during the last session of Congress. We
had 128 government and 170 nongovernment witness, all of whom
were traditionally placed under oath. It has no other meaning
other than the fact that investigative hearings are taken under
oath.
Our final witness on this panel is Mr. Andrew Lack,
President of NBC News, New York, New York. Mr. Lack, if you
will kindly go through the procedure with me.
You are aware the committee is holding the investigative
hearing and when in doing so the practice is we take testimony
under oath. Do you have an objection to testifying under oath?
Mr. Lack. No, sir.
Chairman Tauzin. The Chair advises you that under the rules
of the House and the rules of the committee you are entitled to
be advised by counsel. Do you desire to be advised by counsel
during your testimony today?
Mr. Lack. No, sir.
Chairman Tauzin. Then if you will kindly raise your hand,
sir.
[Witness sworn.]
Chairman Tauzin. You are properly sworn and you may give
your testimony, please, sir.
TESTIMONY OF ANDREW LACK, PRESIDENT, NBC NEWS
Mr. Lack. Mr. Chairman and members of committee, thank you.
I would expect that you would like to get to some questions and
answers pretty quickly, so I'm not going to use all of the time
that's allotted to me for my initial remarks. You already have
my prepared statements and you know that that summarizes the
findings of our election night broadcast and details the steps
that we are taking and the steps that we have taken to avoid
the errors that occurred that night. Make no mistake about it,
we are embarrassed by those errors, and you have heard chapter
and verse today from some of my distinguished competitors. I
join them in ensuring that we are absolutely intent on avoiding
them and making sure they don't happen again.
But I would, if I may, like to spend just a few moments to
take a different cut at this and call your attention to
something that is far more embarrassing to me and more
important to me as a reporter and as the President of NBC News.
It is something which was exposed to the country in the days
and weeks following November 7, and it's something that I worry
may get just a little bit lost in the context of these
hearings. Where was our reporting before November 7 about the
potential impact of ineffective voting machines or confusing
ballots or inadequately staffed polling sites? What was the
potential impact of a system that might in fact be protecting
felons who vote? We knew this was going to be a close election,
and I just don't understand how it quite turned out that we
didn't know much about automatic recount standards which are,
as you know, somewhat arbitrary and incredibly arcane, at least
for me.
We know now that if you are registered to vote it doesn't
always mean you will be permitted to. We know if you're in the
military and you mail in an absentee ballot it doesn't
necessarily mean it will be counted. We know if you're poor in
this country it means it will likely be that you will have a
little bit more difficulty voting than if you are rich. It
occurs to me that a good question for us would be for the price
of a new Federal highway could we have gotten this whole system
fixed?
Millions of votes are thrown out in election after election
in this country. Now that is a story. Now there's a screw-up.
We didn't do nearly enough digging, it seems to me, into those
facts. And if we'd asked some of these questions before the
election and had some answers, we might have been in a whole
lot better shape on election night than we were. We booted it
in more ways than one.
So I just ask respectfully of this committee that as this
hearing winds down, that you extend your focus beyond just the
problems that the networks experience on election night and
look at the problems that the voters experience on Election
Day. As a journalist I wish I had.
And I yield, Mr. Chairman, the balance of my time.
[The prepared statement of Andrew Lack follows:]
Prepared Statement of Andrew Lack, President, NBC News
Thank you for the opportunity to speak here today.
I know that many of you have expressed concern about the way NBC
and the other networks reported the presidential election results on
the evening of November 7th. Let me assure you that no one is more
concerned about what happened on election night than we are at NBC
News. That is why we have moved aggressively to determine what went
wrong and to begin to put in place the checks and balances needed to
ensure that our future election night reporting is in keeping with our
proud tradition.
On election night, NBC News had to take back a call made in an
important race. And we had to do it not once, but twice. We reported at
7:49 p.m. Eastern time that Al Gore would win the state of Florida, and
then had to withdraw that call at 10:16. Next, we reported, at 2:17 in
the morning, that the state of Florida and the Presidency would go to
George W. Bush--a call we had to withdraw at 4:00 a.m. because the vote
count, in fact, was still very much in dispute. At that point, as Tom
Brokaw said, we didn't just have egg on our face, we had omelet all
over our suits.
The next morning, we began a top-to-bottom review of NBC's own
election night procedures and insisted on an independent assessment of
Voter News Service as well. In the process, we have learned a lot about
how we cover elections, about the quality of the information that is
supplied to us on election night, and about some of the steps that we
need to take to improve our election coverage.
I'd like to outline some of those conclusions for the Committee,
but there are two important issues that I'd like to address first: one,
the charge that political or partisan considerations are a factor in
our election reporting, and two, the allegation that our election
projections cause voters to stay home from the polls.
I have heard the allegation of political bias, and as I understand
it, the charge is that our projections for certain states were
delayed--that some of the so-called ``Gore states'' were called more
quickly than some of the ``Bush states''--that, somehow, we favored the
Democratic candidate.
Let me respond by stating, unequivocally, that I am absolutely
certain that political bias played no role in NBC's election night
reporting. We have one mission when it comes to reporting results on
election night: to do so as accurately and as quickly as possible. We
believe our election night viewers want to know who is winning and who
will win that night, and they want to know as soon as we have the
ability to tell them. We aim to provide that information--with no
agenda other than timely and responsible reporting. We have reviewed
each of the projections we made on election night; we have confirmed
that each projection was made just as soon as our highly experienced
election experts determined, in good faith and based on the data in
front of them, that a call was appropriate.
We are mindful, of course, of the concern that projecting election
results while polls are still open in some places may discourage some
voters from voting. We have decided that, for states in which polls
close at different times (such as Florida), we will no longer project
until after the last scheduled poll closing time in that state. We
think this change will help avoid confusion among potential voters
about what time the polls in a state actually close.
As for the charge that voters in the western time zones stay home
on election day because of televised coverage of results in the east:
the Committee no doubt is aware that no reliable study has ever
concluded that voters are actually so affected by our broadcasts. We
have looked closely at this issue, however, and we continue to keep an
eye on all of the research and data available. And, although, as a news
organization, we tend not to take positions on legislative policy
initiatives, we do support, as we have stated in the past, a national
uniform poll-closing time.
It's worth pointing out as well that, during our election-night
broadcasts, our anchors frequently remind our viewers that polls may
still be open in their communities and that they may still have the
opportunity to vote. On November 7th, Tom Brokaw made statements to
this effect a number of times during the broadcast.
Of course, we did make mistakes on election night. Our review of
NBC's performance that night is ongoing, but we have already learned
and concluded that a number of fixes have to be put into place. I make
these observations having accepted the recommendations of a review team
that included two NBC executives and the dean of the Graduate School of
Journalism of Columbia University:
* VNS has to be retooled or replaced. We now have the benefit of an
extensive report by Research Triangle Institute, an independent
non-profit firm, and RTI has confirmed what we immediately
suspected: that the data our experts relied upon on election
night was flawed in a number of ways. We now know that, while
the fundamentals of the VNS methodologies are sound, VNS
performance suffered when it came to implementation and quality
control in a contest as close as the Florida presidential race.
In addition, the VNS computer system, a critical part of the
process, appears not to have been up to the complexity of the
job.
As for whether we should continue to use exit polls and early
vote returns from key precincts to project election results, we
continue to believe that these can be valuable sources of
information for our viewers. It is obvious, however, that
polling methods need to be further adapted to address recent
trends and changes in voting patterns, including the increase
in absentee voting. For instance, we may need to increase
election day exit polling and pre-election telephone polling
for races expected to be close. We have stated already that we
are prepared to increase our budget to make such improvements.
* We also need to make certain that we use supplementary sources of
data as a way to verify the accuracy of the information we get.
We learned from election night 2000 that even the tabulated
vote as reported by election officials can be inaccurate. We
are hopeful that, by the time of the next elections, official
data from states and counties will be more accurate as well as
more easily and quickly available, on the Internet or through
direct computer feeds.
* And, perhaps most important, we need to better explain to our viewers
the type of data that we have, where it comes from and what it
means. We must differentiate for our audience the projections
that we make based on exit polling and those that we make based
on actual tabulated votes, and we need to define our terms,
like ``projection'' and ``too close to call''. Our nomenclature
and our graphics must be clear, understandable and precise.
Finally, I'd like to address the ultimate issue of whether we
should stop making election night projections at all. We think not.
Projections answer the questions our viewers want to know: how have
people voted, and who will win a given race. And, of course, holding
back this information would hardly stop the flow of unreliable
information coming from the Internet and elsewhere, raising the
potential for far greater confusion.
In summary, we at NBC News have already learned a great deal about
what went wrong on election night and what we need to do in the future
to ensure that our viewers have the most accurate, reliable and
understandable election night information.
It is no consolation to us that the closeness of the election
revealed far more serious problems in the American election system than
the ones we experienced, that many citizens had trouble voting or
having their votes count. It will be up to the members of Congress and
other elected officials to spend whatever time and effort are required
to fix those problems. In the meantime, we are committed to fixing our
own.
Chairman Tauzin. You do that very well. The Chair
recognizes himself and members in order. Let me first again
thank you all for voluntarily appearing and for participating
under the rules that are set for these types of hearing. And
second, I can assure you that as we go through this process
I've asked all the members on our side and I know that Mr.
Dingell has similarly requested all members to respect the
boundaries that we've discussed here today.
Let me first ask you a bit about your thoughts on exit
polling and the value of exit polling in this process. The
testimony we just heard and a lot of our own research indicates
that exit polling has not gotten much better and it may have
gotten worse over the years in being a good determinator in
which to make a projection, that, as you heard witnesses say,
that the number of nonresponses drives up the margins of error,
and we've seen some statistics indicating as much as 16 percent
errors in those election projections because of that and other
problems that you identified, Mr. Savaglio.
Do you disagree that exit polling is getting worse not
better? Anyone. Anyone thinks it's getting better?
Mr. Lack. Well, I don't know that I know the scientific
answer to that from a statistician's point of view, but my
sense is that in fact it is getting better. From our
conversations at our decision desk and with the people that we
speak to at VNS, one of the problems that occurred on election
night is they didn't have enough exit polls arguably to make
some of the projections or to use it as one of the tools that
they were using as part of the projection process.
Chairman Tauzin. Does anyone else think--Mr. Ailes, do you
think it's getting better?
Mr. Ailes. I think--was there some question of bias toward
the Democrats in the exit polls? Because that's a question that
I heard earlier today. There seemed to be some statistical bias
in that direction.
Chairman Tauzin. No, I'm just asking do you think it's a
better reliable indicator of where voters had voted than it was
5 or 10 years ago.
Mr. Ailes. It is very hard to tell. Some pollsters will
tell you they're getting more accurate. Most pollsters will
tell you even in polling, let alone exit polls, that fewer and
fewer people are willing to offer information about themselves
personally and privately. But I don't know the empirical data.
I do know that when Republicans come out of polls and you ask
them a question they tend to think it's none of your business
and Democrats want to share their feelings. So you may get some
bias there that is inadvertent, just because it's a cultural
thing and unless you send the Republicans to sensitivity
training you're not going to get them to do that.
Chairman Tauzin. Mr. Ailes, that was actually one of the
conclusions of the analysis of VNS that was made, that indeed
it may be a cultural difference between members of different
parties.
Mr. Westin.
Mr. Westin. Yes, our statisticians and our analysts have
taken a look at this and you go back over the last three
elections, including the 2000 election, and you look at the
overall statistical skew of the exit polls nationwide, you will
find that whereas in 1992 it was roughly 5 percent, in 1996 it
was just under 2.5 percent. I think it was 2.3 and in 2000 it
was 1.4 percent average overall. So in terms of statistical
skew. Now, as you know, Mr. Chairman, the difficulty with
statistical skew, while it is true on average over time it
tends to overpredict Democrat candidates more than Republican,
in any given race it can favor a Republican or a Democrat. And
even in 2000, if you go back and look at the elections, there
were some larger States and key States where there was a
statistical skew in favor of then Governor George W. Bush. For
example, Pennsylvania was one we delayed 80 minutes, and I
believe that was in large part because in fact our exit polling
was overpredicting George W. Bush.
Chairman Tauzin. Yes, it did go both ways.
Mr. Westin. It did go both ways.
Chairman Tauzin. Let me ask you, do you as a group believe
that the original opinion of the networks, you know, we've got
the opinions going back to 1980, we were pretty hard on exit
polls. Walter Cronkite was tough on them. Robert Whistler, the
Executive VP of CNN back then, said this is information of
their own manufacture and posed a significant risk that the
actions in producing these polls and reporting on them might
influence the electoral process themselves. How do you answer
the criticism that they really are produced by the news
networks themselves? They are literally not news you find but
news you create.
Mr. Westin. Right, two or three points. First of all, at
ABC News, and I know this is true of my colleagues, we would
not use exit polls if we didn't think it increased the accuracy
of our reporting. It doesn't mean they're always accurate in
and of themselves, but if that additional input into our
consideration doesn't increase the accuracy we wouldn't be
using them as a practical matter. That's why we use them. And I
think if we took exit polling out of it, our reporting would
tend to be less accurate than it is now. But to your specific
question, there is an awful lot of reporting that ABC News does
and that all of my colleagues do on any given day in this
country that has to do with what is thought to be likely to
happen, whether it's conflict in the Middle East or what Alan
Greenspan is likely to do or even what Congress is thought to
be doing. We would be doing something bad for this country if
we said from now on news organizations don't report anything
until after it's over and surprise everyone.
Chairman Tauzin. Let me understand the agreements that I
think most of you are trying to make with the American public
today in terms of using those polls. Most of you speak about
not using the exit poll until all the polls have closed in a
given State. By that do you mean the hours that polls have
closed or when the voters in line finish voting. Anyone?
Mr. Boccardi. I think what we are saying here today is that
the hour when the polls close would be the problem.
Chairman Tauzin. I think that's what you all meant, as I
read your statements.
Mr. Boccardi. The thing that you are raising would be
almost impossible.
Chairman Tauzin. I don't know how you would administer it,
and it is one you could set a time, you could say 15 minutes
after, 20 minutes after. But do you understand there is still a
lot of concern that there are a lot of people standing in line
even at the hour the poll closes. In fact, we had testimony
that there were 2 hours of voting in some places going on even
after the polls close. So even an agreement that says we won't
report an exit poll until the hour the poll closes may leave a
lot to be desired, although it goes a long way. I'll give you
that.
Mr. Boccardi. If they're on line with their radios
listening for----
Chairman Tauzin. Or their cell phones. Gosh, we have all
kinds of communications today. But understand that has been
raised to us as one of the concerns regarding how we resolve
this even if we go to a uniform poll closing time.
I wanted to do one last thing with the very last minute
I've got. On the VNS screens, when you post the information to
the networks on those screens, we have to tabulate the screen
tabulations for 2:10 in the morning. The screen that was
presented to all the networks when the networks made the call
that the election was over, George Bush won, those screens
indicate there were only, according to VNS estimates, only
179,000 voters left in Florida to vote. Was that a correct
number, Mr. Savaglio?
Mr. Savaglio. No, that was a pretty substantial
underestimation.
Chairman Tauzin. About two to one. They had almost twice
the number left to go. If you could take the mike.
Mr. Savaglio. Thank you. That was----
Chairman Tauzin. Again, so we can make sure we get the
understanding correctly. The screen that presented the
information on Florida, the VNS screen, indicated that you had
96 percent of the vote in, 3 percent was out. 179,000 votes was
your estimate of votes out. That proved to be fairly
inaccurate, right?
Mr. Savaglio. Well, it was inaccurate, correct. It was a
considerable amount, 359,000 votes.
Chairman Tauzin. Almost twice as much were really out. The
screen also says that for Al Gore to have a chance to carry
Florida at that moment on that screen that he needed to get 63
percent of vote.
Mr. Savaglio. Right.
Chairman Tauzin. Was that information forwarded to all of
the networks at that hour of the night?
Mr. Savaglio. Well, that was on their screens and I
appreciate that you've asked me the one question that you don't
have to be a statistician to answer because I'm not. The
outstanding vote calculation is a very simple, call it crude if
you want, calculation. It simply states the number of precincts
that have reported out of the total number that are to report.
It takes and divides that and assumes all precincts are the
same size.
Chairman Tauzin. But they're not.
Mr. Savaglio. But they're not. And so it's simply a ratio.
If 50 percent of the precincts are in and there's 1,000 votes,
it assumes the next 50 percent are going to be another 1,000
votes. And the reason the calculation is put in that way is to
give the information that's available. It's not possible to--in
most cases to get the particular precincts or at least it's not
possible in a reporting fashion to put into our system the
specific precincts and their size and the number of votes from
each one as to come up with a more specific number.
Chairman Tauzin. But the bottom line is you gave the
networks that night on the VNS screen some relatively
inaccurate information, right?
Mr. Savaglio. Yeah, there's no question about that.
Chairman Tauzin. Thank you. Mr. Waxman.
Mr. Waxman. Thank you very much, Mr. Chairman. There has
been criticism, and I mentioned it earlier in my opening
statement, that John Ellis, who was President Bush's cousin and
who had talked with Mr. Bush, Governor Bush, now President
Bush, throughout the election night was responsible for Fox's
decision to be the first network to declare George Bush the
winner of Florida and the presidency.
Mr. Lack, I'd like to know if you would walk me through the
process at NBC on election night. Who at NBC had the
responsibilities assigned to John Ellis at Fox?
Mr. Lack. A gentleman, Dr. Sheldon Gliser, who is director
of our election desk.
Mr. Waxman. There is an allegation making the rounds that
Jack Welch actually intervened in NBC's decision to call the
election for George Bush. I don't know if you've heard that
rumor before. I would like to lay it out there and have you
comment on it.
Mr. Lack. I have heard the rumor and it's untrue.
Mr. Waxman. Well, I would hope the allegation is untrue. If
it were, it would be absolutely inappropriate. But I've been
told that Mr. Welch's actions were observed by others and in
fact were even captured on tape, filmed by NBC's advertising
and promotions department. It's difficult for me to believe
it's true, but it seems there is a simple way to either verify
or debunk this allegation. I would like to have you, if you
would assure us that we would get that tape and, Mr. Chairman,
I would like to ask you if you would to make sure that we have
any subpoena that might be necessary so that if there is such a
tape that we have it available to us.
Chairman Tauzin. Pursuant to our rules, we will take that
under advisement.
Mr. Waxman. Mr. Lack.
Mr. Lack. You're certainly welcome to the tape. I know that
advertising and promotion was around there. I don't know if
there is a tape for you to look at. I was aware that Mr. Welch
was there. I observed him. He was in the building to attend a
political party, network party, and he was invited down to
observe on a very historic night and a very exciting election
night how we were doing and what we were doing, and that's
precisely the manner in which he was there. I think it's
unfortunate that some rumors would get started that because he
observed our election night process at that point that that
would somehow like in a Rashoman-like tale turn out to be that
he intervened in the election process, which is untrue and
rather foolish, but that's rumor.
Mr. Waxman. And you yourself were there at the time?
Mr. Lack. Yes, I saw him. I can state categorically that
it's just a dopey rumor, truly dopey.
Mr. Waxman. Thank you very much. Thank you, Mr. Chairman.
Chairman Tauzin. Very good. Mr. Stearns I believe is next.
He's not here. Mr. Burr.
Mr. Burr. Mr. Chairman, thank you. And since I was not here
when this panel was seated let me take this opportunity to
welcome all of you. I am not sure if the chairman asked but I
would like to get a response from each of you as to whether
your representative medias plan to use exit polling in the
future.
Mr. Lack. I'm sorry, could you ask that again?
Mr. Burr. If in fact your company plans to use exit polling
in the future.
Mr. Lack. We do.
Mr. Burr. You do.
Mr. Ailes. We will also do it. It's under review what we
will tell the audience about exit polling. We believe there was
not enough disclosure and definition to some of the terms used,
light projection and exit polling and so on, and so the answer
is we will use the information as long as we believe it is not
misleading to the audience in some way.
Mr. Johnson. CNN will use exits polls provided we are
convinced of the reliability of the exit poll process, the exit
poll methodology.
Mr. Burr. Could I stop you there and just ask you what
would it take to convince you of the accuracy of the exit
polling given that the person who is being polled has the
ability to tell you the truth or an untruth?
Mr. Johnson. We would need to determine statistically, if
possible, to what extent we were getting misrepresentation. I
do not believe there is a good answer to an outright lie.
Mr. Burr. There is certainly not a perfect answer, but you
raised exactly the point I wanted to get at, that at this point
in this process the statistical accuracy was important. I would
like for the rest to answer the question and then I would like
to go back to the importance of statistical accuracy.
Mr. Heyward. Yes, Congressman, we intend to use them as
well judiciously in a larger context. I think it's important to
remember that exit polls are just one of many tools even in
making a projection. Every single projection that was made
solely of the basis of exit polls was correct. I think there's
a danger of demonizing them a little bit. But having said that,
they have to be improved. I think that the statistical sample
and the population habits and voting habits have changed. We
have to try to adjust for that. But they are one of many tools
even to go into a decision but, yes, we intend to continue
using them and we'll do our best to get them judiciously
obviously. The penalty for error is very severe.
Mr. Westin. ABC News will continue to use exit polls as
long as we believe it will increase the accuracy and timeliness
of our reporting. At this point I believe, particularly with
the improvements that we've specified and have been discussed,
I believe that it increases the accuracy and timeliness of the
reporting, which is not to say they're perfect. I haven't found
a perfect system yet.
Mr. Burr. Nor have we.
Mr. Boccardi. We will continue to use them in the same way
we have used them before. We will be looking very carefully at
how they are done at VNS and what needs to be done to make them
even more reliable than they have been. I repeat a point that I
said at the outset in my opening statement, and I don't think
you were here, we made one mistake election night and that was
on the early call for Gore in Florida. And exit polls were
integral to what we did that night.
Mr. Burr. Let me go, if I could, to the question of the
accuracy of the statistical data, the statistical model. How
many of the companies expressed concern about the model that
VNS was using? I'll either let the companies address that----
Mr. Westin. Prior?
Mr. Burr. Before election night.
Mr. Westin. There's ongoing discussion through the board
members at VNS about statistical models and they're changed
fairly regularly in a somewhat minor regard. As I said earlier,
to some extent we were the victim of our own success, if you
will forgive me. This had been a system that had allowed us to
project many, many elections over a long period of time
accurately and it led to hubris on our part. I will only speak
for ABC News in that regard. But we came to believe that we
really could do this with a very, very high level of accuracy,
and it also let us as we went on the air overstate the
certainty with which it was done. I think that the experience
of November will discourage some of the hubris.
Mr. Burr. If I could, let me go to the comment that was
said earlier of the caveat to your statement by using exit
polling, which was as long as we have the confidence and the
statistical model of its accuracy, and given that I think most
of you would agree with that caveat, were you assured prior to
this election that you thought your news entities thought that
the statistical model was sufficient?
Mr. Heyward. I think we had a lot of confidence in our
models. As David says, obviously we've taken a very hard look
and we will continue to do so to make sure these mistakes are
never repeated. We had a collision between a system that worked
very, very well for many years and over a couple of thousand of
races with an extraordinary election that was so close that it
took the Nation 5 weeks to figure out what happened and the
Supreme Court to intervene, and that certainly exposed flaws we
hadn't seen before. But we certainly went in with a great deal
of confidence, yes, and to the degree that we have to readjust
our perception I think we have and will.
Mr. Burr. Is it safe to say that nobody at the table
questioned the statistical model prior to the election?
Mr. Heyward. Except for the ongoing process of calibration
and evaluation that David talked about.
Chairman Tauzin. The gentleman's time has expired. The
Chair recognizes the gentleman from Massachusetts, Mr. Markey.
Mr. Markey. Thank you, Mr. Chairman. There's three matters
which I would like to get the witnesses on the record so that
we can use it perhaps to understand what we can anticipate 2 or
4 years from now if it be the case. Can I start with you, Mr.
Boccardi? Each of you may have referred to one or another of
these subjects in the course of your testimony, but I don't
think any of you have referred to all of them. Do you support
Congress enacting a uniform poll closing bill?
Mr. Boccardi. My personal view, and the AP does not take
positions on public issues, but my personal view is that it
would be a good idea.
Mr. Westin. Wholeheartedly.
Mr. Heyward. Yes.
Mr. Johnson. Yes.
Mr. Ailes. Yes.
Mr. Lack. Yes.
Mr. Markey. Second, will each of you, beginning with Mr.
Boccardi, reaffirm your commitment not to close--not to call
any State before all of the polls within that State have
closed?
Mr. Boccardi. Unequivocally.
Mr. Westin. Yes, as I said earlier.
Mr. Heyward. Yes.
Mr. Johnson. Yes.
Mr. Ailes. Yes.
Mr. Lack. Yes.
Mr. Markey. Third, do you believe that the solution to this
VNS problem in the year 2000 is to collectively throw more
resources at VNS so that the problem is not repeated in
subsequent years or is it in introducing more competition to
VNS so that each of you or groups of you have your own separate
polling system so that in and of itself serves as the check on
the others jumping the gun too quickly for fear then that the
other is available to accurately call the case?
Mr. Boccardi. My view at this point, Congressman, is to see
if we can fix VNS.
Mr. Markey. Fix VNS.
Mr. Boccardi. That's our view.
Mr. Westin. If I can answer both a little bit. I believe
that devoting more resources and fixing VNS in the way we've
discussed is the best way to ensure the accuracy and timeliness
of your reporting. Having said that, ABC News is open to a
second source if we believe it is as accurate or more accurate
and will work. We don't foreclose that possibility in addition
to it, but right now our attention is devoted to fixing VNS.
Mr. Markey. Mr. Heyward.
Mr. Heyward. We've opened both options in our report,
Congressman Markey, one to fix VNS or to go a separate way. But
realistically, I think it's important to acknowledge that VNS
extends our reach even, as Joan Konner acknowledged, far beyond
what any of us would be able to do individually. However, I
think we are going to try to develop additional sources,
especially in close races, our own unilateral polling,
additional forces on the ground to give us more checks and
balances so that we are not solely reliant on VNS. But I think
that consortium--there are great advantages in economies of
scale that really if used properly and if the recommendations
that have been made by the outside report and by others are
valid, I think VNS will be stronger and there is a good chance
it could serve the public even better than it has. But we do
agree that even more sources in close races are useful.
Mr. Markey. Mr. Johnson.
Mr. Johnson. At CNN our intention, and I underline
intention rather than decision, our intention is to stay with
VNS provided the changes can be made. A news leader is in place
already at VNS. We expected that there will be new technology.
There will be improved methodology. So provided we are assured
that that set of changes, that those sets of changes and others
that have been suggested are made. In addition to that, CNN
will have a second source. We intend to have an entirely
independent research or polling firm which will do research in
close election States. It will examine actual vote returns in
key precincts and then compare that with VNS data. So it's a
double checking system for us and we welcome joiners in the
expense of that undertaking.
Mr. Markey. If I could follow up with that. The Election
Day polling.
Mr. Johnson. Yes--no. First of all, we would determine
prior to Election Day what were the close States, where there
is likely to be a very close race, and then we would choose
precincts within those States to do on Election Day.
Mr. Markey. On Election Day. Is that the same process you
would be looking at, Mr. Heyward?
Mr. Heyward. Yes, but there would also be the option of
telephone polls to identify potential absentees in States where
that's likely to be a factor. That would happen beforehand.
Mr. Johnson. And that would also be our allocation.
Mr. Markey. Thank you. Mr. Ailes.
Mr. Ailes. We're the new kids on the block and we are the
smallest in terms of distribution of the networks and the
newest, and therefore this was our first experience with VNS.
Smallest in terms of distribution. We are doing quite well in
the ratings. But the answer is that, sure, competition is great
but this is an enormously expensive process. We've heard much
made today of the commercial pressures on news. Those
commercial pressures sometimes are good. If you don't make any
money, you can't buy cameras and you can't put extra crews out
and you can't pay for polls and you can't do some other things
that you would like to do. Is it best to stay with VNS? If you
reinvent a new model now and put it out in the field 3 years
from now, are we all going to wake up on election night and
say, gee, the new one didn't work as well as we thought it was
going to, or should we try to fix the one we have? I think all
of those questions are on the table. I think all of us lean to
fixing VNS and perhaps setting up some sort of checking system
against that so that we can take a sample of the sample and try
to determine whether or not we can build in a redundancy there
that will keep us from falling on our faces again. But I think
the bias right now is toward staying with VNS, although other
options are being looked at.
Chairman Tauzin. The gentleman's time has expired, but Mr.
Lack.
Mr. Lack. Yes, I would just quickly add that we would like
to see more resources put into VNS. We support that. We believe
that VNS can be improved, fixed if you will. The emphasis for
NBC News, which I referenced in my opening remarks, is to do
some original reporting on our own. It may support the data
collection that we're seeing during the course of an election
night. And we will explore, if we feel that we can't get the
satisfactory resources supporting VNS, an alternative system.
But for the moment we believe that we can get there. We are
going to explore whether they need another system of checks and
balances, and we're focused more on our own internal original
reporting for election night.
Chairman Tauzin. The gentleman's time has expired. The
gentleman from Florida, Mr. Stearns, is recognized.
Mr. Stearns. Mr. Savaglio, has VNS attempted in recent
years to get more money from the networks to update its model
computer system? If so, has this met with resistance? I think
if you could give us a little history. In the first panel we
talked to them about the problem being not just of recent
vintage but has been in existence for some time. And you might
comment about your discussions with the networks and has there
been any attempt internally on you folks to try and do
remodeling?
Mr. Savaglio. Well, the answer to the specific question
about asking for money, that's by implication been denied. The
answer to that is no. The budgets have been adequate to do the
job and have increased over the election cycles. In 1993, the
members invested a considerable amount of money in upgrading
the computer systems and that resulted in a new unified data
base, which would be a part of any new system. In addition to
which, in every election cycle, and this has to be done in the
periods between the elections, because so much of the energy is
devoted to covering those elections during the cycles, a
considerable amount of updating is done, reprogramming,
segments of the systems have been rewritten and redesigned. The
fact of the matter is that it's been pointed out that there are
some very glaring errors in November, but by and large our
computer system did function quite well, handled the stress and
the load of a very complicated system and did pretty well.
Just to this, one more point on that. There has been
discussions about updating, of integrating the system, and one
of our recommendations is that we plan to work to do that.
Mr. Stearns. What about the absentee ballots? You've known
about that absentee ballot problem for some time. Why didn't
you correct that earlier?
Mr. Savaglio. Well, the absentee ballot situation is a more
difficult situation. Simply put, it's not just the matter of
updating the system to accommodate it, because the absentee
ballots sort of by definition are it's information that is not
known. So you have to find alternative ways of getting it. The
fact is in the States where the largest amount of absentee
voting occurs on the West Coast, California Oregon and
Washington, we have implemented steps to take those into
account. We did telephone polls in those three States and our
information was pretty reliable. We do a lot of work outside of
the computer system itself in terms of assessing the absentee
votes, recording it, and the issues that have been raised here
just mean that, as we've said in our recommendations, that we
need to do further work and because absentee balloting is
spreading and growing in so many other States that it's going
to have to become a larger part of what we do.
Mr. Stearns. Now, I have here that the VNS system has been
under development and you've had an analysis since 1993, and it
says here because of budget limitations the project has not
moved beyond the creation of a partially unified data base. And
so even in your discussion, why didn't you make it part of the
computer system early on? I don't understand why the delay, and
why not doing it proper when it has such significance.
Mr. Savaglio. Well----
Mr. Stearns. Isn't it true since 1993 you've been working
on this problem?
Mr. Savaglio. I wouldn't say that. There was a project in
1993 where the members spent several million dollars on
updating the system and that did result in the unified data
base. What took place in terms of working on it between then
and now I'm really not aware of.
Mr. Stearns. It says partially unified data base. Why did
not we complete it?
Mr. Savaglio. Again, it's a very complex system and there
are portions of the projection system that are accounted for on
the election night that are not in the unified data base.
Mr. Stearns. They're mapping the genome here during the
last several years. Certainly you could map here the parts of
the unified data base, couldn't you?
Mr. Savaglio. Well, again as I said, the system, the
projection system, which are the two basic systems, there are a
number of systems, but the projection system operates on an
election night file that is generated from the unified data
base, but does operate separately.
Mr. Stearns. I yield back.
Chairman Tauzin. The gentlemen's time is expired. The Chair
recognizes the gentleman from Florida, Mr. Deutsch, for a round
of questions.
Mr. Deutsch. Thank you very much, Mr. Chairman. In previous
questioning I mentioned some statistical analysis that the
Miami Herald has done on the election. I just want to briefly
quote and submit it for the record.
[The material referred to appears at pg. 225.]
Mr. Deutsch. The ballots in majority black precincts were
discarded at a rate three times higher than those in nonblack
precincts. Nearly one in every 10 ballots in the majority black
precincts went unrecorded. In the majority white precincts the
discard rate was less than one ballot in 38. In fact, 19 of the
20 precincts with the highest spoilage rate in the State were
heavily black neighborhoods in Duval. The analysis found all
had at least a fifth of their ballots tossed out. A fifth of
their ballots tossed out, 20 percent. And I guess I bring this
again to raise the issue and actually ask Dr. Edelman just to
respond to it. I don't think there's any question that you can
make an analysis to compensate for the fact that African
Americans' votes were not counted. You could do a statistical
analysis and say because of literacy issues and because of the
fact that black votes were not counted as much, we can make our
analysis more accurate.
That's one thing you can do. And I don't think that we
should ask you to do that as the U.S. Congress. I think what we
ought to be talking about, Mr. Chairman, is what I keep
bringing back. We can correct for the reality of what happened
in this election, an election that when, as I mentioned, if
this was a foreign country and we had American election
observers, we would--no one would consider this a legitimate
process. If I can ask Mr. Edelman to directly respond to that.
Chairman Tauzin. Would the gentleman yield for a second? I
think we had an understanding that he would not testify today.
Is that correct?
Mr. Edelman. Nobody told me.
Chairman Tauzin. I don't think he was sworn in. That's the
problem. Is that it?
Mr. Deutsch. He's on the witness list, Mr. Chairman.
Chairman Tauzin. I see. We have to go through that process
if you want to direct questions to him.
Mr. Deutsch. I would be happy to because he's the only
statistician and I would like to ask him a very simple
question. The fact that black----
Chairman Tauzin. If the gentleman will suspend for a
second, let me do the formalities. Sir, do you realize that
under our practices the process of taking our testimony under
oath is our practice. Do you have any objection to testifying
under oath?
Mr. Edelman. No, sir.
Chairman Tauzin. The Chair then advises you under the rules
of the House and the rules of the committee you are entitled to
be advised by counsel. Do you wish to be advised by counsel
during your testimony here today?
Mr. Edelman. Yes.
Chairman Tauzin. But your counsel will not testify today,
is that correct?
Mr. Edelman. Yes.
Chairman Tauzin. Raise your right hand.
[Witness sworn.]
Chairman Tauzin. Sir, you are properly sworn and you can
answer the questions.
Mr. Deutsch. As you and I spoke, I guess it was yesterday,
if you can just respond to the issues that I've raised. The
fact that black voters votes were not counted, could that have
an effect on them misjudging the results?
Mr. Edelman. The exit poll measures people who believe they
voted. We interview people as they leave the polling place and
they fill out the questionnaires. We do not know if their vote
was counted. That's a whole new development that I think all of
us are becoming much more aware of. So the exit poll in Florida
was overstating the Democrat. So we don't know yet just where
the under and over vote is. It is still being counted and it is
still being looked at, but it certainly could be a factor in
the error to the exit poll.
Mr. Deutsch. Could you respond to the quote, the facts that
I just mentioned, that for instance in 19 of 20 precincts 20
percent of the votes were uncounted, where statewide it was in
the neighborhood of 2 percent? What kind of statistical
significance could that have in actually determining the
outcome of the election? I guess the premise of what happened
in Florida is that it wasn't 500 votes of people who thought
they were voting for Al Gore. It wasn't 5,000 votes. It really
was 50 or 75,000 people left the polling place net who thought
they were voting for Al Gore whose votes didn't count. How
would that affect your results or your exit polling results?
Mr. Edelman. Well, if I knew that--the contribution error
of the error in the exit poll was 2.6 percent. So it's
conceivable since the remaining percent was 3 percent that that
could account for that. But that's a fairly big assumption. But
we would never give that as a reason for the wrong call in
Florida. I mean it would be one factor in the exit poll being
off.
Mr. Deutsch. Right, and again I guess I know you're not
giving it as a reason, but you're also saying statistically it
could be. Can I just ask why haven't we investigated that? Why
aren't we focusing on that issue?
I'm just asking. Maybe your results weren't as bad as you
think they were. Maybe you did a good job or a lot better than
this committee is alleging. Maybe you did what you were
supposed to do because it would seem that the statistics and
the numbers that I'm pointing out really get back to the point
that the exit polls were a truer reflection of what people's
intentions were than the actual vote count that occurred in the
State of Florida.
Mr. Edelman. It's certainly possible what you're saying.
I've been waiting to see the results of the counting before I
can make any kind of judgment like that, and I frankly have had
my hands busy with a lot of the reports and analyses and
everything else that's been going on.
Mr. Deutsch. Can I just follow up a little bit, and again
because I don't think we want to do this, but you clearly could
make a statistical analysis that in black precincts discount
for the fact that people's votes don't count so you can make
your results more accurate through that type of adjustment?
Mr. Edelman. That's very risky, from what we've known about
this possible error in the exit poll for a good while. But just
how much do you adjust for it? We don't know. If it were a very
simple number that I could just come up with and put in the
computer, I certainly would do it. I don't know. I don't know
how many people were affected.
Mr. Deutsch. Let me mention there is no reason to believe
that--again Florida is the center of the post-election
analysis, but I don't believe there is any reason to believe
that this phenomenon of African Americans being discriminated
against in terms of voting is any different in any other
location in the United States of America. And I think when we
talk about that 34 States with a bias being Democratic, I think
the phenomenon that we are really looking at is a socioeconomic
phenomenon of illiteracy in America, of functional illiteracy
in America, of issues of discrimination that basically come
back to exactly where we see--I don't see any statistical basis
to say that African Americans in Florida are any worse off than
African Americans in any other State. Is that not possible that
the phenomenon you see in the other States is exactly due to
this phenomenon as reported in Florida?
Mr. Stearns. Would the gentleman yield?
Mr. Deutsch. I will be happy to.
Chairman Tauzin. I would ask that you ask an additional
minute.
Does the gentleman ask the gentleman be extended an
additional minute?
Mr. Deutsch. Yes.
Chairman Tauzin. Without objection, so ordered. The
gentleman will yield.
Mr. Stearns. There are two things. The committee has
analyzed this and they have a graph that shows that 34 States
plus the District of Columbia overestimated for Vice President
Gore. If that's true, then how does what you're talking about
affect--I mean what difference would that make because it's
already been overestimated?
Mr. Deutsch. Right. Taking back my time, the whole point of
what I'm saying is if the overestimation is due to the
phenomenon I'm talking about, and which you have a statistician
saying absolutely it could be, then is our concern saying to
the statistician you're not doing a good job or should our
concern as Congress be, hey, we've got this incredibly awful
situation in the United States of America where hundreds of
thousands of people's votes are not being counted because of a
societal bias against African Americans?
Mr. Stearns. Would the gentleman yield? The point is you're
not providing any evidence to show that's true. You're
speculating.
Mr. Deutsch. That's not----
Mr. Stearns. There is no evidence to show what you're
talking about.
Mr. Deutsch. You're absolutely incorrect.
Mr. Stearns. Give me the documentation.
Mr. Deutsch. I'm giving you the documentation again, and I
will provide more to the committee, and I wish we spent as much
time from the committee staff and from the Congress' staff
looking at these kind of numbers than what we are at this
committee. But I just read some statistics from our own State,
Mr. Stearns, that 19 out of 20 precincts with the highest
spoilage rate, that the ballots in black precincts, the rate
that they were spoiled, not counted, was 1 in 10. In white
precincts it was one in 38, almost a 400 percent larger number
in the African American precincts.
Mr. Stearns. We are talking about exit polls.
Mr. Deutsch. Right, but the exit polls----
Mr. Stearns. We're not talking about what you're talking
about.
Mr. Deutsch. I am going to try one last time to try to
explain a little bit of statistics on this.
Chairman Tauzin. The gentleman's time has expired, so if
you will respond very quickly.
Mr. Deutsch. I will. What the exit poll accurately counts
is what people say who they supported, who they thought they
voted for. But in African American precincts the percentage of
people who said they thought they voted for a person, Al Gore,
and that vote did not count is much higher than in non-African
American precincts, and because of that there is a statistical
bias. If you don't adjust for that phenomenon, there is a
statistical bias that would say Al Gore did better in terms of
the actual count, and that is exactly what we are talking about
here. That's the phenomenon.
Chairman Tauzin. The gentleman has made his point. His time
has expired. The gentleman, Mr. Greenwood, is recognized for 5
minutes.
Mr. Greenwood. Thank you, Mr. Chairman. As I said in my
opening remarks, as uncomfortable as you folks in the media
make us when you grill us incessantly, we're probably--at least
this member is--more uncomfortable grilling you than being
grilled. And some of you have expressed your concerns about the
propriety of this hearing, but I hope as this process winds
down you come to realize that nobody in this committee or this
Congress has any intentions of legislating with regard to your
ability to project elections. You can project the elections the
day before the election if you want. It's none of our business,
frankly, whether you do or you don't. But we do in the Congress
have a decision to make as to whether to try to respond to what
you do, and of course we've all been talking about unifying the
closing of the polls in an effort to do that. And you've come
here to tell the American people today that you're going to use
exit polls and you're going to continue to project elections.
And since we have some concern about whether that impacts the
election process itself, this will educate our decision as we
decide whether to make the people in California get up earlier
in the morning to go to the polls or not. They may not welcome
that without information we've gleaned today.
On that subject I know that Mr. Johnson has been quoted sa
saying, Johnson expressed concerns that calls of States in the
Eastern part of the country affect voter turnout in the West
where polls are still open. He questioned surveys that claim no
such effect. Quote, from my days in California I had so many
friends who did not go out and vote when they heard that the
election was over. I believe it's an influence, a big
influence.
I believe at the end of the day it's a central question for
us as to whether that is a phenomenon that happens or not. I
think each of you relatively heartily endorsed this legislation
to close the polls at the same time, and I would assume the
reason you endorsed that is because you have some concerns
about this phenomenon.
Can I ask each of you to let the committee know do you
believe or do you at least worry about whether or not folks who
have not yet voted and hear projections from the East are
affected by that call, those projections?
Mr. Lack. I don't believe there is, as has been discussed
earlier today, any solid evidence that would indicate that
there is a voter suppression as a result of projections in the
East, but the reason I support your point of the uniform poll
closing is any perception that there might be should be avoided
and should be eliminated and to the extent this legislation can
put an end to it, put a rest to the whole subject, I'm fully
supportive of it for that reason.
Mr. Boccardi. I think we think if you pass this law we will
have to stop to answer the question.
Mr. Johnson. Mr. Greenwood, as the person who you quoted, I
spent 13 years in Los Angeles either as president of the Los
Angeles Times or publisher of the Los Angeles Times. There were
many different surveys that I saw during those 13 years which
indicated to me a wide array of results. But it was my personal
view based on those 13 years, having met with many Members of
Congress, many city officials and others, the university
communities, that there is a definite effect on voters as it
relates to the Presidential election. In the local races many
people still seem to go out on the local issues, on sheriff and
many of the State propositions, for example. But it is a
personal belief of mine, and yet I feel it so strongly that it
is a part of the reason why I strongly recommend the uniform
poll closing act.
Mr. Greenwood. Mr. Ailes.
Mr. Ailes. Yeah. I think it's impossible to determine
exactly what affects--I don't think that will ever be
determined--information affects people and if it's a blowout
election there is no question about it. If it's not a blowout
election, then it might affect some people to go out and other
people to say home. There's no way to determine how it affects
people. But in the end I think Ms. Konner who was on the--if
I've got her name right, said people have to vote. In the end
people have to take responsibility to vote. Our job is to
gather and distribute, report information, and we will do that,
and we should make it clear to people that what we're doing on
the screens should not affect their vote.
Perhaps we maybe even need to say that. Please go vote.
That's fine, I think, the more information the public has and
the more responsibility the public takes to go vote, because we
learned in this election every vote counts and it may be too
close to call. So I support the legislation. I think it's
simplifies things for everybody. It eliminates some of the
problem, but we are never going to know the extent or in what
direction it swings people.
Chairman Tauzin. The gentleman's time has expired, but I
see some of you want to respond. You're welcome to respond, Mr.
Heyward.
Mr. Heyward. Very briefly. I think voting is about a matter
of personal responsibility. I think we're speaking here as
citizens as much as journalists. I think of the different
solutions, Congressman Greenwood, this may be the simplest. The
others seem to raise not only journalistic issues of
withholding information that is easily available but also
States rights issues, as one of your colleagues mentioned
earlier. To ask States not to report results of an election
that's completed in their State seems problematic from your
perspective.
So I think we're trying to be good citizens, even though as
Mr. Boccardi said we don't normally take positions on political
issues, but that's really behind the endorsement.
Chairman Tauzin. Mr. Westin.
Mr. Westin. I don't know the answer to the question of the
effect. I've read some of these studies, in fact, and from what
I've read, they generally conclude there isn't enough data and
we don't know. It would be somewhat ironic for me to sit here
today before this committee and say when we report something
such as the projecting a race, that no one is paying attention
to it. I would hope that some of our audience is paying
attention.
Now, what they do with that information I don't know. I
don't know whether it encourages them to go to the polls or
discourages them to go to the polls. I don't think this
committee or our audience would want a world in which ABC News
or some of my colleagues believe that they knew who the next
President of the United States was, but they sat on the
information for a few hours. Given the way the world works,
that would be dangerous, I think.
Therefore, I support this legislation wholeheartedly
because I think once and for all put this behind us, and we
wouldn't have to talk about it in 4 years or 8 years or 12
years.
Mr. Heyward. I would add that Mr. Johnson is a person of
extraordinary integrity. If I have to explain to him why I
hadn't voted, I would definitely blame the networks rather than
say I was too lazy, just felt like going home.
Mr. Boccardi. If I might make one short point, Mr.
Chairman. You know, you start down a road here that I think can
get pretty dangerous and pretty difficult. If you go to accept
some of the theses now on the table, what about the preelection
polling? If there is a poll that says this candidate is 20
points ahead, as been known to happen in a congressional
district in Louisiana, even more, is that keeping people from
the poll? Well then maybe we shouldn't do those. You see where
the trail goes. I don't think that's a trail we or you would
want to start down.
Mr. Greenwood. Worse yet, if the predictions come in too
early, people stop contributing to our campaigns, and that
would be a disaster.
Chairman Tauzin. Thank the gentleman.
The gentleman Mr. Stupak I think is next for 5 minutes.
Mr. Stupak. Thank you, Mr. Chairman.
It's been alleged throughout this hearing that there's some
bias, network bias, probable bias, political bias by the
networks. So do you believe your networks lean either way,
Democrat, Republican? Mr. Lack, let's start with you.
Mr. Lack. No. We've testified to that fact. We vigorously
believe that there is no evidence of any bias in our reporting.
Mr. Stupak. Mr. Ailes.
Mr. Ailes. I agree with that, and I say the same thing for
Fox.
Mr. Stupak. Mr. Johnson.
Mr. Johnson. No bias whatever, sir.
Mr. Heyward. Categorically no bias. Emphatically not.
Mr. Stupak. Let me ask you this question: Hopefully you all
saw the video that we had at the beginning of the hearing. Do
you believe the video you saw accurately depicted your coverage
on election night?
Mr. Heyward. Let me jump in on that, because I think--I
certainly hesitate to accuse Congress of misleading editing and
distorting what had happened. That would be a dramatic role
reversal. I do think that the video, while it was effectively
done, did not actually convey the viewer's experience of
election night. We actually looked at our transcript very
carefully, Congressman. And I think the sense the video gave
that Vice President Gore was on an unstoppable roll I don't
think was reflective of the reality. In fact, if anything, our
outside expert, Kathleen Hall Jamieson, who I mentioned before,
concluded that we might have overstated the degree to which
President Bush was on a roll. But certainly we displayed the
popular vote graphic 15 times between 7 and 11. President Bush
was ahead every single time; on the electoral count, 75 out of
100 times.
Mr. Stupak. So you don't think the video then reflects----
Mr. Heyward. The video that gave the impression that the
networks were saying Gore's got it in the bag I believe was
misleading, yes.
Mr. Stupak. Anyone else care to comment now?
Mr. Westin. I would be first to admire a good editing job,
and that was a good editing job, no way about it. I don't think
it fairly and fully reflected all the statements made. I mean,
I could even read you into the record statements were making
around 8:15 in the evening saying this was a very close race.
In fact, we said at 8:18 a candidate could win the big three
States and still be in danger of not winning the electoral
college, as one example out of several.
As I say, I respect the job that was done, but I don't
think it fully reflects the experience of the viewer, which was
a much more balanced, much closer race throughout the night.
Mr. Stupak. Let me ask this question. Instead of going all
the way down the line, maybe we want to go down--do you believe
we here today in Congress, this hearing, do you believe we have
accurately reflected on your coverage in the 2000 Presidential
election? All of us up here, do you think we've accurately
reflected how you handled the 2000 Presidential election?
Mr. Westin. You know, with respect, I don't think that's my
position to judge. I'm here as a member of the Fourth Estate.
We made some serious mistakes that we have to adjust for. The
way--as I refer to my remarks, the way it works in this
country, which I firmly believe in, is we don't look to the
government to correct the press. We look to the people and the
people's representatives as represented by this body. We take
criticism. We don't necessarily enjoy it, but we're open to it.
We should take it into account, we should listen to it, and we
should then take it upon ourselves to decide what we should
adjust to and what we should do in the best interests of our
audience. And if we fail, the audience will judge us, and
they'll move somewhere else.
Mr. Stupak. Anyone else care to comment on that?
Mr. Johnson. Congressman, I would only say in a way it
related to interviews we had with Members here. You will have
many accomplishments as a part of your record during the year
and will perhaps make one bad vote or one mistake, and you
accuse us, of course, of focusing on that.
Clearly I think we had an excellent year of coverage of--
all the networks, I think, dedicated an enormous amount of time
and effort, too, and I would say that despite what Andy Lack
has said, I agree with what Andy said, I believe we did a very
good job in a campaign year. We did make a mistake that night.
You got us. And I think we are here not only to assure you that
we will fix it, but it won't happen again.
Mr. Stupak. No different for any of us up here. We make one
vote our opponent don't like, and that's the whole campaign.
So, I mean, just wanted to see how you felt on being on that
side.
Let me ask you this question. Mr. Boccardi hit it a little
bit. I wanted to ask it. There's been a lot of focus on exit
polling, but what about those preelection polls? You believe
they influenced voters; when you do an exit poll on the east
coast, it influences the west coast? What about the preelection
stuff where you have Gore up by 5 points going into the
election, and all of your networks that we watched, you had
wild swings in your polling throughout this campaign season.
Should that be tightened up? Should that give a more accurate
depiction? Because I don't know how could you have such wide
swings from network to network.
Mr. Boccardi. If you're squabbling about whether exit polls
on Election Day affect outcome, we'll be here for 2 weeks
talking about the other polls.
Mr. Stupak. I don't think you can make that determination,
so that's why I asked about preelection polls.
Mr. Boccardi. I know of no evidence that people go to the
poll or don't go to the poll because--to vote because there was
a poll.
Chairman Tauzin. The gentleman's time has expired.
Mr. Stupak. Thank you, Mr. Chairman.
Chairman Tauzin. But I want to take the opportunity to
declare unequivocally there was no intentional bias in the
editing of that tape.
Mr. Ailes. Mr. Chairman, we felt Fox News was
underrepresented in that tape and that, therefore, it was
biased.
Chairman Tauzin. Could have been. Who knows? We'll never
know because there's no empirical evidence, I swear. You won't
find it like we didn't find it.
I also want to tell you, by the way, and this is from the
heart, I don't think any of us would have hired an outside
consultant to be as critical of the CNN job, to criticize our
votes. That was an act of extraordinary, I think, contribution
to this effort. I want to thank you for doing that. And thank
you all for the kind of self-evaluation you did. I think
America has something to be grateful for today. I really mean
that. I thank you for that.
The Chair now yields to--who's next? Mr. Ehrlich.
Mr. Ehrlich. Thank you, Mr. Chairman.
This issue of early eastern projections impacting close
races in the West is interesting. Mr. Johnson, I heard your
previous comment, and I think intuitively it makes sense to me.
It make sense to most people.
Mr. Heyward, I think your earlier testimony--in your
earlier testimony you took a different view. Florida was called
by ABC at 8, CBS at 7:50, and NBC at 7:49. We had real close
races in Iowa, Minnesota, New Mexico, Washington and Wisconsin,
as you all know, within a percentage point to one another.
So, it seems to me an Iron Triangle State, we all know the
conventional wisdom was those three States were vitally
important to the candidates. They spent their time there, they
spent their money there. I think it makes sense that those
early calls do impact Western States.
With regard to Michigan--and I'll let you answer, by the
way, I just want to make another point--I guess what really
bothers me is VNS makes its call at 9:23, and ABC makes its
call at 8:06. CBS makes its call at 8. NBC makes its call at
8:02. Fox makes its call at 8:05, and CNN makes its call at 8,
maybe the critical State in the election, and I understand
you're not necessarily following precisely of VNS, but that
concerns me.
So I would like you to respond to one or both points I just
made, one with regard to VNS. I think you have in your earlier
testimony. But clearly intuitive, I think for most people it
makes sense if you have close races in the West, early
projections in critical States in the East may impact the
bottom line of a Presidential election.
Mr. Heyward. As has been said before and a lot, I don't
believe there's conclusive evidence, but I don't think we have.
And there's a network solution to the issue of effects in the
West. I think the solution, as we've said, is the uniform poll
closing, because we've all said we're going to withhold the
call at any given State until all the polls are closed there.
So that would take care of the Florida problem going forward.
In other words, whatever we agree or disagree or be able to
figure out what happened in the Panhandle, but that's not going
to be an issue next time because everybody is going to wait
until those polls are closed. So that, I think, is as far as we
have all said we're going to go.
In terms of the West, I don't see what the networks could
do. I don't think it's realistic to ask us to hold back the
news of an election that's complete, Secretary of State is
reporting the results, because of this perceived effect. I
think this other solution, as I said earlier to Congressman
Greenwood, is much simpler.
I didn't quite understand your question on Michigan.
Forgive me.
Mr. Ehrlich. You all quite appropriately point out
Pennsylvania, Florida, Michigan, the Iron Triangle. It's
interesting to me that you all look to VNS and use VNS, and not
necessarily in lockstep, but certainly very relevant. Yet you
make--all of you made your calls almost 1\1/2\ hours before
they made their call in such a critical State. Pennsylvania,
Eastern State, very important State, VNS made their call at
9:24, yet CBS made their call at 8:47, and CNN made their call
at 8:48. These are really critical States that people are
looking to.
As I said here again with regard to impacts on the West, I
think it's relevant.
Mr. Heyward. I think what this really illustrates, we're in
the curious position here of facing the suggestion that we're
both not competitive enough and too competitive in the
situation. I think each decision desk is honestly trying to
evaluate the data as best as it can in sifting many
ingredients. We, coincidentally, given the chairman and the
ranking minority member, happened to call both Michigan and
Louisiana right at poll closing, which VNS did not. That's
because our decision desk conscientiously trying to be accurate
believed that we had enough information to make the call, and
it turned out in both cases to be correct.
There is--there is a journalistic imperative to be accurate
and to be timely. I think being accurate, as we've heard here,
is much more important than being timely. But to some degree
those imperatives have collided, but not in those two States
and not in Pennsylvania where we made calls that were accurate.
I don't see what the problem is if we call a State at poll
closing based on the data that our decision desk has and it's
right.
Mr. Ehrlich. It's interesting given the relevance of VNS, I
guess--and you all have appropriately, I guess, criticized the
exit polling earlier with regard to Michigan. I'm getting
information that only 29 of 45 exit poll precincts were
reporting when the call was made. That's--that's not real good
science, I guess.
Mr. Heyward. Our decision desk was very confident of the
call. In fact, Michigan has a large exit poll sample, and the
other indicators that were used, as I mentioned earlier--the
exit poll is just one of many tools that these analysts used--
were in sync, and the feeling was that this network, and
obviously some of the others that made early calls there, that
we could go ahead. We're not entirely dependent on VNS. We also
have this independent decisionmaking body, and in this case the
call was correct, as was the call to Louisiana.
Chairman Tauzin. The gentleman's time has expired. If
anyone wants to respond.
Mr. Ailes.
Mr. Ailes. I would just like to comment, two quick
comments, the first one a personal comment. It was my personal
view after looking at the coverage that we all made so much out
of Michigan, Pennsylvania and Florida without going on to say
there are other mathematical models for people to win. That
there was a perception created on election night that focused
on three States, and once those States went down, people walked
away. I think I was with a group of people, and, frankly, when
Florida went, 40 percent of the room emptied out and said,
well, it's over, we're going home.
So I think there was some--it's just a personal view, and
I've told our journalists that we ought to be careful about
telling people that this is how the election--you're going to
be able to know who won, because we all hire outside
contributors and consultants and political people to tell us
these things, and they all came to that conclusion. It was
certainly true they were all important, but I think that we
probably overstated and oversimplified it.
The other thing I would say is that, having watched the
decision desk this time and watched in other cases, these are
incredibly dedicated people, Republicans and Democrats. We
happen to have three Democrats, one Republican on our decision
desk. These are not partisans. These--at that night, these are
people who are trying to get it right. They are desperate to
try to get information and look at those screens and check
facts, and they are professionals, and they feel the worst
about screwing up, which is what happened on election night. It
was a monumental screw-up. Everybody did it. But these are
individuals who are hard-working, dedicated, and really not
trying to do anything weird. They're trying to get something
right.
Mr. Ehrlich. Mr. Ailes, I am out of time. I appreciate your
forthright answer; both of you, in fact. I had a similar near-
death experience in Timonium, Maryland, at 7:52 p.m. that
evening. So I appreciate it.
Chairman Tauzin. The gentleman's time has expired.
The gentleman from Texas Mr. Green is recognized.
Mr. Green. Thank you, Mr. Chairman.
I know it's been a long day, and I appreciate these folks
being here.
It's interesting, Mr. Ailes, you said you have three
Democrats and one Republican on your decision desk. That isn't
the one Republican who kept calling Austin, was it, that we
heard earlier from one of my colleagues?
Mr. Ailes. You want an answer to that, or was that
rhetorical?
Mr. Green. You can answer it. I thought it was interesting,
though, because of the publicity.
Mr. Ailes. It's incorrect. We didn't keep calling Austin.
But go ahead.
Mr. Green. The exit polls versus before election polls, I
think, is correct. I think each of the networks have realized
the problem we have. But, again, if we back it up, if we
remember, the before election poll showed George Bush winning
the popular vote and Al Gore squeaking by on electoral victory.
So I don't know if that impacted the election any because it
was actually the reverse when--after the U.S. Supreme Court got
through.
I guess I feel frustrated like you do because blaming the
election process, not just election night, now that was you
all's responsibility, but the next weeks were literally a
nightmare for those of us getting up every morning. I felt like
I was in Groundhog, the movie, because every morning I would
get up and see Tallahassee or see Palm Beach. I didn't go to
Florida. Coming from Texas I wouldn't go, Mr. Buyer, unless
they'd let me count the votes, but because of the problems that
we were seeing literally played out for our Nation.
And so that was--I think that's what our focus ought to be.
You've agreed the things that need to be done, and I think each
network independently and by the CNN analysis report talked
about what needs to be done.
Now, the question I have, though, is we realize that you're
going to correct what was done on an election night and be more
assured of what's happening. But the voter education, it
frustrates me when I see the number of ballots that are
discarded of people that said they came out and voted for Al
Gore or George Bush, but their vote didn't count. What can the
networks do to help that prior to the election to say, look, we
have five different ways, or how many ways we count votes in
our country or how we ballot.
I know in my own area in Houston, we have two local
affiliates, CBS and NBC, who for the last 2 years have been
providing time for candidates to go in. We appreciate that,
although, I always said we really don't need 5 minutes; you can
do everything in 2 minutes. But can the networks do something
to help on voter education to make sure that those folks can--
well, they know how to vote?
Mr. Boccardi. I think it goes without saying that next time
before the election there's going to be a raft of stories
illustrated with graphics and almost step-by-step instruction
about how to vote. I think you can count on that. Whether
that's going to make any difference, I haven't the slightest
idea.
Mr. Ailes. Unless we can put that in the middle of
Temptation Island, I don't know if the people are going to
watch it. So I think that trying to do voter education is--in
the middle of commercial television is not the easiest thing to
do.
Mr. Green. I love the setting on Temptation Island.
Any other responses?
Let me just say a little personal note, because I notice in
one of the points that one of my colleagues made with--the exit
polling was really the story, and oftentimes it became in--
today it became the story, the exit poll. One of the things I
noticed, and it happens on network news and even our local
news, is oftentimes Temptation Island may be promoted, or
whatever is going on on Temptation Island gets more coverage on
the news than sometimes other hard news items.
Has anybody ever talked--since you're all here today, it's
interesting, because I like--all of us are news junkies. I will
see things that come up that are reporting on what is maybe
happening that night on a sitcom or something in the news
segment. Has that ever been discussed in some of the--I know it
has nothing to do with today, but it's interesting.
Mr. Westin. I guess I'll speak for ABC News. In our news
programs we exercise our own independent editorial judgment
about what is newsworthy or of interest to our viewers or what
isn't, frankly sometimes to the chagrin of the company that
surrounds us because they would either not like us to report on
someone else's program or do so on ours. But we're entirely
independent. There's a church and State issue there. If you see
it on a newscast on ABC News--and I'm sure this is true for my
colleagues--if you see it on an ABC newscast, it's because ABC
News has made an independent editorial judgment that we think
this is newsworthy and something that our audience is
interested in.
Chairman Tauzin. Thank the gentleman.
The Chair recognizes Mr. Shimkus for a round of questions.
Mr. Shimkus. I'm sorry. I've been in and out the whole day,
but I appreciate the testimony.
A simple question that is posed by the charts, if you will.
Based upon the research done by committee staff, Michigan and
Pennsylvania were aggressively called when they had similar
high crit numbers. The question I want to ask is why wasn't--
why weren't--especially with all these literary and broadcast
types and English majors, why weren't Ohio, Georgia and
Virginia not called as aggressively as Michigan and
Pennsylvania?
Mr. Westin. Just speaking for ABC News, Virginia was called
28 minutes after poll closing, which was much faster than
Michigan or Pennsylvania. Pennsylvania for us was 79 minutes
after poll closing.
Mr. Shimkus. Michigan was called at poll closing; is that
correct?
Mr. Westin. No. For ABC News it was actually 6 minutes
after; but very close, that's right.
Mr. Shimkus. And Virginia was 28 minutes after poll
closing.
Mr. Westin. Yes, sir.
Mr. Shimkus. When the numbers are--would justify as quick
of a calling as Michigan and Pennsylvania----
Mr. Westin. Well, you have to forgive me because I can't
read that chart.
Mr. Shimkus. We can fix that for you. We'll just put it
right up.
Mr. Westin. So I don't know what you're looking at. But let
me explain something to you. There's been some talk about this
legendary crit.
Mr. Shimkus. Let me get to--my question was based on the
research of the staff. Let me just give you the numbers, and
you can make a determination of how you can respond to the
question. If you look at Michigan and Pennsylvania, the crit
needed to win was 2.6. The crit was 2.4. Likewise Pennsylvania
crit needed to win was 2.6. Crit was 2.2. Here the crit in
Ohio, Georgia and Virginia was in excess of the crit needed to
win, but they were called much later than Michigan and
Pennsylvania.
Chairman Tauzin. I think for clarity--would the gentleman
yield? I think for clarity we ought to again explain to those
who are watching and listen the crit, that number that Mr.
Shimkus is citing is that critical statistical number that is a
minimum requirement before a State under VNS analysis can
safely be called for a candidate. And the question, as I
understand this question, is that in two States, Pennsylvania
and Michigan, it appears as though those States were called
before the crit numbers were achieved, and yet a number of
other States achieved their critical numbers and were not
called for a significant length of time.
Thank you, Mr. Shimkus.
Mr. Westin. If I could respond, and I apologize if this is
a little detailed. First of all, there's not just one crit
number on the decision screen that appears that our decision
team looks at. There's a separate crit number for each one of
the models of which there are several, and those different
models have different sorts of information put into them. So it
depends--it's hard for me on the basis of this to respond
because often the crit number you may be looking at is actually
the composite.
Mr. Shimkus. Actually the ones that we're looking for are
the best numbers.
Mr. Westin. Right, but sometimes in the model, the best
model that kicks out on this decision screen is the composite,
which is heavily weighted by prepolling, preelection polls,
sort of a weighted average of preelection polls, which is not
as valuable. So it depends.
The main point I wanted to make----
Mr. Shimkus. But all the other models will show lower
crits. This is the highest one.
Mr. Westin. No, but that--with respect, that is not exactly
the point. But the point I'm trying to make is this.
Mr. Shimkus. But that's my point.
Mr. Westin. There is no magic to crit numbers. A crit
number simply is a comparison of the margin that is being shown
in a given statistical model either from the exit poll or from
the actual vote tally, a comparison of that with the margin of
error, which could be big or could be little. There is no magic
crit number. For VNS they have a limit. That's right. We do
not. That is one of the factors we take into account. There are
a number of other factors. That's why we employ all the people
we do on the decision desk to exercise their judgment based on
past experience and their knowledge of the models and things.
Mr. Shimkus. So what's your tolerance for risk?
Mr. Westin. I want to make sure it's accurate.
Chairman Tauzin. Would the gentleman yield a second?
Would you just answer this Mr. Westin: Was Michigan an
aggressive call?
Mr. Westin. I believe that Michigan was reasonable, but,
yes, it was aggressive. Yes. There were other aggressive calls
made as well. Alabama, in my judgment, looking at the decisions
reached today, was an aggressive call the other way, but it was
a reasonable call based on the data in the decisions reached as
they're replicated now.
Mr. Shimkus. Mr. Edelman, you agree there was too
aggressive of calling; am I not correct?
And as you answer that, are some of these statistics in
line with that evaluation?
Mr. Edelman. Well, let me say a couple of things about the
screens. There's some problems in the ways that you all are
reading the screens, and I think that's leading to some of the
confusion. So if could have a minute, I'd like to explain that.
First of all, the screen is only one piece of information
that we use when I make a call. There's other additional
information that goes into it. There's--we have research on
absentee voting, and we have that available to us when we're
making the projection. So even though that number is not on the
screen, it's available to a person making a decision.
The other information we have is we have the history of
that State in terms of any kind of errors in the exit polling
in the State. So we have that information as well. So we have
that kind of information as well as we have information about
what's been going on that night, and that is also affecting how
we do that.
So to just take a number from the screen is not a very--
it's not a sufficient way of commenting on our process, or on
commenting on the risk.
There's another problem as well, and that is like in
Michigan you say there are 29 out of 45 exit poll precincts
that I believe that--I would have to check your screen, but the
screen--and we overlay exit poll precincts when we get real
vote, so it may look like there's less exit poll precincts, and
it is because we have real vote in for those precincts. With
all those in mind, it's much more a judgment. The decision is
something where the person making a decision has to take all
these kinds of factors into mind.
Mr. Shimkus. You can understand us laymen trying to figure
out these formulas and variables and stuff.
Mr. Edelman. Oh, yeah. I have been working with your staff.
Chairman Tauzin. Thank you. The gentleman's time has
expired. Thank the gentleman.
The gentlelady Ms. DeGette is recognized.
Ms. DeGette. Thank you, Mr. Chairman.
Now, all of you gentlemen in your opening statements talked
about the reason why there's this push for early calling and
exit polling is competitive--competitiveness among your
networks, right? I would say that would be accurate. All of you
came in here, in response to Mr. Greenwood's question, some are
more enthusiastic than others about a uniform poll closing
time. But all of you said you don't really think that the
current lack of a consistent poll closing time is changing
elections. Right? Would that be accurate? Do you think that
because we have disparity in poll closing times that's changing
elections? Do any of you think that?
Mr. Ailes. We think we haven't been able to determine
empirical data to determine that.
Mr. Johnson. I personally do think.
Ms. DeGette. You think so.
Here's my question: Let's say we adopt a uniform national
poll closing time, which, in my mind, being a westerner, will
raise some issues. You're closing the polls at 9 in the East.
You're kind of fudging with daylight savings time in California
and other places. Don't you think networks will find some other
way to then make this election interesting? Mr. Boccardi talked
about it a little bit when he said, I mean, after all, we do
pre-Election Day polling. What about--I mean, exit polling
right now doesn't happen at 7 p.m. when the polls close. What
are the networks going to do earlier in the day now that they
can't start projecting New York and some of the other Eastern
States well before California and the other Western States?
Mr. Heyward. We already for a long time have refrained from
characterizing races, and we're very careful what kind of
language we use. As you know, Congresswoman DeGette, this
information starts coming in very early afternoon. In fact, a
lot of Members of Congress call us for it even though we don't
release it to the public. But--so I think we, you know, it's
sort of our problem. And I don't say that to be flip at all.
It's not Congress's job to make our life easier, make the
election interesting. We will report.
Ms. DeGette. But that's my point. I think you will report.
And I think, you know, you come in here and you testify, look
we won't--we won't report on--we won't call a race until a
State is closed. If you have a national time, then all States
will theoretically close at once. So don't----
Mr. Heyward. What I'm saying is----
Ms. DeGette. So don't you think that a practice will
develop earlier in the day where you may not call a State, but
you don't have to call a State?
Mr. Heyward. I don't think so, because we have good
evidence that that's not the case. We've already--and actually
in some of the halcyon days that the chairman was referring to
with those quotes earlier, there was characterization before
some of the agreements that were made with Congress in the mid-
'80's where races were characterized right and left. The exact
kind of the effect that you're worried about happened, kind of
a wink and a nod and here's what's really happening.
But we've exercised enormous restraint, and we have--I
think you would be--I hope you would be impressed by the
intensity of the debate at 6:30 when the evening news goes on
in the East exactly what we can and cannot say to comply with
what we said.
Ms. DeGette. I understand that.
Mr. Lack, do you feel that way as well?
Mr. Lack. I don't believe it's going to be an issue for us.
We will report it when we get it, and we'll report it in a way
that it ought to be reported, but we're not going to be sitting
around, gee, because we can't report it earlier, we have
nothing to do.
Ms. DeGette. No, no. It's like what Mr. Heyward saying the
wink and the nudge, you know, you're not calling it, but you're
saying it's looking pretty good for----
Mr. Heyward. My point is that we don't do that, and we
won't.
Ms. DeGette. My question is will you slip into that because
now you don't have----
Mr. Heyward. I don't think so.
Ms. DeGette. [continuing] the horse race?
Mr. Lack. No.
Mr. Boccardi. I think one thing, that an assumption that's
sort of built into the questioning here is that uniform poll
closing is a magic bullet that is going to solve everything.
All the polls will close at the same time.
The reporting processes in all the States are not uniform
now from the current closing. Some States come in more quickly,
some States come in more slowly. This morning there was a
reference to Pittsburgh and Philadelphia. So there's going to
be a race after the poll closing that will engage us all in
trying to understand what's happening. What we're taking off
the table is the poll close.
Ms. DeGette. Thank you.
Mr. Ailes.
Mr. Ailes. When Walter Cronkite first went on the air with
his evening newscast at CBS, it was 15 minutes in length. The
biggest concern at CBS was of everybody running around in the
halls trying to figure out is there enough news in the world to
fill 15 minutes. Now that you have these endless news channels,
we are very creative. I think what it will do is force the
creativity level down the ticket to more referendums that are
going on around the country, other interesting side-bar stories
related to the election.
Ms. DeGette. Thank you.
Mr. Boccardi just made my point exactly. It's not--the
magic bullet here is not a uniform poll closing. I think that
anybody who thinks that that will solve the problem, I think it
helps, but I think that there may be other problems we need to
look at as well, and I think that's part of it.
Another issue, and perhaps some of our experts from VNS can
answer this----
Chairman Tauzin. The gentlelady's time has expired, but
she'll be allowed to go ahead and finish that question.
Ms. DeGette. Thank you, Mr. Chairman.
With increased absentee voting, particularly in some of the
Western States, I understand you factor in some absentee
voting. In Colorado in the last cycle, 40 percent of the
ballots were absentee. This doesn't even compare to places like
Washington and Oregon. Now, places like Colorado are going to
mail-in voting, and we are beginning to see some thought of
Internet voting nationwide. How on Earth are you going to be
able to do projections of that when it happens on such a
widescale level?
Thank you, Mr. Chairman.
Mr. Edelman. Well, we've been doing that--telephone polls
in Washington and California and Oregon, and we've done it in
Texas in previous elections, and it appears to be pretty
accurate. And I think I--we'll have to see how it goes. We did
very well in Oregon this year, and that was all mail ballot.
Chairman Tauzin. Thank the gentlelady.
I have a quote I might tell the gentlelady from Tim Worth
in a 1984 hearing. It's rather interesting. He asks the
question, why should someone bother to go to a poll and cast a
vote when the conclusions have already been announced? Here's
his own recollection: I vividly remember voters in west Denver
precincts leaving the polling places in lines in which they
were standing without exercising their franchise as soon as
they heard the results were announced. So it may not be a magic
bullet, but it is certainly worth consideration.
Ms. DeGette. If the gentleman will yield. There's a lot of
issues in urban areas like Denver. Part of it is early
announcement of results; part of it is people having to wait in
line for 2 hours or more to vote, which is part of what we saw
in Florida this last election.
Chairman Tauzin. Thank the gentlelady.
The Chair recognizes Mr. Buyer for a round of questions.
Mr. Buyer. Thank you. Gentlemen, I noted that you didn't
particularly agree with the comments that I made earlier since
most of you referred to them, that we shouldn't look at this
and say the networks made early aggressive calls for Gore
calls, and that there is no evidence that there was bias at the
decision desks. In my home State of Indiana, it's not even
debated whether there's bias among some of the news. We come to
this debate and say, well, we recognize that Fox is more
conservative than some of the others and that jokes are made
about CNN over the last 8 years.
So when I look at the events of election night 2000, and
I've pored through all the testimony and evidence you've
provided, it seems you have an editorial problem. The analysis
by Dr. Edelman is pretty tough on the decision desk. This whole
question about what is an acceptable level of risk when the
decision desk makes the call is a key aspect.
I want to compliment Fox for being very open with the
committee investigators about who were the individuals that
were on its decision desk. Fox revealed their political
affiliations. Others weren't as free with this information. I
can understand why. You say, that's your prerogative, those are
your decisions and stay out of your business, but you invite
this scrutiny. You invite this when such huge mistakes are
made.
I really don't think Congress should be micromanaging in
your business. I welcome your thoughts on it, but I really
don't think we should be. And I really believe that you want to
get it right, because if you don't get it right, you pay a
price to the public.
Let me do two things before I ask a question. There are two
articles that I would like, to be placed in the record. One is
an article by Alicia C. Shepard on How They Blew It, authored
from the American Journalism Review. It's on what happened at
ABC. The other is an article titled A Hard Day's Night, by John
Ellis, a firsthand account on election night in Insight
Magazine. I would offer these both to be introduced.
Chairman Tauzin. Without objection, they're both accepted
into the record.
[The information referred to follows:]
[January 2001/February 2001]
How They Blew It
By Alicia C. Shepard, Senior Writer, American Journalism Review
a behind-the-scenes look at the television networks, dismal performance
on election night.
As votes began streaming into Voter News Service's headquarters
after Florida's 7 p.m. poll closing, it seemed clear to many network
prognosticators that Al Gore was going to clobber George W. Bush in the
Sunshine State. What a story. Florida's governor could not deliver the
votes for his older brother.
But not all experts hired to help the TV networks on election night
thought Florida was a done deal for the vice president 50 minutes after
the polls closed. Some didn't trust the accuracy of Voter News Service
projection models, models which ABC, CBS, NBC and CNN have relied on
for 10 years and which only once had incorrectly projected a major
race.
In the end, the doubters were right. At least three ABC analysts
warned against calling Florida for Gore, but their advice went
unheeded.
It didn't match the VNS models. Besides, raw votes from 120 key
sample precincts and votes trickling in from counties were tracking
with exit poll data collected that day at 45 sample precincts. Exit
poll data and sample precinct votes churned through models that
analyzed past voting patterns, factored in exit poll biases, and
correlated how candidates stacked up against previous contenders.
Around 7:45 p.m., exit poll data, which began coming in at lunchtime,
showed a 6.6 percent lead for Gore over Bush. But election analysts
knew only a fool would call Florida for Gore based on exit poll
information alone.
As votes arrived from sample precincts carefully chosen to
represent voters across the state, the model predicted a 5.4 percent
lead for Gore. It indicated Gore needed a ``critical value'' a
statistical degree of certainty of 2.6 or higher before any network
could comfortably hand the vice president Florida. At 7:50 p.m., the
``critical value'' showed 3.2 for Gore. The Voter News Service model
was more than 99.5 percent sure Gore would carry the state.
Under ``status,'' at 7:50 p.m. the VNS screen said: ``Call.''
In the race to be first, NBC ``won,'' jumping even before VNS at
7:49 p.m.
CBS waited one minute. Warren Mitofsky, who invented exit polls in
1967, has been in the race-calling business for 33 years. He ran CBS'
election unit from 1972 until 1990 and is known for his caution.
Mitofsky, working for CBS and CNN, had vote totals from 12 of 120
sample precincts and data from 38 exit poll precincts. Gore was doing
so well that he concluded exit polls had been overstating Bush's
numbers. ``The real votes were telling us Gore was ahead,'' says
Mitofsky. ``The exit poll data gave us a slight lead for Gore, and the
overlap of the two was telling us that the exit poll data should have
given Gore more support.''
There are three sources of data that VNS uses for its projections.
Exit poll results, the least accurate of the three, come in three times
during the day. They are only used to project winners. Once the polls
close, raw votes from sample precincts are phoned in and measured
against exit poll data. The tally that counts the actual vote total
comes in throughout the evening.
At 7:50 p.m., Mitofsky and his partner, Joe Lenski, confidently
instructed CBS and CNN to call Florida for Gore. Fox News Channel, in
the presidential projection business for only the second time, followed
suit at 7:52, joined by the Associated Press at 7:53 and CNN at 7:55.
``The sad fact is that was a straightforward call,'' says Jonathan
P. Wolman, AP's executive editor. ``VNS' projection material provides a
guidepost that warns you statistically if there's a bias in the
material that might skew the results. In this case, that bias indicator
said it might be underestimating Gore's advantage.''
But not everyone saw things that way.
At VNS' temporary quarters on the 93rd floor of Manhattan's World
Trade Center, two political scientists working for ABC, each with a
strong statistical background, didn't think the Florida result was
clear-cut. Nor did they completely trust the VNS model. When the
decision desk telephoned the two analysts asking, ``Can we make the
call?'' both men advised against awarding Florida to Gore.
Polling places in Florida's Panhandle in the Central Time Zone
wouldn't close for 10 more minutes. Only 237,115 actual votes had been
tabulated in a state with 8.8 million registered voters. But other
factors involving statistical probabilities and VNS models troubled
Kenneth Goldstein, of the University of Wisconsin, and Christopher
Achen, of the University of Michigan, where election surveys were
pioneered in the 1940s.
Achen had flown to New York City six days before the election to
prepare for the big night. He spent four days studying VNS models,
trying to pinpoint why they had screwed up in picking the winner of the
1996 New Hampshire Senate race.
The networks ABC, CBS, CNN and NBC and the AP created VNS in 1993.
The idea was to save money by pooling resources and receiving data
amassed by a single source. Fox joined the consortium in 1996.
For the most part, the setup has worked well. VNS projections have
largely been accurate, to the point that they have virtually been
treated as facts, a state called for one candidate moments after polls
close is seen as decided. But four years ago, the system failed: Each
network, relying on exit polls, prematurely called New Hampshire for
Democratic Senate challenger Dick Swett. When the votes were counted,
it turned out Republican Sen. Robert C. Smith had been re-elected.
``I dug in and went over that as extensively as 30 years of
experience would allow me,'' says Achen. ``I worked out a rule of thumb
to protect myself on election night from it happening again.'' Achen
was so determined to prevent a similar error that ABC colleagues began
referring to him as ``Mr. New Hampshire.''
Looking at the Florida data around 7:30 p.m., Achen noticed the
Gore exit poll numbers were higher than VNS had predicted. This
bothered him because exit polls tend to have a Democratic bias. Plus,
not all exit poll data was in before 8 p.m. And what about absentee
votes? Analysts expected 10 percent of Florida votes to be absentee.
With about 6 million people expected to vote, that's 600,000 votes that
exit polls know nothing about.
``But first and foremost,'' says Achen, ``when I applied my rule of
thumb to protect ABC against mistakes, it indicated it was too soon to
call.''
And so, while other networks were falling all over one another to
declare Gore the victor in Florida, Achen held back. While the VNS
model indicated more than 99 percent certainty for Gore, Achen saw it
as more like 85 percent.
Neither he nor Goldstein advocated making the call. There was too
much at stake to move precipitously. ``It wasn't like we were calling
an off-year dog catcher race in North Dakota,'' Goldstein says.
``Besides, what's the hurry?''
At ABC headquarters on Manhattan's West Side, Paul Freedman was
part of the six-person decision team formed to call the Senate,
gubernatorial and presidential races. Freedman, a University of
Virginia political scientist, also thought it was too early. ``It's
fair to say the three of us wanted to be more confident before making a
call, because we thought there was too much uncertainty in the
estimate,'' says Freedman. Others, he says, were also endorsing
caution.
But they weren't advising in a vacuum.
By 8 p.m., the other networks were flashing Florida for Gore.
Pundits were proclaiming that a Gore win there just might put him in
the White House before the 11 o'clock news. Despite the misgivings of
its experts, ABC's team couldn't resist the competitive pressure, and
ABC decision desk chief Carolyn Smith made the call.
None of the advisers claims to be a white knight. They could have
argued their case more forcefully, but they didn't. They are academics
hired to share their wisdom, not adrenaline-charged journalists
impatient to make a decision.
So, at 8:02 p.m., anchor Peter Jennings joined the pack. ``ABC News
projects that Al Gore wins the state of Florida and its 25 electoral
votes,'' said Jennings. ``Give him the first big state momentum of the
evening. This is the biggest state where the race has been close, the
fourth biggest electoral prize.
As Jennings spoke, Goldstein turned to Achen. ``I think they may
have fallen into the New Hampshire trap,'' he said.
Gore's lead began to shrink within 10 minutes of ABC's call.
``Think how we would have felt if we had really had the courage of
our convictions and if ABC hadn't called it,'' Achen says. ``But we
didn't. The team was wrong, and I'm part of the team. I don't want to
say it wasn't my fault.''
Election night is showtime for the networks. The story is huge and
fast-unfolding, and competition is fierce. With so much on the line,
each network prepares extensively, beginning years before the
presidential vote. They hire experts, spend lavishly on dazzling
graphics, design eye-catching sets, and do more research than a Ph.D.
requires.
The mission is simple: Get it right. During a November rehearsal,
Smith lectured ABC election night personnel for 15 minutes, stressing
that ``we weren't in a race to be first,'' recalls Craig Ammerman, a
former executive editor of the old Philadelphia Bulletin who has worked
every national election for ABC since 1982. ``We were there to get it
right. Especially since the presidential race was so close.''
But the drive to be first is powerful. This is a tricky dance when
all of the networks are getting the same information at the same time
and in the same way. It's particularly dicey in the case of an
exceptionally tight election.
Sharing election data among networks began in 1964. That's when
ABC, NBC and CBS banded together to form News Election Service to
collect poll data and enter it into computers. While they shared vote
counts, each news division ran its own election unit and made its own
calls. Exit polls were used in the 1970s, but only to flesh out voting
patterns and trends. That is, until NBC scooped everyone in 1980 by
calling 11 states based on exit polls. That enabled the network to
declare Ronald Reagan president at 8:15 p.m., while ABC and CBS waited
for more votes to be counted.
``ABC didn't call the election until 10 minutes to 10, right before
Jimmy Carter made his concession speech,'' says Mitofsky, who was then
running CBS' election shop. ``CBS didn't call it until 10:20 p.m.,
after Carter made his concession speech. That was my doing. Yeah, it
was hard. I didn't need any lessons in exit polling from NBC. It
bothered me that they were using them and we weren't.''
After 1980, each network conducted its own exit polls on a massive
scale. But costs mounted and network news executives, far more focused
on the bottom line than in the past, scrambled to save money. In 1990
they created Voter Research Surveys to conduct exit polls, offer
analysis and make projections for all networks. Mitofsky was named to
head the new operation. By joining forces, each network would save $9
million over a four-year period, according to a report cited in David
W. Moore's book, ``The Superpollsters.'' It also made VRS research
affordable for CNN.
Philip Meyer, a pioneer in what is sometimes called ``precision
journalism,'' spoke out against a network consortium at a 1991 meeting
of the American Association for Public Opinion Research, calling it ``a
bad idea.'' If VRS makes a mistake, Meyer said then, there could be
``terrible consequences'' because there would be no other exit polls to
serve as a counterweight. If VRS was wrong, he reasoned, everyone would
be wrong.
``My concern then and now is when you share responsibility like
that across news organizations that normally compete, if everybody's
responsible, then nobody's responsible,'' says Meyer, who teaches
journalism at the University of North Carolina. ``When they were
competing organizations in 1988, they gave different answers, and that
was embarrassing to the networks. The visible conflict is good because
it reminds everybody how delicate these instruments are. Competition
produces better results.''
In 1993, VRS and NES merged to create Voter News Service, an effort
to save even more money. Mitofsky left and Murray Edelman became
editorial director. The terrain shifted that November when ABC broke
from the pack and surprised everyone by making its own calls ahead of
VNS, albeit using VNS data.
Thus began the age where networks artificially compete, using
identical information supplied simultaneously but reaching their own
conclusions on their own timetable.
On Election Day, VNS, payroll, which includes about 30 permanent
employees, swells to about 45,000, including election year staff, data
input operators in the Cincinnati area and New York City, exit poll
interviewers and people who collect county and precinct votes in each
state and the District of Columbia.
In Florida, VNS-trained interviewers conducted scientific exit
polls in 45 precincts with 4,356 people after they voted and staffed
120 sample precincts and 67 county election offices.
VNS ``reporters'' collect vote totals and phone them in to
operators. Statisticians at the World Trade Center crunch the numbers
through a variety of statistical models based on historical and
geographical voting patterns. Then the computer comes up with a
probability that statisticians use to project a winner.
Some critics say VNS models are outdated and not statistically
sound. ``I do think we need a software update on giving estimates of
probability that takes into account changes in the country, especially
on absentees,'' says ABC analyst Achen. ``What we saw this year is a
sign that there's work to do.'' CBS analyst Lenski agrees VNS models
``need to be adjusted to fully account for the intricacies of absentees
and early voters.''
``I'm one of the people who understands statistical models,'' says
analyst Achen. ``On election night, I can sit there as could other
people on the team, and say, `The computer is saying this, but the way
the model works, that's not very trustworthy,' and that's what they
hired us to do. The model makes certain assumptions that are true most
of the time, but not always. The machine can say it's 99 percent likely
and it really is, or it says it's 99 percent likely and that's not
true. That's what happened at 8 with the Gore call.''
Once a race is called, network election analysts put it aside and
turn to other contests. Too much is happening to double-check a called
race. Yet that's just what Goldstein did 15 minutes after ABC awarded
Florida to Gore. Raw votes from Florida's 67 counties were pushing Bush
ahead.
Goldstein happened to be watching the Florida Senate race screen
when results from Duval County in the Jacksonville area came in. The
data was odd. It showed Republican Rep. Bill McCollum gaining on
Democrat Bill Nelson in the race for the Senate. Goldstein switched to
the presidential screen. Gore was surging.
``That didn't make sense,'' says Goldstein. ``A Republican closed
the gap in the Senate and a Democrat widens in the presidential race.
That tells you there's a serious problem. People don't vote like
that.''
CBS and CNN analysts Mitofsky and Lenski lost confidence in their
Florida projection at about 9:15 p.m. ``After we made the call, I was
fine with it for the next hour,'' recalls Mitofsky. ``Then it started
getting suspicious.''
Their decision screens showed numbers that didn't jibe for the
northern Florida region. ``We looked at all the counties in north
Florida and saw Gore was getting 98 percent in Duval and Bush only
getting 2 percent,'' says Lenski, executive vice president of Edison
Media Research. ``We called Murray.''
At 9:07 p.m., a VNS operator had accidentally added an extra digit,
pushing Gore's total in Duval County to 43,023 instead of the actual
number, 4,302. The mistake was corrected by 9:38 p.m.
At 9:54 p.m., after watching the suspicious numbers for almost 45
minutes, CBS stripped Gore of his win and sent the race into the
undecided column.
At 10:16 p.m., VNS sent a message to its members and 100 TV, radio
and newspaper subscribers: ``WE'RE RETRACTING OUR CALL IN FL BECAUSE WE
DON'T HAVE OUR PREVIOUS CONFIDENCE.'' By then, everyone but NBC had
already pulled Florida back.
``After VNS deleted the bad data,'' says Lenski, who's worked every
election for CBS or VNS since 1988, ``we realized it was dead even.''
They advised CBS polling chief Kathleen A. Frankovic to take Florida
from Gore.
Word traveled from CBS News Executive Producer Al Ortiz into Dan
Rather's earpiece. ``Bulletin,'' sputtered Rather. ``Florida pulled
back into the undecided column. Computer and data problem. One of the
CBS News election night headlines of the hour. This knock-down-drag-out
battle drags on into the night, and turn the lights down, the party
just got wilder.''
Mitofsky hasn't made many wrong calls in 33 years. ``They are
embarrassing,'' he admits. ``I'm chagrined by them. But if you're
wrong, you're wrong.''
But he had company: Everyone using VNS data had jumped the gun. It
may have been a tad more embarrassing, though, for Mitofsky, since
Rather had earlier assured viewers that they had settled on the
reliable channel.
``Let's get one thing straight right from the get-go,'' said Rather
at 7 p.m. ``We would rather be last in reporting a return than to be
wrong. And again, our record demonstrates that to be true. If you hear
someplace else that somebody's carried a state and if you are off, as
you shouldn't be, watching them, then come back here because if we say
somebody's carried a state, you can pretty much take it to the bank,
book it, that that's true.''
Fifty minutes later, the promise wasn't worth much.
So what, exactly, had gone wrong?
The bad Gore call was not because of flaws in the exit polls or a
data entry error, despite dozens of inaccurate media reports to that
effect. Experts agree there was no bias in the exit polls, as there had
been in New Hampshire. And the Duval County mistake, made an hour after
the Gore call, played no part in the blunder, VNS Editorial Director
Murray Edelman explained in a confidential November 14 memo to members.
Nor were other errors to blame, such as one made by a VNS staffer
who inaccurately recorded figures for Lake County at 9:01 p.m. and
again at 10:47 p.m., coming up with totals larger than was possible. By
11:59 p.m., the errors were corrected.
Nor was it due to a VNS operator shortchanging Gore by 4,000 votes
in Brevard County, punching in 93,318 instead of 97,318 at 10:13 p.m.,
though that error may have played a role later in the evening since it
wasn't corrected until 3:51 a.m, according to a VNS memo.
``I still believe the biggest problem in the model is that we did
not correctly anticipate the impact of the absentee vote,'' Edelman
wrote in the memo.
Edelman declined to be interviewed by AJR for this article. The
network consortium will not allow any VNS employees to speak to the
media about what happened on election night. And network officials also
will not discuss the situation, beyond saying that they are
investigating what went wrong.
University of Wisconsin political scientist Ken Goldstein is not
surprised that absentee ballots played a key role in a bum call. Exit
polls, while painting a portrait of Florida voters who cast ballots at
precincts, tell analysts nothing about absentee voters. ``One of the
things you are looking for in a close race is you want to be sure you
know what's been going on with the absentee ballots and that you are
counting them properly,'' says ABC election analyst Paul Freedman.
But Goldstein and others say Edelman was cognizant of the absentee
factor going in.
``If you don't include absentee votes in your model, you are going
to be off,'' says Goldstein. ``Everyone knew that Florida is 10 percent
absentee and has always been. We got memos ahead of time that said,
`Don't forget about absentee.' Murray produced a lot of paper before
the election on absentee voters. We had Excel spreadsheets on
absentees, plus we had reports for each state on absentee history.
Yeah, the absentees could be the explanation, but you knew about
that.''
Academics, statisticians and news people working election night
know uncertainty goes hand-in-hand with predictions. They expect and
try to compensate for bad data or human error. Making projections is a
science. But it is not foolproof.
What does ABC do when everyone else has picked a winner and ABC
hasn't? ``There's got to be social pressure,'' says the University of
North Carolina's Meyer. ``Who wants to watch an anchor who doesn't know
who won when everybody else says they know?''
Says Freedman, ``As long as the context is calling the winner
first, you are going to see built-in incentives to be a little risky.''
Given the intensely competitive environment, it's reasonable ``to
expect that people will make decisions with competition in mind and not
simply with the data in mind,'' he adds. ``It's inevitable. It's built
into the way the networks structure election night.''
But why are the networks in the projection business at all? who
cares which network is first? Most people merely want accurate results.
many critics among them election expert Curtis Gans argue the networks
do democracy a disservice by declaring a winner before the final
results are in. ``In almost every election there's this rush to
judgment,'' says Gans, vice president and director of the Committee for
the Study of the American Electorate, ``and there's something
inaccurate reported. Networks are creating the news by projecting
winners, not reporting it. No data is as accurate as tabulated
results.''
A poll by the Pew Research Center For The People & The Press
reported nine days after the election that 87 percent of the American
people want the networks to stop predicting winners before the votes
have been tallied. ``There are a lot of people out there who are fed up
with the networks,'' Gans says.
What happened in the wee hours of the morning after changed perhaps
forever, but certainly for many years to come how much Americans trust
television networks. What will they believe in 2004 when a network
projects a winner in Florida?
``What everybody is going to remember is not the campaign coverage
but the election night coverage,'' says S. Robert Lichter, president of
the Center for Media and Public Affairs. ``So even if we could give
[the networks] a brilliant grade, I think it would be a little like
saying that the Titanic was doing just fine except for the iceberg.''
Almost 41 million households were tuned to the four major broadcast
networks and three cable news channels on election night, according to
Nielsen Media Research. No waiting for tomorrow's paper. No competing
to get on a crowded Web site. And TV graphics beat radio. Turn on the
TV and Jennings and Rather and Tom Brokaw and the cable guys are there
for reliable updates,
Reliable?
By 1 a.m., it was clear that whoever carried Florida would move
into the White House in January 2001. As predicted, election night 2000
had been a wild ride. Now, all that had to happen was for someone to
win Florida. Then all of those viewers struggling to keep their eyes
open could finally go to bed.
At 1:30 a.m., everyone staring at election data screens at the
networks, at the AP, at VNS was jumpy. Ninety-five percent of Florida's
precincts had been tabulated, according to VNS, and there was still no
clear winner. At 1:30 a.m., Bush had about a 60,000-vote lead, says AP
Florida Bureau Chief Kevin Walsh. But how many votes were out? Where
were they concentrated?
At 1:52 a.m., Walsh's figures showed the lead shrinking to 56,000
Bush. ``It was very intense,'' recalls Walsh. ``We were getting calls
here and in Tallahassee from editors all over the country wondering if
we were going to call the race, or looking for guidance.''
At 2:10 a.m. in Fox's New York studio, election analyst John Ellis,
George W. Bush's first cousin, saw the same VNS data as those at the
other networks. It indicated Bush held a 51,433-vote edge. Only 179,713
votes were outstanding, according to VNS. (It would turn out that VNS
had grossly underestimated that number. Instead, twice as many votes
remained uncounted, according to a VNS post-election memo.) To close
the gap, Gore needed to win 63 percent of them. It seemed an impossible
task. Ellis advised the network to call the state for his relative.
NBC decision desk head Sheldon Gawiser was worried, too. On the
phone with Edelman, Gawiser repeatedly asked why NBC couldn't make the
call. ``When he couldn't give me any reason not to call the election,''
says Gawiser, ``then I told him I was going to go ahead and take a look
at it myself. And we then broke the connection.''
Fox anchor Brit Hume declared George W. Bush winner of Florida's 25
electoral votes and crowned him the next president at 2:16 a.m. One
after the other, the networks tumbled, like so many dominoes, CBS and
NBC at 2:17 a.m., CNN at 2:18 a.m.
Before 2:00 a.m., the ABC decision desk asked Achen about making
the call. Again, he said it was too soon. The VNS model couldn't assure
him the absentee votes had been counted, and he knew that at the end of
a long night, errors in totals were likely. But Achen says he didn't
lobby strenuously against the decision.
At 2:20 a.m. again last, ABC gave Florida and the election to Bush.
Ironically, VNS itself did not award the state. But CBS, and CNN's
Mitofsky says with 97 percent of the vote reported, Bush leading by
50,000 and about 180,000 votes not yet tallied, he thought it was an
easy call.
But the AP hesitated. At 2:16 a.m., with 99 percent of the state's
5,884 precincts in, Walsh's count showed the Bush margin narrowing to
30,000 votes. ``We didn't feel there was any way we could make that
call,'' says Sandy Johnson, AP's Washington bureau chief and the one
making final decisions on all of the races. ``It just wasn't there.''
As CNN anchor Bernard Shaw declared Bush the winner at 2:18 a.m.,
Republican operatives projected Shaw on a drive-in movie-size screen
outside of the Texas State Capitol. The Republican crowd in Austin
erupted in cheers. In Nashville, Gore supporters standing in the rain
had another reason to feel glum. Soon, they were told, their man would
arrive to give a concession speech.
CBS' Rather was awash in Ratherisms. ``Let's give a tip of the
Stetson to the loser, Vice President Al Gore, and at the same time, a
big tip and a hip, hip, hurrah and a great big Texas howdy to the new
president of the United States. Sip it. Savor it. Cup it. Photostat it.
Underline it in red. Press it in a book. Put it in an album. Hang it on
the wall. George Bush is the next president of the United States.''
CBS' screen flashed: ``Bush elected president.''
Rather and CBS stars Ed Bradley, Lesley Stahl and Bob Schieffer
began speculating on why Gore lost, what he would do next, whether he
could have run a better race.
Still the AP remained silent.
Across the country, newspapers were bumping up against their final
deadlines. Editors were edgy. Dozens of the AP's 1,500 daily newspaper
members began calling New York, Washington, Miami. Why hasn't the AP
called the race? In New York City, the AP's president and CEO, Louis D.
Boccardi, called his D.C. team at 2:30 a.m. He, too, wanted to know
what was going on. So did Executive Editor Wolman, who was in D.C.,
although Bureau Chief Johnson was honchoing election coverage. ``I had
to wonder if there was an overabundance of caution because of the bad
Gore call,'' Wolman says.
Yes, the AP team was wary. It couldn't afford to blow it, not with
all of those newspapers depending on it. But Johnson had confidence in
an invaluable tool the AP had that the networks didn't: a backup
system. Long before the AP joined VNS in 1990, it had cobbled together
its own network for gathering election results.
``We count the vote in every state,'' says Tom Jory, the AP's
liaison to VNS. ``We count the same national races VNS counts as an
independent backup to VNS and use VNS to edit-check our report. We
often use VNS as a primary source, but we weren't using VNS in
Florida.''
Like VNS, the AP hires freelancers to gather vote counts. The
freelancers are strongly encouraged during dress rehearsals the weekend
before election night to call bureaus early and often. ``As a result,''
says Walsh, ``we were consistently out front of VNS and out front of
the Florida secretary of state's Web site.''
When VNS models showed Bush's lead jump from 29,386 at 2:05 a.m. to
51,433 five minutes later, network analysts thought the race was over.
But AP's count was radically different. It showed that Bush's lead at
1:47 a.m. was 56,486. By 2:16, it had plummeted to 30,513.
The AP decided it wasn't ready to call the race. At about 2:30
a.m., seeking reassurance, AP decision editors in D.C. summoned Will
Lester, who for 11 years had participated in the wire service's
coverage of Florida elections as a writer and an editor.
``You didn't have to sit down and do fancy math to figure out that
if Gore could win a substantial share of the outstanding votes, it was
clear that the margin Bush had could evaporate,'' says Lester, the AP's
poll writer. ``Broward and Palm Beach are big metropolitan counties.
Those counties coming in late definitely could wipe out a Republican
lead.''
``Can we call it?'' Johnson asked Lester. ``We need to know when we
can call it.'' Lester eyed the data again. ``You can't do it,'' he
responded. At 2:30 a.m., as Rather deconstructed Gore's downfall, the
AP's numbers showed Bush's edge dropping to 19,000.
Wolman was feeling enormous pressure. ``After the celebration
started in Austin,'' says the AP executive, ``we spent all our time
crashing numbers on calculators, trying to understand whether the race
was over or was it possible for the Democrats to catch up?''
At 2:37 a.m., Johnson and others concluded Gore could pull it out.
The wire service issued an urgent update, cautioning that while the
networks were calling the race, there was ``the narrowest of margins''
between the candidates with votes still being tallied.
But everyone at the AP's Washington bureau was extremely nervous.
``You have the weight of the other VNS members making the call,'' says
Wolman. ``By that time newspaper Web sites were printing `Bush Wins.'
Gore had conceded to Bush. There was doom and gloom in Nashville. We
felt extremely lonely. We were thinking, `Florida's the whole ball
game. Don't blow it.' ''
Many were thinking the same thing. ``That would be something if the
networks blow it twice in one night,'' said NBC's Tom Brokaw less than
an hour before the network took Florida away from Bush at 4:00 a.m.
They did.
Could the debacle have been avoided if the networks had subscribed
to the AP's special election-night wire, which flashed vote totals up
to 12 times an hour, as much as four times as frequently as the main
wire? None of them did.
``I wish we had,'' Lenski says. ``At 3 a.m., we went into Yahoo! to
find AP vote totals.''
Some network analysts explain the second blunder by blaming a
computer glitch in one machine in one tiny precinct in Volusia County,
which includes Daytona Beach.
In precinct 216 in DeLand, election workers began having trouble
with one voting machine shortly before the polls closed. After 7 p.m.,
a poll worker drove the malfunctioning laptop-size machine to the
Department of Elections to see if someone could get it to work.
``We would have had to remove the memory card and upload it
ourselves, which is what we did at 10:02 p.m.,'' recalls Denise Hansen,
the department's assistant supervisor.
Within 15 minutes, she says, the county attorney called. He'd been
watching results stream in on a screen in the county's council
chambers. Gore's vote was going backwards!
By 10:30 p.m., the county knew there was an error. But where?
Officials tried for hours to track down the AccuVote machine vendor in
Texas. After running a precinct-by-precinct printout at 1:24 a.m., the
error jumped out. ``You don't see a negative 16,000 votes for
anybody,'' says Hansen. Not to mention a five-digit vote in a precinct
with 585 registered voters only 219 of whom had voted.
When VNS entered the Volusia glitch at 2:08 a.m., Gore's count in
the county dropped from 82,619 to 72,152. The problem should have been
caught by a VNS operator. Its computers flag anomalies, alerting
operators to potential data problems that need to be checked.
``When there's a big decrease in a candidate's vote of 10,000, it
has to be approved by a manager at VNS,'' says Lenski. He adds that if
they'd received ``weird data'' from Volusia, they would have taken a
close look at the county's totals.
Says the AP's Walsh, ``I think everyone was affected by that
erroneous data. However, our vote totals in all the counties were far
ahead enough of VNS, that it affected us less. We were recording votes
so fast in the Miami bureau that the Volusia problem did not have as
dramatic an impact on the overall Bush margin for us as it did for
VNS.''
AP didn't notify VNS because while it could tell something was
wrong in Volusia, ``we had no idea what the specific vote drop was,''
Walsh says.
What looked like a 56,000-vote Bush lead according to VNS was
probably more like 30,000, says CBS' Mitofsky. ``Now we always expected
his lead to close up somewhat, because the missing votes were in
Democratic areas. We certainly didn't think there were more than
180,000 votes out. If we had known that there were as many as 360,000
votes out, as we do now, we never would have made a projection.'' It's
not clear what caused that miscalculation.
At 2:30 a.m., the AP's count gave Bush a 19,000-vote margin while
VNS, margin was almost twice that. To the AP, with votes in Democratic
strongholds outstanding, it appeared Gore could pull ahead. To those
eyeing VNS data, Gore didn't stand a chance.
At 3:11 a.m., the AP sent out a cautionary advisory. Bush's lead
had dwindled to 6,000.
On CBS, Ed Bradley was holding up the AP report. Rather turned and
Bradley said slowly, ``The Associated Press believes that the uncounted
votes in Broward and Palm Beach counties could allow a change of the
lead in the Florida vote.'' There was laughter, groaning, amazement on
the set. ``Hello, 911? Cardiac arrest unit please,'' said Rather,
pretending to make a phone call. Bob Schieffer covered his face in
disbelief.
Rather recovered. ``Let's not joke about it folks,'' he advised.
``You have known all night long and we've said to you all night long
that these estimates of who wins and who loses are based on the best
available information we have. CBS News has the best track record in
the business, over a half century plus, for accuracy on election night.
But nobody's perfect.''
Congress is holding hearings on the election night fiasco. Networks
are hiring outside consultants to study the VNS system. Already, ABC,
NBC and Fox have promised they won't call a state until all of its
polls have closed. The networks are once again lobbying for a uniform
poll-closing time. ABC has announced it will no longer allow television
sets in decision desk rooms. Critics are beseeching the networks to
stop projecting winners based on exit polls or sample precincts.
After all of the investigations are over, and Congress has studied
the misfire into oblivion, and network executives have pounded their
chests with mea culpas, and anchors have promised to never, ever
project winners in tight races, and television sets are forever banned
from election night decision rooms, the best solution may be simply to
force everyone involved to repeat this mantra 100 times:
``It's not an off-year dog catcher race in North Dakota.''
Mr. Buyer. I would like to ask each of the networks about
this issue of acceptable levels of risk, and whether you agree
with Dr. Edelman's comments on that; and second, as you are
doing your review, whether you're going to keep the present
setup that you have about the decision desk and how you make
those decisions.
I can almost see it. All these people at the different
decision desks at the networks know each other, right? Some are
friends, some are colleagues, some aren't friends, but they
know each other, and they're all competitive, and they've been
that way for decades. I can almost see the camaraderie. I can't
believe so and so made that call. He's going to fall on his
face. I'll bet you 10 bucks. I can see that stuff going back
and forth through the night. Everyone wants to get ahead of the
other. And they lay the networks credibility on the line in
something that is a big game to them. But they also want to be
right because they want to be hired back.
But I'm curious about whether you're doing internal reviews
on whether you restructure that desk or not. So the two
questions are do you agree with the doctor's assessment about
that level of risk; and second, are you going to be
restructuring the decisionmaking desk to be more accurate?
Thank you.
Mr. Johnson. Congressman, I'd like to say first that I take
strong exception to your comments, to your jokes about CNN. I
see we live in a word where jokes are made, Members of
Congress, even the President, and I think both the Pew study
and the NBC/Wall Street Journal study showing CNN to be the
most trusted network is something that we take great pride in.
Mr. Buyer. Was that a paid advertisement? That was for
free, wasn't it?
Mr. Johnson. I think everybody has said today that we are
determined to make certain that we take the steps that are
necessary to assure that our future conduct on election night
is as solid, as credible as we know how to make it. That cuts
across no matter what area here we may represent in terms of
where all the networks stand on a political spectrum.
Mr. Buyer. You will continue to team with CBS on how you do
your decisions at the desk?
Mr. Johnson. We may, or we may not. We have not made that
decision. We have tremendous respect for CBS. We also have
great respect for the other news organizations here. I think
you'll find that each of us does our job in the way we think
best serves the public.
Mr. Buyer. Did you agree with Dr. Edelman's assessments on
the issues on risk?
Mr. Johnson. I think that we want to take no risk that that
risks our credibility.
Mr. Buyer. That's fair.
Chairman, would you permit each of them to answer the
question?
Chairman Tauzin. The gentleman's time has expired, but our
rule is if you would like to answer, you certainly can.
Mr. Lack. There is no acceptable risk to be wrong. You
don't want to go out there and report something that just might
be wrong.
I think it's important to point out it hasn't been
mentioned in this discussion today that 75 percent of the races
are not close at all, and there isn't any risk involved. Then
you get down to that last 25 percent, you're starting to ask
yourself, what's the risk here?
I think the answer is we don't want to take any risk
calling a State where we're wrong.
I guess the gentleman representing VNS today Mr. Savaglio
pointed out that they have been right 99.8 percent of the time,
but they blew a doozy, and that's not acceptable to any of us.
We're not going to take the risk that that is going to happen
again.
With respect to the way NBC News is going to look at its
decision desk, we are going to take a look at it. I don't know
what or how it's going to reconfigure itself between 2000 and
2004, but we're not going to play with a pat hand. We're going
to tear it apart and tear up the rug and start all over again
and see if we've got the right people in the right places doing
the right things.
Mr. Buyer. Thank you.
Mr. Lack.
Mr. Ailes. I think as far as the desk at Fox is concerned,
I am certain that we will review that between now and when we
have another election. What we will keep, however, is that we
had a situation where every single person on that desk had to
give a yes vote before it went to air, and then it when through
a vice president, senior vice president for the network, who
recanvassed the floor and said, are you sure of this, should we
make this decision. So we had a unanimous desk. The last thing
I said to the people at 5:30 that evening after a briefing was
nobody will get fired here for being late. I don't care if you
call every single race last. I want you to be fair, and I want
you to be as accurate as you can be. That was my instruction at
5:30 to 35 of our journalists who were meeting in a conference
room.
And I think that there is this competition and so on, but I
do think the idea of having everybody in the room agree it's
time to go is helpful.
Mr. Buyer. Thank you.
Chairman Tauzin. Mr. Heyward.
Mr. Heyward. Very briefly. We already said in our report
that we're going to toughen up the oversight of the decision
desk. We have already announced some change. We're going to
have a new executive who over sees it and will serve as a
potential check or balance on decisions that might be
premature, let's say. But, you know, in terms of the risk,
again, I think one of things we haven't really said much at
this hearing, the penalty for us for a mistake is extremely
severe. I think that also speaks to some of the concerns about
perceived bias, or, you know, I think one of the themes of
election night is just waiting until you can be sure to get it
right. And clearly we have to rely on some better models than
the ones we've had.
The--our decisions are there for the world to see, and the
penalty when we're wrong, in addition to having the spend the
day here, which is fine, is that our credibility with our
public is impaired, and ultimately our very business would be
threatened. If people couldn't count on us to be accurate, we
would be out of business. So we have a tremendous incentive
independent of any of the issues that's being raised by you,
Congressman Buyer, or any of the other Members, tremendous
incentive not to let this ever happen again, tremendous, built
into the marketplace.
Mr. Buyer. Mr. Westin?
Mr. Westin. Yes, sir. On the two questions, first
acceptable level of risk, I said in my opening statement, and
I've tried to reiterate, that we at ABC News, as my colleagues
do, do everything possible to ensure the accuracy and the
timeliness of what we report.
Now, I want to be honest with the committee, out of respect
to the committee. To manage for zero risk in journalism is not
to be a journalist; it is to be a historian. The only way to
manage for zero risk, as we learned in this last election, the
Presidential election, is wait until after the electoral
college has finally voted. So we are going to do everything we
can.
The problem we had here was there were risks in this system
we were not aware of. We thought we were managing for closer
tolerance and risk than, in fact, we were, and that's what we
have to fix and get rid of.
On restructuring the desk, we said 2 weeks after the
election, we want to remove as much as is feasible our decision
desk from some of the pressures of watching monitors and see
other people who can do it other ways, but it is human nature
that when you are seeing everyone else with graphics up
announcing the 43rd President of the United States, that you
start to think maybe you should get moving.
That's a fundamental change that we are going to make. But
I also have to just come back on the end on the allegation of
bias, of political bias. You don't know the political
affiliation of anybody working on a decision desk because I
don't know, and if I did know, they wouldn't be working on my
decision desk.
It is a policy at ABC News, you do not take a political
position. Now, you may have it in the privacy of your home, you
may vote it, but you better not express it in any way that I
know about it. And that's why you don't see it.
It is, frankly, unfair to the men and women that I work
with to suggest that they come into the office every day and
bandy about their political views, because it is simply not the
way it works. We may get it wrong, and we do, and we get called
to task for it, as we should, but these are hard-working,
dedicated men and women who are trying to get it right the best
they can.
Mr. Boccardi. I would like to make the same point about the
people who work for me. I don't know their politics, and I
don't care about it, but on this question of risk,
philosophically David is right, we don't work in a zero risk
business. But there are none of us who would say, well, let's
make that call; there is a risk it's wrong, but let's do it. We
just don't do that. When we make a mistake, we make a mistake,
as in the Florida early Gore call in Florida by us. We felt
that it was true.
Chairman Tauzin. The gentleman's time has expired. The
Chair will now recognize a gentleman who, believe it or not,
has sat in this chair and held longer hearings than I have, the
gentleman from Michigan Mr. Dingell.
Mr. Dingell. Mr. Chairman, I thank you.
Gentlemen, you have been here a long time. I would note
that most everything that can be said has been said, but not
everybody has said it.
I would observe that the last few minutes of the hearing, I
think, have been very helpful.
Mr. Chairman, I do want to commend you for the way this
hearing has been conducted.
I would note that CNN, Mr. Ailes has come forward with some
rather scathing criticism of their own performance, and they
have suggested a number of changes that they think they should
make. I note that others here have come forward with similar
statements about needs for changes.
First of all, what then, gentlemen, do we do so that we
don't have another situation like this? It was, I think, pretty
important that we have fair, accurate reporting, reporting here
that didn't affect or impact one of the great and important
elections in the history of this country.
Starting with you, Mr. Lack, what do we do about these
matters?
Mr. Lack. Well, I think we have all generally agreed that
we need to reevaluate how VNS goes about collecting their data
and providing it for us on election night. So we need to
retool, fix, use whatever verb you like, to get VNS up to speed
so that the mistakes that they did encounter on election night
won't occur again in 2004.
In addition to that, we need all of us individually be
satisfied that our own organizations have backup systems and
enough original reporting to help them evaluate the course of
events as election night unfolds.
Mr. Dingell. Does that include the contract that you have
with your contractor to do certain work for all of the
networks? Does that need any work, additional expenditures?
Mr. Lack. I believe it will take additional expenditures,
and NBC and many of my distinguished colleagues here have also
spoken straightforwardly about that. It is going to take more
money, and we are going to pony up.
Mr. Dingell. Mr. Ailes, what do you have to say to the
question?
Mr. Ailes. I think everybody on this panel is probably
going to pay very close attention to this personally. I think
it will cost more money. We all know that. I think we will be
involved in frequent meetings to try to solve this problem.
I think that beyond that we have an obligation to inform
the public a little better about the election; put disclaimers
on the screen about what they are seeing. If they are seeing
something and they could think it is something else, an exit
poll is an exit poll. It may or may not be accurate. You should
never base your vote on it.
I think we have to be careful about oversimplification of
how to win the race or not win the race. I think it has caused
a real review of our process internally about information that
we give to the public to make their judgments.
Beyond that, I think there are a lot of things, a review of
the desks and so on and many of the suggestions, which I won't
go back through here. We all came to pretty much the same
conclusions, but it is clear we are all going to have to
rebudget.
Mr. Dingell. Mr. Johnson, what do you suggest? We have
talked about software and modeling and things of that kind.
What are your comments?
Mr. Johnson. First, that we fix VNS. Second, at least in
our organization, that we get a second source of data to make
projections. Third, that we be very careful about the use of
exit polls in the close races. As we have suggested, I really
believe that we will withhold calls when the vote margin is
extremely close, make it clear we cannot make that call, and
also, as we have recommended, a uniform poll closing act.
Congressman, I also believe, though, as we have said
earlier today, that there is almost an equally important need
for us to upgrade the technology throughout the United States.
I mean, in an era where we all can go to an ATM machine and
reliably dial in a code and take out money, we should all be
able to go and reliably be able to make sure that our votes
count.
Mr. Dingell. Sure can't count on those punch card machines,
can you, or the butterfly ballots, either one?
Mr. Johnson. We need to change those soon.
Mr. Dingell. Mr. Heyward.
Mr. Heyward. Congressman Dingell, rather than just restate
our list of recommendations, they are very similar to what my
colleagues have said, I would just add that maybe as an overall
theme one benefit that the public will get out of this, in
addition to, I think, more reliable data and perhaps some
commendable caution when the races are close, is that the
process will be demystified a little bit. I think that the
notion that there is something mysterious about it is not a
good thing, and I think the degree to which we report on why we
make calls, why we don't make calls, the nature of exit
polling, the nature of our projections, how votes are going in
particular States and not try to have this air of omniscience
will be a good thing. I think that's one of the lessons that
comes out of 2000 is that accuracy, clarity, directness with
the viewers are very important to us, and I think that the
citizens will benefit from that next time around.
Mr. Dingell. Mr. Westin.
Mr. Westin. Four things. We need to redo VNS and put more
money and more resources into it, correct some of the errors
that we didn't fully appreciate and know that were in there.
Second, we have to be open to other alternatives to VNS if
they will make our reporting more accurate and more timely. If
those are available, we should take advantage of them.
Third, uniform poll closing we have said we are in favor
of, and I wholeheartedly endorse what Mr. Heyward said. As I
said earlier, I personally look forward to a day when the
technology is such that we get an instantaneous and perfectly
accurate count of the actual vote, and we don't have to make
projections at all anymore. That would be best for the country
and ultimately best for the news divisions.
And finally, and I think as important as anything else, I
think that we as news divisions, and certainly speaking for ABC
News, need to have a bit more humility in our reporting. Now,
that doesn't mean that we are less vigorous. I don't think any
of us want a less vigorous press, but in our vigor we have to
be explaining to our audience what we know and what we don't
know so that they can then incorporate that and make use of
that information in a more constructive way.
Mr. Dingell. Mr. Boccardi, you have expressed some distaste
for the hearing and some questions of pride of this kind of
event. Why don't you tell us what you folks in the media need
to do.
Mr. Boccardi. Well, I think a lot of good things have been
said in the last couple of minutes, and it may sound like an
odd thing for me to say at the end of a long day devoted to the
problem of the 2000 election, but we don't think AP had a
broken election. We had one problem. It was the early Gore
call, and it was made because of some faulty data we received
from VNS, of which we are one of the owners. So fixing VNS is
an important part of what we think needs to be done.
Mr. Dingell. That includes software and money and modeling
and all that sort of thing; does it not?
Mr. Boccardi. Yes. It needs some or all of the above.
Mr. Dingell. Mr. Edelman, do you have some comment?
Mr. Boccardi. Just say yes.
Mr. Edelman. Well, I agree with everything they have said,
and I look forward to having a much larger budget.
I would like to make a couple of comments, that the models
that you were evaluating in the earlier part, you were taking
one number off of the screen and didn't take into account that
there was a lot of other information available.
The issues of absentees and bias and stuff we have been
studying very seriously for these very many years. There was
just not one number fixed to put into the system, so I chose to
give the information and make that available. So when you used
that number and then talked about the systematic bias and all
of that, I think it is quite misleading. That doesn't mean we
don't have a lot of problems, and it doesn't mean that we don't
have a lot of things to make much better.
Mr. Dingell. Mr. Chairman, I am over the time, but may I
ask Mr. Savaglio to comment on this?
Chairman Tauzin. Of course, Mr. Dingell.
Proceed.
Mr. Savaglio. Thank you, Congressman.
As we outlined in our statement, we have identified a
series of improvements in the system that include reworking the
models, using larger sample sizes and working to improve the
accuracy of the exit poll with regard to the response rate, and
also the issue of absentees. So I think that the process has
given us a number of things that we need to do that we know
will make it better, particularly the things that were in the
Research Triangle Institute report, which will help us have
greater accuracy in the future.
Mr. Dingell. Mr. Chairman, thank you.
Gentleman, thank you. I hope we don't see you here after
the next Presidential election.
Chairman Tauzin. Thank you, Mr. Dingell.
The gentleman, Mr. Walden is recognized.
Mr. Walden. Thank you very much, Mr. Chairman.
First I want to thank our panelists for their internal
analyses and external as conducted. I thought they were very
thorough, very complete, had excellent recommendations.
Your decisions to withhold the numbers until all the polls
have closed in the State is appreciated, especially for
somebody from the West.
I am sorry and disappointed that some of you seem to be
offended at having to come here today and testify on these
issues. Mr. Boccardi, maybe you won't take offense to this, but
I have been a dues-paying AP co-op member for 15 years in my
business, so maybe you can look at me in that respect.
Mr. Boccardi. That would never offend me.
Mr. Walden. Good. I will keep paying the dues.
But I think the issue for me is that at the end of the day
we have the closest election in decades. The outcome exposed
serious flaws in the election process as elections are
conducted, we have heard a lot of different issues today, and
how those elections get covered through the VNS formula model
and the problems there.
The States don't like us prying into how they run
elections, as they are guaranteed the opportunity under the
Constitution, any more than you like us prying into how you
cover those elections. But not once today have I heard of any
legislative intent to influence VNS through legislation or to
undermine in any way your first amendment rights to conduct the
election.
Gentlemen, what we run into, as we go home to our district,
are constituents who never have access to people like you to
ask the questions we are asking today. So, please, don't take
offense to being here and sharing the answers, because I think,
frankly, the answers you have shared are very valuable to your
own credibility as well as our own. So I think that it is
important to have these oversight hearings to get to the bottom
of the issue. I think in the end it will strengthen the first
amendment, frankly.
I do have a question, Dr. Edelman, referring to your
testimony, your report, with all members--if I could read from
it, Mr. Chairman--with all members using the same data, one
would expect that all members would call a given race fairly
close to the same time, say within a few minutes of each other.
However, an analysis of the calls in this election reveals gaps
of as much as an hour between the times the same race was
called.
It is very unlikely that a difference of that amount of
time could exist between calls unless some members are assuming
a much larger level of risk than others. If the first member
who calls a race makes the call with some insight into the
data, that insight becomes immediately transparent to the other
members as soon as the call is made, and if the level of risk
is acceptable, the other members would quickly follow suit. It
would appear that calls are being made at the minimum
acceptable tolerances for risk, with very little allowance for
error.
Can you explain further what you meant by that statement,
Dr. Edelman?
Mr. Edelman. Yes, sure. If a member calls a race at 8, and
another one calls it at 9, the one at 9 has a lot more
information than the one at 8. So that member at 9 is taking
less of a risk when they are doing it.
However, a member at 8 is making their determination based
on all of the information that they have at that time, and they
could easily perceive this as a very safe thing. So I made that
statement--so that doesn't necessarily mean that that person,
that member at 8, is taking a bigger--not aware that they are
taking a bigger risk. It is only in retrospect that they might
be. And I made that statement to point out that we should look
at those gaps in time to see if it is based on more information
that someone has or a different interpretation of the data.
Mr. Walden. And as you analyzed the various calls that were
made and the times they were made, I am assuming you did that,
is that correct, a postelection analysis; did you look at when
the different networks made the different calls?
Mr. Edelman. No, I didn't really do that. I had my hands
full, as you may have noticed.
Mr. Walden. Okay. Haven't we all. Thank you.
I have one other question for VNS. It seemed to me, going
into this Election Day, there was this agreement not to--among
the networks not to reveal the results of the data prior to
polls closing. But wasn't there an Internet company that--it
may be a subscriber to VNS--that said, we are going to ignore
that agreement?
Mr. Savaglio. No. There had been some issues with Internet
organizations through the primaries where some of the early--
what were reported to be early numbers were published on the
Internet, but there was no source available on the Internet of
the VNS data.
Mr. Walden. Okay. So nobody in general got access, no
member--could a member have--do members have access to the
initial wave data?
Mr. Savaglio. The system goes up to the members in the
afternoon, yes.
Mr. Walden. And while you have agreement among the networks
here not to release the data ahead of any poll--well, before
the polls have all closed in a given State, do you have that
same agreement with the other recipients of the data?
Mr. Savaglio. Yes. It is in the subscriber contract that
they have to abide by the same regulations.
Mr. Walden. And what is the penalty if they do not?
Mr. Savaglio. Well, there are penalties, I believe, in the
contracts, financial penalties.
Mr. Walden. Thank you. All right.
Chairman Tauzin. The gentleman's time has expired.
Just for the record, are the two political parties
subscribers to VNS? I think they are.
Mr. Savaglio. Not during the day. There is a--I am told
here that the Republican Party buys the tabulated vote once the
tabulated vote starts coming in.
Chairman Tauzin. Once the tabs come out?
Mr. Savaglio. Yes.
Chairman Tauzin. They are not subscribers to the exit
polls?
Mr. Savaglio. No.
Chairman Tauzin. The gentlemen Mr. Bass is recognized.
Mr. Bass. No questions.
Chairman Tauzin. Mr. Shadegg.
Mr. Shadegg. Thank you, Mr. Chairman. I will be brief, or
at least I will try to be brief.
First of all, I thank you for your time here today. I also
want to thank you for the reviews you have each done. I think
they do reflect a desire on your part to be responsive to the
concerns of the American people, and I would echo my colleague
Mr. Walden's comments about the fact that we do have an
obligation to our constituents who don't get to talk to you and
perhaps but for this hearing wouldn't know of the efforts you
have made to examine what went wrong.
This is a very troubling issue for me. I am--for a number
of years in my life, I was an election law attorney. I
specialized in the Voting Rights Act. I advised the Arizona
Secretary of State on election law matters, handled automatic
recounts and other recounts in a State which was using punch
card balloting. So I have been fascinated with this whole
issue.
I am, however, a rabid proponent of the first amendment and
have never thought there was an easy solution to this problem,
and I still don't think there is an easy solution to this
problem. A part of me is concerned about the efforts you make
to make an early call, in fact affecting elections. And since I
report Arizona, a west coast State, there is a part of me that
is deeply concerned about that, and I think, in fact, there is
evidence that it does affect voting.
Indeed, you received a letter, I believe, signed by all 50
Secretaries of State back in 1998 reflecting their requests
that you not report or that you not call elections in advance
of the closing of the polls on the west coast. While I
sympathize with that request on their part, there is a part of
me that says it is unreasonable to ask you to do that. I think
the only solution to that problem is a uniform closing time,
and I commend the chairman for his efforts in that regard.
I simply think you cannot ask people to voluntarily
restrain when you are in the business you are in and people
have a different expectation, but I think the uniform closing
time will do a great deal of good, and I commend those of you,
and I guess it is almost everybody, if not everybody, who has
said here today that you support that effort.
I do have a serious problem with calling an election before
the polls close in a State; for example, what happened in
Florida this year. And it kind of mystifies me why networks
would call Florida 8, 10, 11 minutes before the Florida polls
closed, because that truly, I think, can affect the results,
and I, quite frankly--if we cannot accomplish a uniformed poll
closing statute, I would favor a statute that at least
prohibits you from calling a State before the polls in that
State close, because I do see that as a serious problem.
In this election I think that it is certainly possible that
voters in the Panhandle of Florida, many of whom are military
personnel, many of whom are solely interested in the national
election, the Presidential election, could have been en route
to the polls, heard that multiple networks had called the race
in Florida and said to themselves, there is no need for me to
go. And while I tend not to sympathize with voters who choose
not to go for whatever reason, I think that it still creates a
disincentive, which is a problem for democracy. As you know, we
don't have enough Americans, a large enough proportion of our
population, voting in our elections to begin with, and so
disincentives, I think, are a serious problem.
I was here much earlier in the day when I heard my
colleague Mr. Stupak say that, and this is a direct quote
because I wrote it down at the time, nothing the networks do
influenced the outcome of this election. I disagree with that.
I don't think he can prove that proposition. Indeed, I think
that it is pretty clear that--it certainly is possible that
some of the things that were done in this election did
influence the outcome of the election, and I certainly don't
believe you can prove Mr. Stupak's claim. But hopefully we get
to the point with a uniform poll closing time where that's not
an issue.
What I do want to express a little bit of concern about is
that in my State of Arizona now almost one-third of all voters
early vote. I have heard just now in the last few minutes
several references to absentee voting. On election night, as I
was being interviewed in Arizona about what was going on, I
said, and I believe, that exit polling is a fatally flawed
practice which can no longer be justified because of early
voting. I am skeptical of your ability to adjust the formulas
and to take into consideration the reality not of absentee
voting--under Arizona law, you used to have to have a reason
for absentee voting. Today you simply say it is more convenient
to me to vote absentee, and you vote by absentee, and Arizona
is 30 percent. Other States, I think, are even higher, and I
think those numbers are going to grow.
I am interested in how it is that you propose to, in a fair
fashion, continue to use exit polling when a third or more of
the population, at least of my State, is going to cast an early
ballot. And I would appreciate your comments, any of you that
have them, on that topic.
Chairman Tauzin. The gentleman's time has expired, but you
are free to comment anyway.
Mr. Boccardi. VNS may want to say something about the
question on polling, but I would like to pick up on one of the
many points you made. As an example of the concern I expressed
in my remarks about the first amendment, as you skipped through
some of your perspective here, one of the things I think you
said was that you support legislation that would ban us from
reporting before a poll closed, and that's an example to me
where your legislative mandate and our first amendment rights
would conflict pretty directly.
Mr. Lack. I think we have already quite clearly volunteered
that we are not going to do that, and I think all of us
collectively came, if not in minutes, hours, and if not hours,
days, that we were going to do that.
It was wrong, even though there was an agreement in place
which I think the chairman has referred to that the networks
were going to limit themselves from calling--or projecting a
winner in a State before most of or the majority of the polls
had closed, but that, it seems to me, is a fig leaf. We ought
to get rid of it and just said no, no more projections before
the polls have closed. So you don't--I mean, I agree with Lou's
point about the legislative aspect of it, but you don't need
to.
I think your skepticism with respect to the use of exit
polling and the early voting and how that's going to impact
exit polling is a very good one, and we are going to be looking
at that very carefully, and VNS has got a very long bridge to
cross with us to assure us that we have dealt with that issue
correctly as we face 2004.
Mr. Savaglio. The only thing I wanted to say about early
voting is that the way the early votes are counted vary from
State to State, and there are some aspects of it that allow you
to get information about it, how many people have returned
their ballots; the counties count them, and prepare them for
tabulation, and release them right at poll closing. So they are
not as unknowable, and obviously as it becomes--as it grows,
hopefully those kinds of things that are helpful will continue.
Mr. Shadegg. If I could make a very quick comment. It seems
to me there are two issues there. One is at least in Arizona
you can find out on Election Day who has early voted or at
least who got an early ballot to vote, and you could perhaps
contact them to adjust your sample to reflect their views.
The other thing you could do, and this is an issue where
Congress might step in, is that Congress might say if a State
is going to allow early voting, it has got to dump in the
results or announce the results of the early voting at a
specific time. Either it has got to be the first result
announced, or it has got to be the last result announced, but
it has got to be announced in a fashion where you can look at
the returns that come in and factor in what happened in the
early voting to combine that with what your exit polling showed
to give you an accurate projection of what actually happened.
Right now the States do that differently. In Arizona most
early ballots are dumped in right at the outset. They are the
first votes in. If you knew that, and you could look at those
results, you might be able to perfect your projection by
saying, okay, the first--we could even say States have to
report, must report, their early ballots as the first thing
they report and identify that that is what they are. Then you
could look at how those results came out, factor them into your
exit polling, and have an accurate projection. But it means to
me it is a problem you have to deal with, and I appreciate the
comments made here today.
Chairman Tauzin. Thank you. The gentleman's time has
expired, and all time has expired.
Happy Valentine's Day, for real this time.
Let me again thank you, and just a comment, and then we
will get out of here.
First of all, I hope you did not think this was a penalty
because you know Election Day was a mess. This was not about
penalty. It wasn't about punishing or piling on or criticizing.
Your own self-criticism was more than, believe me, anyone else
in the country probably would have leveled here today.
Second, you know, I don't have to tell you, we make errors
on this side of the estate. Generally, when we make them, it is
because we have bad information, too, and we have to be big
enough to say that when we do. I will never forget a vote I
cast on a bill called SALVO which had to do with reporting
emissions, and I thought the information we had was that it was
overbureaucratic, overburdensome, wouldn't do the job. I have
to tell you, it was a bad vote, but it has been a good act. It
has cleaned up more emissions in my State than any other single
tool we have ever used in the petrochemical industry, and I am
proud to tell you that I admit I was wrong, and we all do that.
You did that today, and I think the American people appreciate
that, frankly.
More importantly, I hope you have not felt you were here
under duress. You came here voluntarily. You know that. I asked
you to come. You came. I appreciate it, and I will give you one
commitment in return. I will fight to the death to protect your
right to keep doing this wrong if you really want to do it
wrong. That's the truth. I will not--no longer not support, I
will fight vigorously any attempt to legislate in the area of
your content. That is wrong for us to even talk about doing. We
won't do it.
On the other hand, I know you know our responsibility was
to make an inquiry about how we could make this system work
better; what we might do and what you were willing to do
voluntarily, and I appreciate not only the self-inquiry that
you literally undertook, but the fact that you were willing to
come all day and spend this time with us so that Americans
could learn that we are all interested in doing a better job.
I know your anchors appreciate this. I have talked to them.
They didn't like having egg and omelet all over their face that
night. Their credibility was at stake that night. They don't
like to report garbage out because they got garbage in. They
would like to know when they make calls and when they make
announcements or pronouncements about what the facts are that
they are really the facts and not some speculation that didn't
prove to be right. I am telling you, they appreciate the fact
that we are all trying, you on your side of the aisle and we on
ours, to try to get it right next time instead of getting it
done faster.
Thank you very much. This hearing stands adjourned.
[Whereupon, at 6:55 p.m., the committee was adjourned.]
[Additional material submitted for the record follows:]
Voter News Service
New York, New York 10122
April 13, 2001
The Honorable John D. Dingell
Ranking Member, Committee on Energy and Commerce
U.S. House of Representatives
Room 2125, Rayburn House Office Building
Washington, DC 20515
Dear Representative Dingell: This responds to your letter of March
26, 2001, which contained twenty-two questions following up on the
Committee's hearing on the network coverage of the 2000 presidential
election. Attached are our responses to those questions.
We appreciate that during the hearing the Committee recognized the
significant First Amendment issues implicated by its review of the
media's reporting on election results. Our attached responses reflect
our endeavor to be responsive to your inquiries while respecting the
strictures of the First Amendment and the need to maintain the vital
independence of the news-gathering and reporting process.
Sincerely,
Ted Savaglio
cc: The Honorable W.J. ``Billy'' Tauzin, Chairman
Committee on Energy and Commerce
questions for voter news service
Question 1. It has been stated in Voter News Service (VNS)
documents that VNS has made only six errors in projecting over 2,800
races. Please describe the dates and circumstances of those errors and
comparisons with the erroneous Florida calls.
Response 1. The statistical models employed by VNS have been
utilized in projecting over 2,800 races over the past 32 years. Since
its inception in 1993, VNS has utilized these models to project winners
in approximately 545 races. Prior to the 2000 presidential election,
VNS made only a single error in issuing a projection. In 1996, VNS
projected the wrong candidate as winner of the New Hampshire senatorial
election.
Question 2. If VNS has done additional analysis of the New Mexico
and Washington calls, please summarize the conclusions and submit the
analysis for the record.
Response 2. Since the February 14, 2001 hearing, VNS has not
reached any conclusions concerning the New Mexico or Washington calls.
Question 3. It was originally alleged that VNS and the networks
delayed calling the following states for George W. Bush, which
President Bush ultimately won by a margin of 6 percentage points or
more: Alabama, Arizona, Arkansas, Colorado, Georgia, Louisiana, North
Carolina, Virginia, and West Virginia.
Would you describe the reasons that these states were not called at
poll closing? If any of these races were ``too close to call'' at poll
closing, please list those.
Response 3. VNS has no first-hand knowledge of when or why the
networks made particular calls. VNS itself called Alabama at poll
closing. A decision by VNS as to when any given state is ready to be
called is based on numerous complex factors, including an evaluation of
the various analytical models presented by the VNS system.
Question 4. Should the error made in the New Hampshire senatorial
race in 1996 have given you a warning that the VNS models might not be
reliable in very close races?
Response 4. VNS works very hard to ensure that its statistical
models produce reliable results. Since the 1996 New Hampshire
senatorial election, and up to the presidential election that was the
subject of the Committee's hearing, VNS was correct in all of the
projections it issued, including projections made in close races.
Question 5. During the Committee's hearing, there was an allegation
of bias in the call of Ohio for George W. Bush. Please describe the
basis for the Ohio call and whether or not there was bias against
President Bush in this call.
Response 5. VNS categorically denies that there was political bias
against President Bush in its Ohio call, or in any other call made by
VNS.
Question 6. Dr. Edelman has conducted research on the impact of the
non-response rates on exit poll. Please list all of his publications
and papers and summarize their conclusions.
Has any other researcher looked at the impact of non-response
rates? Please list all of his/her publications and papers and summarize
their conclusions.
Response 6. Dr. Edelman's publications relating to non-response
rates on exit polls are listed (on page 70) as references to the
Research Triangle Institute's Report to the VNS Members dated January
22, 2001 (the ``RTI Report''), copies of which previously have been
provided to the Committee. In addition, copies of Dr. Edelman's
publications previously were provided both to majority and minority
counsel to the Committee.
Question 7. Mr. Biemer of Research Triangle Institute testified
that if VNS had communicated all of the deficiencies and uncertainties
on the decision screens to the networks, they would not have made
either of the incorrect Florida calls. Since VNS itself made the first
Florida call, do you agree with that statement? Why or why not?
What uncertainties did VNS itself ignore in the first Florida call?
Response 7. VNS believes it inappropriate to speculate about
hypothetical situations involving VNS members. VNS' own actions
relating to its call of Florida for Vice President Gore already have
been described in VNS' Submission to the Members, dated December 8,
2000 and the RTI Report, both of which previously were provided to the
Committee, and in testimony before the Committee given by Ted Savaglio
and Dr. Murray Edelman.
Question 8. One of the witnesses at the hearing stated that when
the networks combined to form VNS, they were able to expand the number
of precincts used for exit polling. Elsewhere, it has been stated that
both CBS and NBC, which were polling independently, had over 2,000
precincts in their polling sample, while VNS had only 1,400. Please
clarify the number of sample precincts used for polling by CBS/NBC and
VNS. If the total number was reduced under VNS, please give the date of
that reduction and the reasons for it.
Response 8. Since its inception in 1993, VNS has used at least 1399
sample precincts in connection with presidential elections.
Question 9. What is an acceptable precinct sample size for exit
polls?
Response 9. The size of an acceptable precinct sample for an exit
poll may vary based upon myriad factors. There is no absolute sample
size which, in any and all circumstances, is required for a reliable
exit poll.
Question 10. Please provide a state-by-state comparison of the
breakdown of the vote for Al Gore and George W. Bush as projected by
VNS from exit polls and actual vote.
Response 10. VNS uses its exit poll data in a series of projection
models, and these different models accordingly reflect a series of
different projections at the time a call is made in a given state.
Therefore, there is no single basis for comparing actual vote with a
VNS projection. VNS previously provided to the Committee a disk
containing all of the projections made by VNS for all of the states.
Question 11. For the 2000 elections, please describe the training
and/or information provided by VNS to the networks' decision desk
personnel.
Question 12. Did VNS have reason to believe that the network
decision desks understood the limitations of the end-of-night
outstanding vote model? Why or why not?
Question 13. Please describe the methods of communication between
the VNS decision desk and the networks' decision desks. In particular,
describe the capability to send messages out on the decision screen.
Question 14. If VNS wanted to send an instantaneous message to its
members during election night, was it technically possible to do so? If
so, why was a message not sent to warn the members against calling the
race for George W. Bush shortly after 2 a.m.?
Response 11.-14. VNS believes that Congressional inquiries into
communications between VNS and its member news organizations that
relate to the editorial process are inconsistent with First Amendment
guarantees of a free and independent press, and respectfully declines
to respond to these questions.
Question 15. Last year, VNS received a report from Batelle
Institute which made recommendations for a new computer system. Please
summarize Batelle's recommendations. Have they been implemented? If
not, why not? When will they be completed?
Response 15. Review of input from Battelle Institute is ongoing.
Question 16. Was there any way VNS could have avoided the New
Mexico situation, which resulted from an error by New Mexico election
officials?
Response 16. VNS employs stringent quality control systems, and is
reviewing and implementing additional quality control measures.
Irrespective of one's field of endeavor, no matter how stringent the
quality control processes one may adopt, one simply may not catch every
conceivable error.
Question 17. Is it accurate to say that the VNS model produces
state-by-state statistical biases that always favor Democrats? Please
explain your answer with examples.
Response 17. It absolutely is not accurate to say that the VNS
model produces state-by-state statistical biases that always favor
Democrats. Statistical biases present in exit polling tend to vary from
state to state, year to year and election to election. In the 2000
presidential election, VNS exit polling reflected a statistical bias
towards the Republican candidate in numerous states including Iowa,
Missouri and West Virginia, among others. Similarly, in the 1996
presidential election, VNS exit polling reflected a statistical bias
towards the Republican candidate in a number of states including
Connecticut, Louisiana and Oklahoma, among others. For Arkansas, like
some other states, VNS' exit poll reflected a statistical bias toward
the Democratic presidential candidate in one election (2000) but toward
the Republican in another (1996). In elections other than for
president, VNS' exit polling often has reflected a statistical bias in
favor of the Republican candidate, such as the 2000 Indiana and Utah
gubernatorial races, and the 2000 Montana senatorial race, for example.
Question 18. Is it accurate to state that exit polling today is
less scientific than it was in the past, and that ``garbage'' is going
into the exit poll model?
Response 18. It is absolutely false to state that exit polling
today is less scientific than it was in the past, and that ``garbage''
is going into the exit poll model.
Question 19. Is the VNS modeling approach basically sound, or will
it be completely scrapped and recreated in the next two years?
Response 19. The VNS modeling approach is sound and has produced
accurate projections 99.8% of the time since 1990, when the first joint
polling and projection effort began. Nevertheless, as VNS indicated
during the hearing, VNS is analyzing and implementing methods by which
to improve its approach.
Question 20. Will you be able to eliminate all of the statistical
biases in the models? Why or why not?
Response 20. VNS is committed to identifying, reducing and taking
into account all statistical biases reflected in its exit polling.
Statistical biases cannot completely be eliminated as some statistical
bias is inherent in all polling.
Question 21. In the 2000 election, was VNS limited by its budget on
the number of state absentee polls it could take?
Response 21. VNS, like all entities, must operate in accordance
with a budget. In the 2000 election, VNS conducted absentee polls in
the states where absentee voting traditionally has been high.
Question 22. The Research Triangle Institute suggested that VNS
consider hiring an outside polling firm. What is your response to that
recommendation? Have you ever used outside polling firms in the past?
Is there any reason that VNS should have its own exit polling
operation?
Response 22. VNS itself has used outside polling firms to conduct
telephone polls of absentee voters.
______
Associated Press
February 5, 2001
Honorable John D. Dingell, Ranking Member
Honorable Edward J. Markey, Member
Committee on Energy and Commerce
U.S. House of Representatives
Rayburn House Office Building, Room 2125
Washington, DC 20515
Dear Rep. Dingell and Rep. Markey: Lou Boccardi has asked me to
prepare this response to your letter of January 30 requesting certain
information about AP's coverage of vote tabulation on election night
last November.
Your four questions and AP's answers follow below and in the
attached spreadsheets.
Question 1. The exact time at which The Associated Press called
each state for Governor Bush or Vice President Gore. All calls for
Florida should be included in this tabulation.
Response. Attachment A provides a table containing the time at
which AP transmitted to its members a projected winner in each state,
and the projected winner's name. The table includes as well the AP's
advisories to its members concerning the Florida results.
Question 2. The person or persons in charge of receiving election
night information from the Voter News Service and recommending to
Associated Press officials that a state be called for one of the
candidates.
Response. AP's election night operations encompass bureaus in every
state, and the decision process is widely distributed. The principal
persons whose assessments and judgment are relied upon include the AP
chiefs of bureau, news editors and chief political writers in all of
the states, along with political editors in Washington, D.C., and New
York. In addition to data from Voter News Service, these individuals
monitor returns from AP's independent vote tabulation system, which is
described in the response to Question 4 below. No single AP person
precisely fits the role you describe, but the individual with widest
familiarity with AP operations and the decisions made on election night
is Sandra K. Johnson, chief of AP's Washington bureau.
Question 3. If the Associated Press conducted exit polls during the
November 7 election, please describe those polls and the purpose for
which they are conducted.
Response. AP did not conduct exit polls during the November 7
election and has never conducted exit polling on its own.
Question 4. Please describe the process by which The Associated
Press collected presidential vote tallies for the November 7 election
and provide copies of all vote tallies for the state of Florida from 7
p.m. EST on November 7 through 8 a.m.
Response. In addition to monitoring the VNS tabulation, AP gathers
tabulated vote returns nationwide on its own. To do this, AP retains
election data reporters in every county of each state. On election
night these reporters position themselves at county election offices
where election officials assemble precinct totals. At frequent
intervals, as the precinct reports come in and official county totals
in each election race are updated and made public, the AP reporter
calls the state AP data center, where the fresh numbers are received
and keypunched into the AP election database. The database continuously
generates a variety of running reports for AP and its members, always
based on the latest available figures.
Attachment B is a table which shows Florida presidential totals for
each candidate as AP reported them throughout election night based upon
official county-by-county figures, and the margin separating the
candidates at each interval.
Please consider these responses as supplementing Mr. Boccardi's
letter to Congressman Tauzin dated November 16, 2000, another copy of
which is enclosed.
I believe these answers are complete, but if anything about them is
unclear, please feel free to contact me at (212) 621-1796.
Sincerely,
David Tomlin
Assistant to the President
c: Honorable W.J. ``Billy'' Tauzin (with enclosures)
Enclosures: Attachment A, Attachment B, Letter from Mr. Boccardi to
Congressman Tauzin
Schedule A
AP WIN FORECASTS
------------------------------------------------------------------------
------------------------------------------------------------------------
Kentucky........................ 600 pm............ Bush
Indiana......................... 600 pm............ Bush
South Carolina.................. 700 pm............ Bush
Vermont......................... 700 pm............ Gore
Virginia........................ 720 pm............ Bush
Georgia......................... 742 pm............ Bush
Florida......................... 753 pm............ Gore*
Connectcut...................... 800 pm............ Gore
Delaware........................ 800 pm............ Gore
District of Columbia............ 800 pm............ Gore
Illinois........................ 800 pm............ Gore
Kansas.......................... 800 pm............ Bush
Maryland........................ 800 pm............ Gore
Massachusetts................... 800 pm............ Gore
Mississippi..................... 800 pm............ Bush
New Jersey...................... 800 pm............ Gore
Oklahoma........................ 800 pm............ Bush
Texas........................... 800 pm............ Bush
North Carolina.................. 808 pm............ Bush
Alabama......................... 825 pm............ Bush
Nebraska........................ 900 pm............ Bush
New York........................ 900 pm............ Gore
North Dakota.................... 900 pm............ Bush
Rhode Island.................... 900 pm............ Gore
South Dakota.................... 900 pm............ Bush
Wyoming......................... 900 pm............ Bush
Michigan........................ 923 pm............ Gore
Louisiana....................... 928 pm............ Bush
Pennsylvania.................... 930 pm............ Gore
Idaho........................... 1000 pm........... Bush
Montana......................... 1000 pm........... Bush
Utah............................ 1000 pm........... Bush
Tennessee....................... 1003 pm........... Bush
Minnesota....................... 1025 pm........... Gore
West Virginia................... 1049 pm........... Bush
Missouri........................ 1052 pm........... Bush
Ohio............................ 1058 pm........... Bush
California...................... 1100 pm........... Gore
Hawaii.......................... 1100 pm........... Gore
Maine........................... 1110 pm........... Gore
Colorado........................ 1140 pm........... Bush
Arizona......................... 1150 pm........... Bush
Alaska.......................... 1200 am, Nov. 8... Bush
New Hampshire................... 1205 am, Nov. 8... Bush
Arkansas........................ 1217 am, Nov. 8... Bush
Nevada.......................... 130 am, Nov. 8.... Bush
Washington...................... 130 am, Nov. 8.... Gore
Florida......................... 311 am, Nov. 8.... **
Iowa............................ 355 am, Nov. 8.... Gore
Wisconsin....................... 608 am, Nov. 8.... Gore
Oregon.......................... 712 pm, Nov. 10... Gore
New Mexico...................... Nov. 17........... Gore
------------------------------------------------------------------------
*At 9:59 p.m. AP transmitted an advisory to its members rescinding its
earlier projection.
**At 3:11 a.m. AP transmitted the following advisory to its members:
EDITORS:
The lead in Florida for George W. Bush has dwindled to about 6,000 in
the vote count. A small percentage of the vote has yet to be reported
in several counties, including two predominantly Democratic counties.
AP believes the uncounted votes in Broward and Palm Beach counties could
allow a change of the lead in the Florida vote. We are watching the
resolution of the actual vote count to assure if there is a change in
the Florida results, which could yet have an impact on the outcome of
the presidential election.
The AP
Schedule B
AP FLORIDA RESULTS
------------------------------------------------------------------------
Bush Gore Bush Lead
------------------------------------------------------------------------
19:22.................................... 13641 12437 1204
19:27.................................... 41123 34245 6878
19:42.................................... 52098 47647 4451
19:47.................................... 118324 112966 5358
19:52.................................... 118605 113106 5499
19:53.................................... 124883 116278 8605
20:02.................................... 181379 185435 -4056
20:17.................................... 274695 287878 -13183
20:22.................................... 638895 631331 7564
20:27.................................... 644233 635895 8338
20:32.................................... 658367 650284 8083
20:42.................................... 712380 699609 12771
20:47.................................... 738680 729491 9189
20:52.................................... 777562 748165 29397
20:53.................................... 779409 749198 30211
21:02.................................... 910052 829501 80551
21:12.................................... 979145 884020 95125
21:17.................................... 1052295 932082 120213
21:22.................................... 1085828 965030 120798
21:27.................................... 1129137 1005241 123896
21:32.................................... 1161106 1037804 123302
21:42.................................... 1298213 1191676 106537
21:47.................................... 1332881 1216992 115889
21:52.................................... 1416422 1280948 135474
21:53.................................... 1418504 1282503 136001
22:02.................................... 1487890 1352114 135776
22:12.................................... 1611442 1447603 163839
22:17.................................... 1650392 1490767 159625
22:22.................................... 1697049 1549320 147729
22:27.................................... 1741816 1597744 144072
22:32.................................... 1781530 1632164 149366
22:42.................................... 1850659 1703923 146736
22:47.................................... 1939188 1788296 150892
22:52.................................... 2015897 1870672 145225
22:58.................................... 2051892 1905072 146820
23:02.................................... 2059462 1914107 145355
23:12.................................... 2142726 2008164 134562
23:17.................................... 2168931 2021991 146940
23:20.................................... 2193550 2040857 152693
23:21.................................... 2198973 2049265 149708
23:22.................................... 2199593 2049699 149894
23:27.................................... 2199593 2049699 149894
23:32.................................... 2261404 2120834 140570
23:42.................................... 2364342 2229006 135336
23:47.................................... 2364342 2229006 135336
23:47.................................... 2387413 2251637 135776
23:52.................................... 2408540 2280202 128338
23:53.................................... 2408540 2280202 128338
0:02..................................... 2427272 2293101 134171
0:12..................................... 2501890 2366565 135325
0:17..................................... 2502581 2366905 135676
0:22..................................... 2521902 2383009 138893
0:27..................................... 2549937 2423468 126469
0:32..................................... 2549937 2423468 126469
0:42..................................... 2564441 2435897 128544
0:47..................................... 2609768 2497657 112111
0:47..................................... 2609768 2497657 112111
0:53..................................... 2609768 2497657 112111
1:02..................................... 2610984 2498403 112581
1:12..................................... 2709650 2627317 82333
1:17..................................... 2709753 2627347 82406
1:22..................................... 2736318 2672793 63525
1:27..................................... 2736318 2672793 63525
1:31..................................... 2736318 2676793 59525
1:42..................................... 2749894 2685533 64361
1:47..................................... 2763925 2701035 62890
1:47..................................... 2799648 2743162 56486
1:52..................................... 2799648 2743162 56486
2:02..................................... 2799648 2743162 56486
2:12..................................... 2822029 2774175 47854
2:16..................................... 2825865 2795352 30513
2:22..................................... 2850199 2834840 15359
2:27..................................... 2873834 2854829 19005
2:32..................................... 2873834 2854829 19005
2:42..................................... 2884625 2870618 14007
2:47..................................... 2884845 2870911 13934
2:47..................................... 2885417 2871463 13954
2:52..................................... 2890277 2879187 11090
3:02..................................... 2899999 2893634 6365
3:12..................................... 2903251 2901571 1680
3:17..................................... 2903251 2901571 1680
3:22..................................... 2903251 2901571 1680
3:27..................................... 2906997 2904519 2478
3:32..................................... 2906997 2904519 2478
3:42..................................... 2906997 2904519 2478
3:47..................................... 2906997 2904519 2478
3:47..................................... 2906997 2904519 2478
3:52..................................... 2906997 2904519 2478
4:02..................................... 2906997 2904519 2478
4:12..................................... 2906997 2904519 2478
4:17..................................... 2906997 2904519 2478
4:22..................................... 2906997 2904519 2478
4:27..................................... 2906997 2904519 2478
4:32..................................... 2906997 2904519 2478
4:42..................................... 2907600 2905712 1888
4:47..................................... 2907600 2905712 1888
4:52..................................... 2907600 2905712 1888
------------------------------------------------------------------------
______
Associated Press
November 16, 2000
Congressman Billy Tauzin
House of Representatives
2183 Rayburn House Office Building
Washington, D.C. 20515
Dear Congressman Tauzin: I write in response to your letter of
November 9 concerning reporting of the presidential election results
last week, particularly the use of exit polling and the projection of
final results.
As you may know, the media partners in the Voter News Service have
long agreed that no winner should be declared in any state before the
great majority of polls are closed in that state. That consensus
originated in 1985 as an understanding between the broadcast networks
and Congress. Although AP was not a party to it, we have abided by it,
and we signed on to it expressly when VNS was formed in 1993.
If the agreement among those with access to exit polling and early
vote tabulations were modified so that winners within a state are not
projected until every precinct in that state--not just a majority--has
reported, AP would be in agreement.
A few sentences of background about AP might be useful.
AP is a not-for-profit industry-owned news cooperative governed by
a board (mostly elected but with some appointed seats) made up of
owners and senior managers of the U.S. media companies we serve,
approximately 1,700 newspapers and 5,000 radio and television stations.
We serve thousands more subscribers overseas, along with a growing base
of Internet and other specialty news distributors. We have bureaus in
every state (two, I might say, in Louisiana) and in more than 100
nations around the world.
We do not operate at what you might call the ``retail'' level. Our
mission is to serve other news organizations with the most accurate,
complete and balanced reporting possible. We express no editorial
opinions. On election night, as always, editors here are aware that not
only AP's reputation but those of thousands of other news distributors
depend on the quality of our work--its accuracy, its timeliness and its
fairness.
As a VNS member, AP has access to the results of VNS exit polls on
election day, and VNS vote counts on election night for federal and
statewide offices.
As you know, AP and the network partners have instructed VNS
management to conduct a detailed internal review of its systems for
processing poll data and tabulated votes to ensure against a recurrence
of last week's problems.
AP also operates a backup vote tabulating system of its own for
federal and statewide races. We also tabulate votes for state
legislative elections and ballot issues, a total of roughly 6,000
contests in all. By contrast, VNS reports votes from about 550 races.
AP decisions to project a winner in major elections represent our
own editorial judgment by reporters and editors with specialized
knowledge of each state's electoral demographics. Debates, conflicting
interpretations, discrepancies are all dealt with as integral pieces of
our editorial process for handling returns. Those processes do not vary
by state.
You ask if a projection on the presidential outcome was made in any
state other than Florida and Kentucky when any polls remained open.
There are 13 states with more than one time zone. We declared Bush the
winner in Kansas while polls remained open to 1 percent of the voting
age population, and in Indiana, where 18 percent still had time to vote
if they hadn't already. As you know, neither state was a close call.
Finally, I would add only that the core here is the editorial
process, human judgment bolstered by sophisticated but not infallible
tools. I have tried to be responsive to your inquiry without
compromising the necessary independence of the news-gathering and
reporting process.
Sincerely yours,
Louis D. Boccardi
President and Chief Executive Officer
______
U.S. House of Representatives
Committee on Energy and Commerce
March 26, 2001
Mr. Andrew Heyward
President
CBS News
524 West 57th Street
New York, New York 10019
Dear Mr. Heyward: I have attached questions as a follow-up to the
Committee's hearing on the network coverage of the 2000 presidential
election. I request that you provide answers to these questions for the
record.
Please submit your response by close of business on Monday, April
9, 2001. If your staff has any questions or needs additional
information, please have them contact Edith Holleman, Minority Counsel,
at (202) 226-3400.
Thank you for your cooperation in this matter.
Sincerely,
John D. Dingell
Ranking Member
Attachment
cc: The Honorable W.J. ``Billy'' Tauzin, Chairman
Committee on Energy and Commerce
QUESTIONS FOR ANDREW HEYWARD, CBS NEWS
Question 1. In 1980, the networks called the election for Ronald
Reagan before the polls had closed in the Central, Mountain, and
Pacific time zones. The 2000 presidential election appears to have
presented a quite different fact pattern. Did CBS call the 2000
presidential election for Al Gore before the polls closed in the
Central, Mountain, and Pacific time zones? Did CBS ever imply that
because Florida was called for Al Gore, the race was over?
Question 2. The attached excerpt from a transcript of a press
conference held on November 14, 2000, by Chairman Tauzin and other
members of the Committee challenged the networks to prove that their
coverage was not intentionally biased. Is it your position that you
have met this burden of proof?
Question 3. Your report states that the CBS decision team made an
assumption that half of the votes were counted in Washington State at
the time CBS called that race. What was the basis for that assumption?
Question 4. What was the ``outer limit of sampling error'' that CBS
accepted in the exit and pre-election absentee polls? Why did CBS
accept this outer limit?
Question 5. Did CBS' own investigation find evidence of biased
modeling? Please explain.
Question 6. In the ten years that Voter News Service and its
predecessor have existed as a consortium, have any of the networks
asked for a ``bottoms up'' review of the operation? If not, why?
CBS News
New York, New York
April 9, 2001
The Honorable John D. Dingell
Ranking Member
Committee on Energy and Commerce
U.S. House of Representatives
Washington, D.C.
Dear Congressman Dingell: I am responding to your inquiry of March
26, 2001. Here is the additional information you requested about our
coverage of Election Night 2000.
CBS News never called the presidential race for Al Gore. In fact,
when CBS News did call the presidential election at 2:17 a.m., there
were no polls still open anywhere in the U.S. CBS News never implied
that the early call for Al Gore in Florida meant the race was over. In
fact, our independent expert, Dr. Kathleen Hall Jamieson, Dean of the
Annenberg School for Communication, noted in the CBS report on Election
Night coverage that Dan Rather explicitly made the point that ``The
Florida [Gore] call does not make a Gore victory inevitable, but it
makes Gore viable.'' (p.46)
Chairman W.J. ``Billy'' Tauzin said at the hearing on February 14,
2001, that ``as an initial matter, I want to take something quickly off
the table . . . [our investigators] found no evidence of intentional
misleading or biased reporting.'' As I wrote in my December 11, 2000,
letter to you, ``We unequivocally reject the notion that bias played
any part in the way that we reported the events of November 7.''
Washington State cast ballots on Election Day in two ways: by
absentee voting, which comprised 50 percent of the votes, and by voters
going to the polls on Election Day, which made up the other half. When
CBS News called the Washington State Senate race at 12:54 a.m. Election
Night, 26 percent of the precinct vote had been counted, and the
Decision Desk assumed that half the absentee votes had been counted as
well. Regrettably, that was not the case.
In Washington State, pre-Election Day polls were used to project
the expected votes from the absentee ballots, while Election Day exit
polls were used to project the outcome of the in-person voting. The
``outer limit of sampling error'' phrase meant the following: in the
Gorton/Cantwell race, the sampling error of the exit polls was plus or
minus 4 percentage points for each candidate's percentage, which could
mean a maximum 8 percentage points on the margin between the
candidates. When CBS News called Cantwell, the polls showed her ahead
by 13 points wrong all Election Day polling place voters. In fact, we
believe she carried those voters by only 5 or 6 points. The Voter News
Service pre-election absentee voter poll suggested that absentee voters
would be evenly split between Cantwell and Gorton. However, it appears
the absentee voters favored Gorton by about 6 points.
These differences are still within the sampling error of three to
four percentage points on each candidate's percentage, but they are at
the ``outer limit'' of the sampling error. It is very rare for two
separate surveys (one a phone survey, the other an in-person exit poll)
to be at that extreme and to be off in the same direction. This is a
case where one reform we have adopted for future Election Nights--
toughening the criteria for making calls in the closest races--might
have prevented the premature call.
As for your question of biased modeling, the models themselves are
neutral. They are not ``biased'' in any political sense. In its
independent review of VNS exit polls, RTI reported that Democratic
strength is more often overstated than Republican, but RTI did not
attribute that to the model itself. RTI suggested it might be the
result of non-response by Republicans. We will continue to improve the
accuracy of the models' predictions.
Finally, there has never been a ``bottoms up'' review of VNS, as
there was never a reason for one. Until this year the system had made
more than 2000 calls with only 6 errors since the 1960s. VRS, the
predecessor to VNS, emerged from the CBS election system. CBS News was
comfortable with the system, one that had performed excellently in the
VRS and then the VNS configuration (with only one error between 1990
and 1998). We assume other members investigated the system when they
joined the consortium and were also comfortable with it. As far as I
know, no ``bottoms up'' review was ever requested by any member after
it joined the consortium.
Thank you for the opportunity to clarify these matters.
Sincerely,
Andrew Heyward
President
______
ABC News
New York, New York
April 9, 2001
The Honorable John D. Dingell
Ranking Member
Committee on Energy and Commerce
United States House of Representatives
2125 Rayburn House Office Building
Washington, D.C. 20515-6115
Dear Congressman Dingell: Thank you for your letter of March 26,
2001, I enclose the responses of ABC News to your questions.
Sincerely,
David A. Westin
President
ABC TELEVISION NETWORK RESPONSE TO INQUIRY FROM THE COMMITTEE ON ENERGY
AND COMMERCE
Question 1. In 1980, the networks called the election for Ronald
Reagan before the polls had closed in the Central, Mountain, and
Pacific Time zones. The 2000 presidential election appears to have
presented a quite different fact pattern. Did ABC call the 2000
presidential election for Al Gore before the polls closed in the
Central, Mountain, and Pacific Time zones? Did ABC ever imply that
because Florida was called for Al Gore, the race was over?
Answer: ABC News did not ever project that Mr. Gore won the 2000
presidential election. Nor did ABC News state during its election
broadcast that the former Vice President's apparent early victory in
Florida decided the election. To the contrary, until the 2:10 AM
projection. for Mr. Bush, ABC News reported throughout the evening that
the 2000 election was extraordinarily close and could be won by either
candidate. This was the case even during the 40 minute period in which
ABC News had projected that Mr. Gore would prevail in Michigan, Florida
and Pennsylvania. During that time, ABC News stressed that Mr. Bush
could lose these three states and still ultimately win the presidency,
depending on the outcome of a number of hotly contested races in
western states.
Question 2. The attached excerpt from a transcript of a press
conference held on November 14, 2000, by Chairman Tauzin and other
members of the Committee challenged the networks to prove that their
coverage was not intentionally biased, Is it your position that you
have met this burden of proof?
Answer: ABC News has demonstrated through evidence and testimony
submitted to the Congress that there was no intentional political bias
in its coverage of the 2000 election. At a subsequent press conference
in February, and at the February 14, 2001, hearing Chairman Tauzin
acknowledged that no such bias was found.
Question 3. In 1990, the networks did not have their own decision
desks independent of Voter News Service's (VNS) predecessor, so there
was no competition among them to call states. In 1994, ABC surprised
all of the other VNS members by setting up its own decision desk to
name winners.
Why did ABC start the competition? Do you think it is reasonable to
expect that competing networks will eliminate their own decision desks
again?
Answer: ABC News consistently strives to provide thorough, timely,
and accurate reporting to its audience. The decision desk is part of
ABC News' comprehensive election unit, and the decision to use an
independent decision desk in 1994 was part of a journalistic judgment
that the audience would be well served by the addition.
ABC News has no information concerning the plans of other networks
in covering future elections.
Question 4. In 2000, ABC hired two statistical experts to look at
the VNS models to make sure that they did not make the same mistake
made in New Hampshire in 1996 and call the race wrong. According to an
article in American Journalism Review, these experts spent four days
familiarizing themselves with the VNS models.
Did they reassure you that a mistake like New Hampshire could not
be made again?
Answer: For the 2000 election, ABC News retained two outside
experts for assistance on various matters, including the mistaken
projection in the New Hampshire Senate race in 1996. These experts
suggested ways to reduce the chances of such a problem recurring, but
obviously could not guarantee that similar problems could not arise in
other elections.
Question 5. Your experts stated that they did not believe the first
Florida call should have been made because of the limited exit poll
data and the lack of knowledge about absentee voting. They did not
believe the second Florida call should have been made because the
absentee vote was not all in, and that there could be errors in the
vote tallies. But these experts were overridden by the journalist
running your decision desk. Please explain why.
Answer: While we will not comment on the precise deliberations of
the journalists who staffed the desk on election night, the decision
team assigned to make the Florida projection took all available
information into account. As shown on the decision screens made
available to the Committee, the data and the statistical models fully
supported the projection of ABC News for Vice President Gore in Florida
at the time it was made. In reviewing the data after the fact, it
appeared that these data and models were mistaken in three respects,
First, the exit poll data included in the models overstated Vice
President Gore's advantage. The check on this possible error in the
system failed because, by coincidence, the first few precincts
reporting actual vote tallies indicated (erroneously) that Vice
President Gore's lead was being under-predicted (rather than over-
predicted) by the exit poll results. Second, the absentee ballots were
more numerous and more favorable to Governor Bush than anticipated.
Finally, the statistical models used the 1998 governor's race as the
past comparison race. This provided a poor comparison that resulted in
an overstatement of Vice President Gore's lead. At the time ABC News
made its first projection in Florida, we knew about the possibility of
these errors. We believed they had been adequately accounted for in the
statistical models. We were wrong.
Question 6. If your decision desk knew that the outstanding votes
at 2:10 a.m. were in the very heavily Democratic counties of Palm Beach
and Brevard, why did they not know that the vote totals for each
candidate could radically change?
Answer: The data available at 2:10 a.m. adequately supported the
projection for Mr. Bush. Specifically, the information available
suggested that even taking into consideration the estimated outstanding
vote, Mr. Gore could not overtake Mr. Bush's lead. As we are now aware,
there were two problems with the data: 1) the model severely
underestimated the number of votes that were outstanding; and 2) a
tabulation error in Volusia county overstated Bush's lead by 20,000
votes. If even one of these had not occurred, a projection would not
have been made.
Question 7. In the ten years that VNS and its predecessor have
existed as a consortium, have any of the networks asked for a ``bottoms
up'' review of the operation? If not, why not?
Answer: All of the networks are and were actively involved in the
operation of VNS. Together, we comprise its governing board. The VNS
statistical models are dynamic, with changes put into place following
each election. Nothing in the prior, exemplary record of VNS suggested
a ``bottoms up'' review was required. Indeed, before the election of
2000, those most closely involved with VNS can remember only one
instance in which a VNS projection ultimately proved wrong (the 1996
New Hampshire Senate race). In that case, VNS retained outside experts
to examine what happened in the race, The experts were unable to
recommend any definitive steps to guard against a similar problem.
Finally, the board members are currently engaged in a thorough review
of the VNS operation as the 2000 election revealed a number of
deficiencies that VNS must immediately address.
______
U.S. House of Representatives
Committee on Energy and Commerce
March 26, 2001
Ms. Joan Konner
Professor of Journalism and Dean Emerita
Graduate School of Journalism
Columbia University
200 Central Park South, #33B
New York, New York 10019
Dear Ms. Konner: I have attached questions as a follow-up to the
Committee's hearing on the network coverage of the 2000 presidential
election. I request that you provide answers to these questions for the
record.
Please submit your response by close of business on Monday, April
9, 2001. If your staff has any questions or needs additional
information, please have them contact Edith Holleman, Minority Counsel,
at (202) 226-3400.
Thank you for your cooperation in this matter.
Sincerely,
John D. Dingell
Ranking Member
Attachments
cc: The Honorable W.J. ``Billy'' Tauzin, Chairman
Committee on Energy and Commerce
QUESTIONS FOR JOAN KONNER
Question 1. When Representative Sherrod Brown asked at the
Committee's hearing on network coverage of elections if there should be
a policy that close relatives of candidates, such as brothers, sisters,
cousins, etc., should not be allowed to be on the networks' decision
desk on election night, you responded by stating that if there was a
``perceived conflict of interest'' there should be ``some layers of
insulation between that individual and the person who is going to make
the announcement on the air.''
John Ellis, a first cousin of George W. Bush, was hired by Fox in
1998 to--in Mr. Ellis' words--``professionalize the poll desk.'' (Mr.
Ellis had worked on election coverage for NBC from 1980-89.) However,
in a July 3, 1999, column in the Boston Globe, Mr. Ellis stated that he
would no longer write about the 2000 campaign because, ``I am loyal to
my cousin, Governor George Bush of Texas. I put that loyalty ahead of
my loyalty to anyone else outside my immediate family . . . there is no
way for you to know if I am telling you the truth about George W.
Bush's presidential campaign because in his case, my loyalty goes to
him and not to you.'' (Copy attached.) In the fall of 2000, Mr. Ellis
was interviewed on a Frontline show called ``The Choice 2000'' as a
family member about his knowledge of George W. Bush and the Bush
family. (Copy attached.)
Despite this, John Ellis was selected again to head the Fox
decision desk in 2000. He then made the first call of Florida--and the
entire election--for George W. Bush. In your opinion, would these
circumstances as described above establish a ``perceived conflict of
interest'' on Mr. Ellis' part?
Question 2. When Mr. Ellis worked at NBC, and his uncle was a
candidate for President in 1988, he did not work on the decision desk,
but wrote the daily political news summary. He told our staff in an
interview that even then his loyalty ``was to my uncle, not to NBC
News.'' How should a network handle such a perceived conflict of
interest?
Question 3. There has been an allegation in the press that the
network anchors talked so much about ``Gore's winning ways'' on the
East Coast that Slade Gorton lost the Senatorial election in Washington
State. Washington is a state with a 50 percent absentee vote. Did your
panel reach this conclusion about Washington or any other state in its
review of CNN's election night coverage?
Question 4. In your opinion, was it complacency or the time crunch
caused by competition that kept the networks from checking the
Associated Press' vote totals before calling the election for George W.
Bush shortly after 2 a.m.?
Question 5. Please describe the relationship between the CNN/CBS
decision desk and the VNS decision desk in terms of pre-election
training and discussion of the model and the communications on election
night. How would you improve it?
Columbia University
Graduate School of Journalism
March 30, 2001
Edith Holleman
Minority Counsel
US House of Representatives
Committee on Energy and Commerce
Room 2125
Rayburn House Office Building
Washington, DC 20515-6115
Dear Edith, attached find the answers to the questions sent to me
by Rep. John Dingell:
Sincerely,
Joan Konner
RESPONSES
Response 1. The answer to your question is ``yes.'' If John Ellis
had, indeed, made comments stating that his loyalties to the Bush
family superceded any commitment he has to his profession or his
employer, then I would judge that to be not only a perceived conflict-
of-interest but a real conflict-of-interest for a journalist. (I did
not know at the time of the hearing that he had made any such
statement.) While that does not disqualify an individual from any
position as a journalist, it would, in my judgement, disqualify that
person for any decision-making role involving reporting on his
relatives during an election. Often friends and relatives are hired by
journalism organizations because of their connections to the
newsmakers. Their access to sources makes them valuable to the
organization. However, the news organization should take every
precaution against placing such an individual in an assignment that
could result in bias in reporting.
Response 2. As you are well aware, many of those who appear as
pundits on the air, and also in print, have worked for political
parties and/or political candidates and officeholders. Their political
background and bias are known facts of their fife. Some individuals
leave politics behind them and become very good objective reporters and
commentators. Many others remain party or candidate apologists.
Personally, I deplore casting someone in the role of a journalist,
reporter or commentator, who clearly is in politics, not journalism.
However, the practice is so widespread that to recommend removing those
people from the air or from print, would be to remove what passes today
as political commentary. I wouldn't object to that, but I know it is
unrealistic to expect that that will happen.
Response 3. I do not think there is any evidence that network
reporting affected the outcome of any election. I am not aware that CNN
or any other network favored the campaign of Al Gore. In fact, most of
the coverage of both Al Gore and his campaign was highly critical.
Response 4. It was not complacency or a time crunch. Our report
stated that hyper-competition, over reliance on polls, and speed
accounted for the errors. There is no deadline for getting it right.
The urge to beat the competition and to demonstrate that you can
interpret the numbers better than any of your competitors were the
driving forces behind the decision-making on election night. Call it
arrogance or hubris, if you want. We did not.
Response 5. We did not examine the relationship between the CNN and
CBS decision desks that night. Our information indicated that the
election coverage as a whole at CNN was planned and operated
independently. We did not examine or report directly on the pre-
election training at VNS, although there were studies by other
organizations that did. One report, in particular, was both
complimentary as well as critical. The documents we examined indicated
that the communication between VNS and its members was not what it
should have been. VNS failed to notify its members of known errors in a
timely fashion and VNS was unable to correct errors promptly. There
were other unacceptable errors and failures--human and technological--
in both the tabulation and communication systems. The report
recommended that every effort be made to examine and analyze the
weaknesses and breakdown in the system and that every effort be made to
fix it or to develop a new system.
______
U.S. House of Representatives
Committee on Energy and Commerce
March 26, 2001
Mr. James V. Risser
Retired Director
Knight Fellowship Program
Stanford University
394 Diamond Street
San Francisco, California 94114
Dear Mr. Risser: I have attached questions as a follow-up to the
Committee's hearing on the network coverage of the 2000 presidential
election. I request that you provide answers to these questions for the
record.
Please submit your response by close of business on Monday, April
9, 2001. If your staff has any questions or needs additional
information, please have them contact Edith Holleman, Minority Counsel,
at (202) 226-3400.
Thank you for your cooperation in this matter.
Sincerely,
John D. Dingell
Ranking Member
Attachments
cc: The Honorable W.J. ``Billy'' Tauzin, Chairman
Committee on Energy and Commerce
Questions for James Risser
1. When Representative Sherrod Brown asked you at the Committee's
hearing on network coverage of elections if there should be a policy
that close relatives of candidates, such as brothers, sisters, cousins,
etc., should not be allowed to be on the networks' decision desk on
election night, you responded by stating that if there was a
``perceived conflict of interest'' there should be ``some layers of
insulation between that individual and the person who is going to make
the announcement on the air.''
John Ellis, a first cousin of George W. Bush, was hired by Fox in
1998 to--in Mr. Ellis' words--``professionalize the poll desk.'' (Mr.
Ellis had worked on election coverage for NBC from 1980-89.) However,
in a July 3, 1999, column in the Boston Globe, Mr. Ellis stated that he
would no longer write about the 2000 campaign because, ``I am loyal to
my cousin, Governor George Bush of Texas. I put that loyalty ahead of
my loyalty to anyone else outside my immediate family . . . there is no
way for you to know if I am telling you the truth about George W.
Bush's presidential campaign because in his case, my loyalty goes to
him and not to you.'' (Copy attached.) In the fall of 2000, Mr. Ellis
was interviewed on a Frontline show called ``The Choice 2000'' as a
family member about his knowledge of George W. Bush and the Bush
family. (Copy attached.)
Despite this, John Ellis was selected again to head the Fox
decision desk in 2000. He then made the first call of Florida--and the
entire election--for George W. Bush. In your opinion, would these
circumstances as described above establish a ``perceived conflict of
interest'' on Mr. Ellis' part?
2. When Mr. Ellis worked at NBC, and his uncle was a candidate for
President in 1988, he did not work on the decision desk, but wrote the
daily political news summary. He told our staff in an interview that
even then his loyalty ``was to my uncle, not to NBC News.'' How should
a network handle such a perceived conflict of interest?
3. There has been an allegation in the press that the network
anchors talked so much about ``Gore's winning ways'' on the East Coast
that Slade Gorton lost the Senatorial election in Washington State.
Washington is a state with a 50 percent absentee vote. Did your panel
reach this conclusion about Washington or any other state in its review
of CNN's election night coverage?
4. In your opinion, was it complacency or the time crunch caused by
competition that kept the networks from checking the Associated Press'
vote totals before calling the election for George W. Bush shortly
after 2 a.m.?
5. Please describe the relationship between the CNN/CBS decision
desk and the VNS decision desk in terms of pre-election training and
discussion of the model and the communications on election night. How
would you improve it?
April 2, 2001
The Honorable John D. Dingell
Committee on Energy and Commerce
U.S. House of Representatives
2125 Rayburn House Office Building
Washington, DC 20515-6115
Dear Congressman Dingell: In answer to the five questions that you
sent me on March 26, as a follow-up to the hearing on network coverage
of the 2000 presidential election, I would answer as follows:
(1) If the circumstances surrounding John Ellis are as described in
your first question, then I would agree that a ``perceived conflict of
interest'' existed on Mr. Ellis' part.
(2) Assuming the circumstances involving Mr. Ellis and NBC News
back in 1988 were as described in your second question, and also
assuming that NBC News knew that Mr. Ellis was a nephew of presidential
candidate George H.W. Bush, the network should not have employed him to
write the daily political news summary. I believe this is true, whether
or not the network knew of Mr. Ellis' attitude of owing loyalty to his
uncle and not to NBC News.
(3) The panel on which I served did not investigate nor reach any
conclusion about the impact on Washington State or other states of Vice
President Gore's supposed ``winning ways'' in the East.
(4) I believe that both complacency and the time crunch were
factors in the networks not checking Associated Press vote totals
before calling Florida and the election for George W. Bush. The
networks were under self-imposed time pressures to make a call, and in
addition they were too complacent about the accuracy of information
received from Voter News Service.
(5) I don't have enough knowledge of the pre-election training, the
discussion of computer models, and the election night communications to
answer this question about the relationship between the CNN/CBS
decision desk and VNS.
Sincerely,
James V. Risser
[GRAPHIC] [TIFF OMITTED] T1490.001
[GRAPHIC] [TIFF OMITTED] T1490.002
[GRAPHIC] [TIFF OMITTED] T1490.003
[GRAPHIC] [TIFF OMITTED] T1490.004
[GRAPHIC] [TIFF OMITTED] T1490.005
[GRAPHIC] [TIFF OMITTED] T1490.006
[GRAPHIC] [TIFF OMITTED] T1490.007
[GRAPHIC] [TIFF OMITTED] T1490.008
[GRAPHIC] [TIFF OMITTED] T1490.009
[GRAPHIC] [TIFF OMITTED] T1490.010
[GRAPHIC] [TIFF OMITTED] T1490.011
[GRAPHIC] [TIFF OMITTED] T1490.012
[GRAPHIC] [TIFF OMITTED] T1490.013
[GRAPHIC] [TIFF OMITTED] T1490.014
[GRAPHIC] [TIFF OMITTED] T1490.015
[GRAPHIC] [TIFF OMITTED] T1490.016
[GRAPHIC] [TIFF OMITTED] T1490.017
[GRAPHIC] [TIFF OMITTED] T1490.018
[GRAPHIC] [TIFF OMITTED] T1490.019
[GRAPHIC] [TIFF OMITTED] T1490.020
[GRAPHIC] [TIFF OMITTED] T1490.021
[GRAPHIC] [TIFF OMITTED] T1490.022
[GRAPHIC] [TIFF OMITTED] T1490.023
[GRAPHIC] [TIFF OMITTED] T1490.024
[GRAPHIC] [TIFF OMITTED] T1490.025
[GRAPHIC] [TIFF OMITTED] T1490.026
[GRAPHIC] [TIFF OMITTED] T1490.027
[GRAPHIC] [TIFF OMITTED] T1490.028
[GRAPHIC] [TIFF OMITTED] T1490.029
[GRAPHIC] [TIFF OMITTED] T1490.030
[GRAPHIC] [TIFF OMITTED] T1490.031
[GRAPHIC] [TIFF OMITTED] T1490.032
[GRAPHIC] [TIFF OMITTED] T1490.033
[GRAPHIC] [TIFF OMITTED] T1490.034
[GRAPHIC] [TIFF OMITTED] T1490.035
[GRAPHIC] [TIFF OMITTED] T1490.036
[GRAPHIC] [TIFF OMITTED] T1490.037
[GRAPHIC] [TIFF OMITTED] T1490.038
[GRAPHIC] [TIFF OMITTED] T1490.039
[GRAPHIC] [TIFF OMITTED] T1490.040
[GRAPHIC] [TIFF OMITTED] T1490.041
[GRAPHIC] [TIFF OMITTED] T1490.042
[GRAPHIC] [TIFF OMITTED] T1490.043
[GRAPHIC] [TIFF OMITTED] T1490.044
[GRAPHIC] [TIFF OMITTED] T1490.045
[GRAPHIC] [TIFF OMITTED] T1490.046
[GRAPHIC] [TIFF OMITTED] T1490.047
[GRAPHIC] [TIFF OMITTED] T1490.048
[GRAPHIC] [TIFF OMITTED] T1490.049
[GRAPHIC] [TIFF OMITTED] T1490.050
[GRAPHIC] [TIFF OMITTED] T1490.051
[GRAPHIC] [TIFF OMITTED] T1490.052
[GRAPHIC] [TIFF OMITTED] T1490.053
[GRAPHIC] [TIFF OMITTED] T1490.054
[GRAPHIC] [TIFF OMITTED] T1490.055
[GRAPHIC] [TIFF OMITTED] T1490.056
[GRAPHIC] [TIFF OMITTED] T1490.057
[GRAPHIC] [TIFF OMITTED] T1490.058
[GRAPHIC] [TIFF OMITTED] T1490.059
[GRAPHIC] [TIFF OMITTED] T1490.060
[GRAPHIC] [TIFF OMITTED] T1490.061