[House Hearing, 110 Congress]
[From the U.S. Government Publishing Office]


 
     GAO BRIEFING TO THE TASK FORCE: REPORT ON THE FINDINGS IN THE 
      INVESTIGATION INTO THE FL-13 CONGRESSIONAL DISTRICT ELECTION

=======================================================================

                                MEETING

                               before the

                           COMMITTEE ON HOUSE
                             ADMINISTRATION

    TASK FORCE FOR THE CONTESTED ELECTION IN THE 13TH CONGRESSIONAL 
                          DISTRICT OF FLORIDA

                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED TENTH CONGRESS

                             SECOND SESSION

                               __________

                HELD IN WASHINGTON, DC, FEBRUARY 8, 2008

                               __________

      Printed for the use of the Committee on House Administration


                       Available on the Internet:
   http://www.gpoaccess.gov/congress/house/administration/index.html


                     U.S. GOVERNMENT PRINTING OFFICE
42-709 PDF                 WASHINGTON DC:  2008
---------------------------------------------------------------------
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512ï¿½091800  
Fax: (202) 512ï¿½092104 Mail: Stop IDCC, Washington, DC 20402ï¿½090001

                   COMMITTEE ON HOUSE ADMINISTRATION

                ROBERT A. BRADY, Pennsylvania, Chairman
ZOE LOFGREN, California,             VERNON J. EHLERS, Michigan,
  Vice-Chairwoman                      Ranking Minority Member
MICHAEL E. CAPUANO, Massachusetts    DANIEL E. LUNGREN, California
CHARLES A. GONZALEZ, Texas           KEVIN McCARTHY, California
SUSAN A. DAVIS, California
ARTUR DAVIS, Alabama
                 S. Elizabeth Birnbaum, Staff Director
                William Plaster, Minority Staff Director

    Task Force for the Contested Election in the 13th Congressional 
                          District of Florida

                  CHARLES A. GONZALEZ, Texas, Chairman
ZOE LOFGREN, California              KEVIN McCARTHY, California


GAO BRIEFING TO THE TASK FORCE REPORT ON FINDINGS IN THE INVESTIGATION 
     INTO THE FLORIDA-13 CONGRESSIONAL DISTRICT CONTESTED ELECTION

                              ----------                              


                        FRIDAY, FEBRUARY 8, 2008

                  House of Representatives,
                          Task Force on Florida-13,
                         Committee on House Administration,
                                                    Washington, DC.
    The task force met, pursuant to call, at 10:09 a.m., in 
Room 1310, Longworth House Office Building, Hon. Charles A. 
Gonzalez (chairman of the task force) presiding.
    Present: Representatives Gonzalez, Lofgren, McCarthy, 
Ehlers and Lungren.
    Staff Present: Liz Birnbaum, Staff Director; Thomas Hicks, 
Senior Election Counsel; Janelle Hu, Election Counsel; Jennifer 
Daehn, Election Counsel; Matt Pinkus, Professional Staff 
Member/Parliamentarian; Kyle Anderson, Press Director; Kristin 
McCowan, Chief Legislative Clerk; Daniel Favarulo, Legislative 
Assistant, Elections; Gineen Beach, Minority Election Counsel; 
and Bryan T. Dorsey, Minority Professional Staff Member.
    The Chairman. I will call to order at this time the 
Committee on House Administration's Florida-13 Task Force, and 
I apologize for the slight delay. But we were having problems 
getting over here, obviously, with traffic and such. So my 
apologies.
    We are going to attempt to be brief today and hear the 
final report by GAO. But first I think that it would be 
appropriate for me to extend my thanks, of course, to my fellow 
members of the task force; and that would be Congresswoman Zoe 
Lofgren and Kevin McCarthy from California. I also want to 
extend my thanks to Congressman Dan Lungren, who was not an 
official member of the task force but, nevertheless, attended 
some of the meetings and offered some very constructive advice 
which actually was followed. And so, Dan, I want to thank you.
    To the majority and minority staff, again thank you. 
Because, at the end of the day, of course, all of the work is 
truly predicated on the product that they produce for us as we 
go through the machinations of the task force.
    And definitely to the late Congresswoman Juanita Millender-
McDonald--and you see her portrait in the back of the room--for 
her fine service and the individual who actually created this 
task force with the simple charge of: Get it right and get to 
the truth, which hopefully that is what we have done.
    I also want to extend the task force's and the committee's 
sincere thanks to the Government Accountability Office for 
their fine work. I think when we first selected them, they 
entered it with some reservation.
    I think the concern of any agency or department many times 
is that Congress or Members of Congress will drag them into a 
political debate, which would not serve, of course, their 
mission nor their purpose and would jeopardize their 
objectivity. That did not occur, and we appreciate that they 
were able to assist us to the degree and the fine manner in 
which they did.
    I will be making an opening statement. Then I will be 
recognizing the members of the task force for opening 
statements.
    Clive Thompson, writing in the January 6, 2008, issue of 
the New York Times magazine in an article entitled ``Can You 
Count on These Machines?'' makes the following observation: 
``The mistrust of touch-screen machines is thus equal parts 
technological and ideological.'' Technology is subject to 
empirical analysis. Ideology is not. So today we address that 
which we are capable of resolving through the scientific 
method, that is, the technological mistrust surrounding the 
November, 2006, election for the office of the United States 
Representative from Florida's 13th Congressional District.
    The challenge presented before the task force was to 
determine the merit and validity of the central allegation 
contained in contestant's notice of contest that the electronic 
voting machines in Sarasota County malfunctioned, resulting in 
18,000 undervotes, thus bringing into question the reliability 
of the vote totals determining the winner of the election as 
reported by the Florida State officials to the United States 
House of Representatives.
    The task force at the outset decided that it would not 
entertain testing and findings by opposing experts designated 
by the contestant and contestee, thus avoiding a ``dueling 
experts'' dilemma. Rather, the task force unanimously decided 
that the United States Government Accountability Office, with 
its credentialed resources, would serve as the impartial and 
independent expert.
    The task force further agreed that it would abide by the 
GAO's findings barring some substantial basis to question those 
findings.
    Today, the Government Accountability Office will formally 
present its final report.
    Following established task force procedure, GAO did make a 
preliminary briefing before the task force on February 6, 2008, 
of its draft report.
    I want to make an important point. The task force has 
respected GAO's established procedures and protocols. Until a 
final draft is prepared and presented, GAO's draft report 
remains a work in progress. On February 6, at the preliminary 
briefing, task force members determined that the contestant and 
the contestee or their representatives should have a copy of 
the draft report. This was to allow them to review the 
materials and convey any questions or concerns to the task 
force in preparation for today's meeting.
    I am aware that the draft report was made publicly 
available, contrary to the spirit of cooperation and respect 
for GAO's procedures and protocol. I apologize to the United 
States Government Accountability Office for this was not what I 
preferred and was contrary to what I stated and requested at 
the February 6th briefing. This development is not conducive to 
creating and maintaining a healthy working relationship between 
Congress and GAO based on trust and mutual respect.
    With that, I will recognize the minority member of the task 
force, Congressman McCarthy, for an opening statement.
    Mr. McCarthy. Well, thank you, Chairman Gonzalez.
    Before we begin, I do want to thank the GAO for the work 
they have done.
    I also notice Kurt Browning, Secretary of State of Florida 
is here and thank you for the work you have done. You did the 
study before; and, as GAO went down, they were very 
complimentary of how you opened it up and made sure as we went 
all the way through.
    And I want to thank this committee and especially Chairman 
Gonzalez, because I think you have laid the groundwork of how 
contested races in the future should be handed out. I thought 
every decision or every vote we came to a conclusion was in a 
bipartisan manner.
    And going to the GAO was something new. And with my own 
experience, actually having been a staffer and worked on 
contested races before, I think this is one that has been more 
thoroughly analyzed than any one before it. So I do appreciate 
the work you have done and all the committee, as well as Zoe 
Lofgren; and I thought everybody worked in a bipartisan manner. 
So thank you for that.
    Today we will hear the report of the GAO's investigation on 
this most studied election. The results are as clear as we can 
objectively expect. After thoroughly combing and analyzing past 
studies and conducting its own studies to confirm past results, 
there is no evidence suggesting any malfunction of the 
processing and counting of ballots on the DREs.
    We will hear from the GAO today and ask about its report 
that concludes that the machines counted votes accurately. With 
this final objective confirmation, I hope that we can finally 
put to rest for the people of the 13th District of Florida the 
notice of contest that the challenger, Christine Jennings, 
filed asking the House to overturn the election results against 
Congressman Vern Buchanan and send a strong message to the 
American people that results of the 2006 election and this 
particular election have been and were always correct. And I 
look forward to hearing the rest of the report.
    The Chairman. Thank you very much.
    And the Chair will recognize Ms. Lofgren.
    Ms. Lofgren. I would just add that I think I agree with the 
other members of the task force. This has been an endeavor that 
has been bipartisan. I think all of our votes have been 
unanimous. We entrusted the analysis to experts at the GAO, and 
we didn't interfere with them in any way. We let them go 
through their study, and now they have reached a conclusion.
    And I would only note that if we had had, as the GAO has 
noted, if we had had a paper trail, we would have certainly not 
had to go through this exercise. So this will never happen 
again in Florida, because Florida, pursuant to the Governor's 
direction I think in the legislature, no longer uses these 
machines and there will always be a paper trail. So this will 
be the last such contest we will see from the State of Florida.
    And, hopefully, as States change their election systems, we 
will never have this type of situation again.
    I would like to thank again the staff, members of the 
committee and especially you, Mr. Chairman, for your terrific 
leadership in this not-an-easy endeavor. But I think we have 
done our duty to the institution fairly and under the 
Constitution.
    Thank you.
    The Chairman. Thank you very much.
    And, even though not a formal member of the task force, but 
members, of course, of the full committee to which we will be 
making a recommendation, hopefully, at the end of this meeting 
today, I do want to recognize for brief opening statements, and 
that would be the ranking member of the full committee, Mr. 
Ehlers.
    Mr. Ehlers. Thank you, Mr. Chairman; and the majority of 
the statement that I want to say is to commend you and the 
other members of the task force.
    It is obvious to me that your judicial experience has been 
invaluable to you and that your judicial demeanor in handling 
this has been exemplary, and I deeply appreciate it. I have 
served on several task forces and chaired one. I know how 
difficult the task can be. And you have been very fair, 
thorough and evenhanded in the handling of this. You have set 
an example for all the task forces of the future, and I really 
commend you for that.
    The second comment I would like to make is a very important 
side benefit of what we have gone through with this. The use of 
the GAO and their work done on this, I think has set a good 
example not just for us but for the Nation. It will restore the 
confidence of the public in the process.
    It was not a coincidence that four of the contests were 
filed from Florida. There was only one contest that was filed 
outside of Florida, and that had to do with residency, not 
machines. But four, the other four were filed from Florida, and 
they questioned the machines. Obviously, because of the 
experience Florida has had over the past decade, there is a 
spillover here. The public has lost confidence in the results. 
And I think the result of this is going to demonstrate to the 
public that the voting process, by and large, is proper, 
correct and believable and that the public should not question 
the accuracy of election results as they have done for the past 
8 to 12 years.
    So you have set an example and a standard that is very 
helpful to the Nation as well as in resolving this case, and I 
thank you for it.
    The Chairman. Thank you very much for your kind words.
    And the Chair will recognize Mr. Lungren for some brief 
opening remarks.
    Mr. Lungren. Thank you very much, Mr. Chairman.
    First of all, thank you for your indulgence in allowing me 
to participate in the hearings of this task force although I 
was not a member.
    Some more than 20 years ago I was a member of the full body 
when there was a contested election, the results of which 
caused bitterness in the House; and partisanship was 
exacerbated. I genuinely did not want us to return to those 
days and was most interested in how this contest would be 
handled.
    And while I did not know the chairman well before the 
hearings, I did serve with his father and enjoyed my 
relationship with his father and only knew the chairman by 
reputation. But I want to thank you not only on my behalf but 
on behalf of all Members for the way that you conducted this 
task force. This was somewhat of a test of our House as to how 
we would handle this, not only because there were questions 
raised about the election results but the nature of the 
contest. And I think your judgment in having GAO get involved, 
which was adopted unanimously by the other members of the task 
force, was very, very essential in the manner in which this was 
handled.
    And while I know we are still awaiting the report, I have 
looked at the draft report; and I think I know what they are 
going to tell us. And I think this does help restore confidence 
in the ability of this House to be able to handle touchy 
matters like this, but, also, it is an example to the American 
people of where we can work in a bipartisan basis, where it is 
easy to be partisan otherwise.
    And I thank the chairman and the other members of this task 
force.
    The Chairman. Thank you very much, Mr. Lungren.
    The Chairman. At this time, we'll proceed with the final 
report by the representative from the United States Government 
Accountability Office.

     STATEMENT OF NABAJYOTI BARKAKATI, PH.D., ACTING CHIEF 
    TECHNOLOGIST, APPLIED RESEARCH AND METHODS, GOVERNMENT 
                     ACCOUNTABILITY OFFICE

    Mr. Barkakati. Thank you, Mr. Chairman.
    Chairman Gonzalez, members of the task force, we have 
submitted the formal statement for the record, and I am going 
to summarize briefly.
    As our statement says I am here today to present the 
findings on our Florida-13 review, based on the testing we 
conducted on iVotronic voting machines used in the 2006 general 
election in Sarasota County, Florida.
    I would like to begin by thanking the task force for its 
overall support of our efforts and specifically for the 
assistance provided in obtaining the resources of the House 
recording studio, which were critical in successfully 
completing our tests.
    At the October 2, 2007, meeting of the task force, we 
proposed and you asked us to proceed with three tests--firmware 
verification test, ballot test and calibration test--in order 
to obtain increased assurance that the iVotronic voting 
machines did not contribute to the large undervote observed in 
the 2006 elections in Sarasota County.
    To conduct the three tests, we developed test protocols and 
detailed test procedures. We met with officials from the 
Sarasota County Supervisor of Elections and the Florida 
Department of State and Florida Division of Elections in order 
to make necessary arrangements to obtain access to the voting 
machines and to schedule and conduct the tests.
    Thanks to everyone's cooperation and help, we have been 
able to successfully complete our testing during November 26th 
through December 4, 2007, in Sarasota County's voting equipment 
facility in Sarasota, Florida.
    Our tests on the randomly selected iVotronic machines in 
Sarasota County did not identify any problems that would 
indicate that the machines were responsible for the undervote 
in the Florida-13 race in the 2006 general election.
    In our firmware verification test, we extracted the 
firmware from a random sample of 115 iVotronics and found that 
in each case the firmware extracted from the machines matched 
the firmware that was escrowed and certified by the Florida 
Division of Elections.
    The statistical approach that we used in selecting these 
machines enables us to say with a 99 percent confidence level 
that no more than 60 of the 1,499 iVotronics that were used in 
the 2006 election could have been running a different version 
of software. Consequently, we were able to place more 
confidence in the tests that we conducted on a smaller number 
of machines, both we conducted as well as the ones that were 
conducted by others in the past, the results of which have 
indicated that the iVotronics did not cause the undervote.
    Prior to the Sarasota testing, on November 19, 2007, we had 
visited the manufacturer of the voting machine, ES&S, in their 
Rockford, Illinois, facility and observed as they rebuilt the 
firmware from the source code that was previously held in 
escrow by Florida Division of Elections. The software that was 
rebuilt from that source code, we observed that that software 
matches the firmware that was held in escrow and that was 
certified by the Florida Division of Elections. This provides 
further confidence in the prior source code reviews that were 
conducted by a team from Florida State University and by us.
    For the ballot testing, we cast predefined test ballots on 
10 iVotronic machines and confirmed that each ballot was 
displayed and recorded accurately. The test ballots represented 
112 ways a voter may have interacted with the iVotronic to cast 
a ballot in the Florida-13 race. These test ballots were cast 
on nine machines that were configured as election day machines 
and repeated on one machine configured as an early voting 
machine.
    Finally, we conducted the calibration testing by 
miscalibrating two of the iVotronics and casting some of the 
test ballots on them. Our tests, involving 10 different 
miscalibration patterns and capturing a total of 39 votes on 
the two machines, found that, although the machines became more 
difficult to use with miscalibration, the selections that were 
displayed on the screen for the Florida-13 race, were the same 
ones that appeared on the review screen and then recorded when 
the ballots were cast.
    Based on the testings that we have conducted, we have 
obtained increased assurances that the iVotronic voting 
machines used in Sarasota County's 2006 general election did 
not contribute to the large undervote in the Florida-13 
contest. Although the test results cannot be used to provide 
absolute assurance, we believe that these test results, 
combined with the other reviews that had been conducted by the 
State of Florida, us and others, have significantly reduced the 
possibility that the iVotronic machines were the cause of the 
undervote.
    At this point, we believe that adequate testing has been 
performed on the voting machine software to reach this 
conclusion, and we do not recommend further testing in this 
area.
    Given the complex nature--complex interaction of people, 
process and technology that must effectively work together to 
achieve a successful election, we acknowledge the possibility 
that the large undervote in the Florida-13 race could have been 
caused by factors such as voters who intentionally undervoted 
or voters who may not have properly cast their ballots on the 
iVotronic machine potentially because of issues relating to 
interaction between the voters and the ballot.
    Mr. Chairman, this concludes a summary of my written 
statement. I would be happy to answer any questions that you or 
other members of the task force may have at this time.
    Thank you.
    [The statement of Mr. Barkakati follows:]

    [GRAPHIC] [TIFF OMITTED] T2709A.001
    
    [GRAPHIC] [TIFF OMITTED] T2709A.002
    
    [GRAPHIC] [TIFF OMITTED] T2709A.003
    
    [GRAPHIC] [TIFF OMITTED] T2709A.004
    
    [GRAPHIC] [TIFF OMITTED] T2709A.005
    
    [GRAPHIC] [TIFF OMITTED] T2709A.006
    
    [GRAPHIC] [TIFF OMITTED] T2709A.007
    
    [GRAPHIC] [TIFF OMITTED] T2709A.008
    
    [GRAPHIC] [TIFF OMITTED] T2709A.009
    
    [GRAPHIC] [TIFF OMITTED] T2709A.010
    
    [GRAPHIC] [TIFF OMITTED] T2709A.011
    
    [GRAPHIC] [TIFF OMITTED] T2709A.012
    
    [GRAPHIC] [TIFF OMITTED] T2709A.013
    
    [GRAPHIC] [TIFF OMITTED] T2709A.014
    
    [GRAPHIC] [TIFF OMITTED] T2709A.015
    
    [GRAPHIC] [TIFF OMITTED] T2709A.016
    
    [GRAPHIC] [TIFF OMITTED] T2709A.017
    
    [GRAPHIC] [TIFF OMITTED] T2709A.018
    
    [GRAPHIC] [TIFF OMITTED] T2709A.019
    
    [GRAPHIC] [TIFF OMITTED] T2709A.020
    
    [GRAPHIC] [TIFF OMITTED] T2709A.021
    
    [GRAPHIC] [TIFF OMITTED] T2709A.022
    
    [GRAPHIC] [TIFF OMITTED] T2709A.023
    
    [GRAPHIC] [TIFF OMITTED] T2709A.024
    
    [GRAPHIC] [TIFF OMITTED] T2709A.025
    
    [GRAPHIC] [TIFF OMITTED] T2709A.026
    
    [GRAPHIC] [TIFF OMITTED] T2709A.027
    
    [GRAPHIC] [TIFF OMITTED] T2709A.028
    
    [GRAPHIC] [TIFF OMITTED] T2709A.029
    
    [GRAPHIC] [TIFF OMITTED] T2709A.030
    
    [GRAPHIC] [TIFF OMITTED] T2709A.031
    
    [GRAPHIC] [TIFF OMITTED] T2709A.032
    
    [GRAPHIC] [TIFF OMITTED] T2709A.033
    
    [GRAPHIC] [TIFF OMITTED] T2709A.034
    
    [GRAPHIC] [TIFF OMITTED] T2709A.035
    
    [GRAPHIC] [TIFF OMITTED] T2709A.036
    
    [GRAPHIC] [TIFF OMITTED] T2709A.037
    
    [GRAPHIC] [TIFF OMITTED] T2709A.038
    
    [GRAPHIC] [TIFF OMITTED] T2709A.039
    
    [GRAPHIC] [TIFF OMITTED] T2709A.040
    
    [GRAPHIC] [TIFF OMITTED] T2709A.041
    
    [GRAPHIC] [TIFF OMITTED] T2709A.042
    
    [GRAPHIC] [TIFF OMITTED] T2709A.043
    
    [GRAPHIC] [TIFF OMITTED] T2709A.044
    
    The Chairman. The Chair is going to recognize himself for 
some preliminary questions and then, of course, turn it over to 
my colleagues.
    But, Dr. Barkakati, I have a couple of questions. First of 
all, that you were specifically charged with simply looking as 
to whether the electronic machines malfunctioned and may have 
been the culprit in the undervote; is that correct?
    Mr. Barkakati. Yes, sir.
    The Chairman. We did not ask you to look into ballot 
design?
    Mr. Barkakati. No, we did not look at the ballot design.
    The Chairman. And we did not ask you to go and look into 
either voter apathy or just whether they were totally turned 
off by the tenor of the campaign and decided to intentionally 
or deliberately not vote in Florida-13; is that correct?
    Mr. Barkakati. Yes, that is correct.
    The Chairman. In conducting your testing and the 
protocols--because, of course--again, for just general 
information, you came before the task force with your intended 
protocols as to how you were going to conduct your testing. And 
in arriving at that protocol, did you entertain suggestions, 
recommendations, basic input from all of the parties, including 
the contestant, the contestee, Florida election officials, 
Florida State University personnel that conducted some previous 
testing and their experts, as well as suggestions and 
recommendations or observations from the manufacturer of the 
voting machine?
    Mr. Barkakati. Yes, sir, we did entertain all of those 
inputs in deciding the test protocols and procedures.
    The Chairman. Okay. Now, there are many individuals that 
are interested in this for a lot of reasons. Because it goes 
beyond this particular Florida-13 contested election. And as 
much as I'd like to just contain everything to the question at 
hand and which will basically form the basis for everything 
that we do--we are not going to expand it. However, we do have 
individuals out there that have expressed great interest; and I 
have one individual who e-mailed some information to my office 
this morning. And so I do want to go ahead and ask some of the 
questions that they--that he has actually posed.
    And one would be, the personalized electronic ballots, or 
PEB cards, used to activate each voter's ballot for the DRE, 
the electronic voting machine, contains firmware separate from 
the firmware for the DRE itself. And this individual believes 
that that could have been a culprit, that could have been--
given you some insight as to whether the machines 
malfunctioned.
    Outside of that which you may have tested specifically, how 
would you address this individual's concern regarding the PEB 
card which you cover extensively in your final report?
    Mr. Barkakati. Sir, the PEB, personalized electronic 
ballot, I mean, does have a little bit of capability to 
transmit back and forth information with the iVotronic machine 
using infrared communication mechanism. However, besides that, 
it is primarily a memory device to hold data; and it holds 
initially the ballot definitions that are going to be used to 
display the ballots on the screen. However, when you open the 
machine, you copy all of those into the iVotronic and then from 
that point on the PEB is only an activator device and it can 
activate the iVotronic to get started. So there were some 
allegations that maybe the PEB was displaying different ballots 
to different people.
    However, it was copied once, and then the same ballot 
displays over and over again. And we have seen that every 
machine that was used in all the 157 precincts had votes cast 
on the Florida-13 race. So we could not have had a situation 
where it was selectively not displaying ballots.
    I mean, we have checked into that part, that the PEB could 
not have been the cause of the problem, including the fact that 
the PEB is not the one that does the primary task of the 
calculations or the displaying of the ballot. It is all done in 
the firmware that is inside the iVotronic machine itself.
    As I say, the PEB does have a small capability to send 
things back and forth and that is all, to the extent it does 
some work. So what I am saying is that we looked at the whole 
situation that the PEB could be a cause--both us as well as the 
Florida State University review team that looked at, you know, 
the same scenarios--and we concluded that it could not be the 
reason for the undervote.
    The Chairman. All right.
    And, my second question, again taken from this particular 
individual--and I am going to read specifically from the e-
mail: ``It is conceivable that--and even likely that were there 
bugs in the user interface in the Sarasota machines such as 
intermittent smoothing filter problems or other anomalies. 
These: (A) Could have occurred regardless of whether the 
software, firmware were identical; and, (B) would not have been 
uncovered without greater volume testing with actual use of the 
interface, especially in conditions resembling election day 
use.''
    Can you respond?
    Mr. Barkakati. Yes, sir.
    We actually--as you might recall, there was some mention 
that the smoothing filter--in one of those letters that came 
from, I think, ES&S to the county saying that we have a 
smoothing filter that we want to change and, you know, it will 
help with the performance of the machine in terms of 
responsiveness.
    Now, the source code review team in Florida, they--they 
said they did not find the smoothing filter. However, when we 
were looking at it--because we already knew that, you know, was 
an issue coming up, we could find certain elements that I was 
able to confirm with the manufacturer that that constituted 
what they were calling a smoothing filter.
    It essentially is very simple to explain. Specifically, the 
machine does not give you--as soon as you touch it, it does not 
assume that you have touched it yet. It records the point and 
then it waits a little bit, like 200 milliseconds or so, and 
then tries to get another data point to see that both are close 
enough. And if they are not close enough, then they assume 
that, one, maybe it is a spurious thing and then I would ignore 
it. And that is the logic built in. And that logic is what was 
called smoothing filter.
    Now, how close the two touches have to be is the threshold 
that they can play with to get it to be, like, more sensitive, 
less sensitive kind of thing. If it is like wide--as you can 
imagine, the threshold is wide--I mean, it doesn't matter if 
another touch is slightly farther off, it still assumes them to 
be the same. Whereas if you tighten them up, it takes longer, 
because the touch screens do have some noisy, you know, data 
going back and forth.
    So, specifically, the bottom line was the smoothing filter 
was not more of a mystery after that. I can understand why it 
was there, and we figured it out, and the proposal they had 
made was to simply to widen the threshold to make it more 
responsive.
    Regardless of all of that, we were looking at a source code 
that handles the display and the recording of the vote. That is 
not going to be affected at all by the filter--other than the 
machine's response might be slow. In other words, it might take 
a little time to get the mark to show up.
    But the machine isn't going to suddenly misbehave because 
it is slightly slower in response, and that actually almost 
explains the other element to which you might have read, like 
nondeterministic behavior.
    Well, because the touch screen--I mean, depending on the 
time, some of the conditions--it may, you know, have other 
spurious things going on whereby that filter that I talked 
about, the closeness of two touches, maybe it's not being met. 
And if it is not being met, the condition is not met, then the 
system waits a little bit longer to get another touch, you 
know. So it looks like it--is it waiting for me to get--you 
know, accept my input. That is what the user feels like.
    But the bottom line of all of that is, regardless of all of 
that, that it isn't going to affect anywhere at all in terms of 
what selection is displayed to the user and what selection is 
recorded. From that we are concluding that, essentially, if the 
machine is displaying a selection, that will be the one 
displayed on the review screen and recorded when you press the 
vote button; and that is the bottom line that we are going 
after in the technological study.
    So I guess it is a very long way of explaining, but I think 
we did understand--we looked into all of that and we understood 
what it was, smoothing filter was, and what the company was 
intending to do when they sent the letter and what its effect 
might have been. And, regardless of all of that, we were able 
to confirm sort of like the same thing, that the source code 
review from--team from Florida State University had said, it 
wasn't going to affect the vote display and recording of the 
machine.
    The Chairman. Thank you very much, Doctor.
    I will recognize Congressman McCarthy at this time for his 
questions.
    Mr. McCarthy. Thank you, Mr. Chairman.
    In reading your report and listening to you when we first 
started out and talking to you about whether you could do this 
work and setting out some parameters to look at it, you told us 
that the closest statistically that anyone can get is 99 
percent. Now, in hearing your report, you say you are 99 
percent sure every single vote that was voted has been counted 
and counted correctly. Is that true?
    Mr. Barkakati. I guess it is a bit nuanced in the sense 
that the statistical portion of it applies to the--confirming 
that every machine was running the same software. So that was--
that does have a 99 percent confidence level, that most 
machines are running the same software.
    The remaining part is based on a logic, essentially, kind 
of like an assembly line or something. When you have a lot of 
the widgets being built, you can take one and sample and it is 
okay because everybody is the same. Extending that logic, we 
are able to say that, while we test a small number of machines, 
they record votes properly and display things properly and 
therefore that applies to the whole population. And then taking 
the abundance of all the testing that has been done in the past 
by the State of Florida and others, we are basically forming 
the judgment which is what we say, that we collectively think 
everything has been done to eliminate the machine as the cause 
of the undervote.
    So, in a way, you are getting the answer, but the 99 
percent number in a nuanced way applies only to the condition 
that the firmware is the same. But the rest should be--since 
that is our--basically and logically, as I explained, was the 
reason behind coming up with a conclusion that adequate testing 
has been done to eliminate the machine. So, essentially, you 
are getting the conclusion without the 99 percent number 
associated with it because statisticians are not going to let 
me say it that way.
    Mr. McCarthy. So in my world we say 100 percent, but I 
understand.
    And I will tell you--and having been an individual that 
worked on staff on some other contested races, this one has 
been more thoroughly investigated than any one I have seen. You 
had an independent Florida State review prior to it ever coming 
to Congress' review, and they were--we even had some of the 
individuals that did the research on that before us testify at 
other times. And I found that to be very thorough and very 
correct.
    In reading your report, you even said you miscalibrated the 
machines, and they still worked properly.
    Mr. Barkakati. Yes. The miscalibration was really meant to 
be a small amount and a larger amount of miscalibration to see 
the effect of it, essentially. And what we observed is that--
and wanted to confirm that, even if miscalibrated, if it is 
showing you something as the selection you are making, then 
that would be the selection that will eventually get recorded 
in the memory inside the machine. That is what our, you know, 
goal was.
    And we did find that all of the machines, even with a 
smaller and larger amount of miscalibration--of course, it gets 
hard to use and they are so obvious in terms of bigger 
miscalibrations, become so obvious that--you probably will--
officials have told us that they would probably have found that 
right away and put it out of service in any case.
    But, regardless--and coming back to your question--we did 
do that, and we found that in each case they definitely were 
displaying--if you can make any selection on the screen, that 
is the selection that is going to be essentially your choice 
that will be recorded.
    So we confirmed that, yes.
    Mr. McCarthy. That is very interesting, really, to the 
American public, too. Because Mr. Ehlers always points out to 
us, even when you went back and you just did paper and you 
would recount, you would always find one or two differences 
because you had human error touching it a number of times. And 
here we had individuals researching numerous different ways in 
the computer, and it came out where no vote was different, and 
it all worked properly. So that is nice to know.
    Because I think our goal here was set out first when the 
challenger came in and wanted the election overturned and we 
went through this research here. We have found time and time 
again that every vote in the 13th district of Florida was 
counted properly and was put forward. So that is really a 
testament to the American public and to this district that they 
know their election was honest, true and correct. And I applaud 
the work you have done and the work that the others have done 
prior because we have never researched it this thoroughly, and 
I thank you.
    The Chairman. Thank you very much, Mr. McCarthy.
    The Chair will recognize Ms. Lofgren.
    Ms. Lofgren. Mr. Chairman, you asked the questions that I 
was going to ask, so I am happy for that.
    I would just note that on page 12 of the report the GAO 
does make a suggestion that a voter-verified paper trail could 
provide an independent confirmation. And, obviously, we can't 
create that retroactively. But, looking forward, I think that 
is something that we hope that the jurisdictions will look to 
do; and we will be hopefully in the near future pursuing that 
kind of endeavor at the Federal level.
    So, as with the other members of the task force, I--there 
was a big undervote. We don't know why. I will say that when I 
looked at the replica of the ballot on page 10 I didn't notice, 
because of the way it is set out, that there was actually an 
election above the State. So I--we will never know. But I think 
that that had an impact here.
    I think that it is important that we have discharged our 
duty here with the help of GAO in a way that has been 
nonpartisan really. Because our obligation is not as Democrats 
or Republican. Our obligation is to find out as best we can 
what happened and to make a decision based on that alone. And I 
think that is what we have done, and that is what we should 
have done. And I feel very satisfied with the process, and I am 
eager to put this behind us and get on to other business.
    So I don't know if there are other questions. At the 
appropriate time, I have a motion to make.
    The Chairman. Thank you very much, Ms. Lofgren.
    And the Chair will recognize Mr. Ehlers, the ranking member 
of the full committee.
    Mr. Ehlers. Thank you, Mr. Chairman; and I will be very 
brief.
    First of all, I just want to thank the GAO that did a very 
thorough job under very difficult circumstances. I commend you 
for your work.
    The other comment I would make is to agree with Ms. Lofgren 
about the ballot design. I had exactly the same experience. 
When I looked at the ballot, I actually missed the 
congressional race. I think this is a valuable experience, too, 
in terms that has been transmitted nationwide. I think every 
county clerk, city clerk, township clerk, and Secretary of 
State will be looking at ballot design more carefully in the 
future.
    With that, I yield back.
    The Chairman. Thank you very much, Mr. Ehlers; and the 
Chair recognizes Mr. Lungren.
    Mr. Lungren. Thank you, Mr. Chairman.
    Dr. Barkakati, I know statisticians don't like to say 
certain things and so forth, but is it fair to say--to 
summarize in part what you said, that based on the work that 
your organization has done and based on your review of those 
other organizations that did testing, you do not see a need for 
any additional testing or are not suggesting to this committee 
that we request any additional testing to answer the questions 
you were requested to answer?
    Mr. Barkakati. Yes, I agree with your statement. We are not 
suggesting any further testing based on all the things we have 
seen so far and done so far.
    Mr. Lungren. I thank you. And I join my colleagues in 
suggesting that the ballot design is something that needs to be 
looked at.
    But I also may be the only one willing to confess that I 
have gone to the ballot at times and intentionally not voted in 
certain elections to show my protest over the way the campaign 
was conducted. And I just note that this was a hard-fought 
campaign not only in the general, but in the primary, and 
voters sometimes respond in that way. And I think we ought to 
recognize--or at least I will admit I have been one that has 
done that. And I have been in similar situations. Maybe you 
would be investigating as to why I didn't vote.
    But I thank the chairman once again for the professional 
way in which he has handled this and all the members of the 
task force.
    The Chairman. Thank you very much, Mr. Lungren.
    Now I am going to finish up with a couple of questions, one 
that was actually posed by a representative for the contestant. 
And even though this was not your charge, and we understand 
that. And I don't want to get too far afield. But there have 
been some--obviously, we all have our take on what may have 
happened, ballot design or simply [a] voter that was simply 
turned off. Probably more ballot design, if we look at it 
somewhat objectively.
    And the reason for that is this only--the undervote was 
experienced in this degree--or extent in Sarasota. So it 
couldn't be just the apathy or the intolerance or the disgust 
stopped at the county line and--nor was it reflected, I 
believe, in the absentee voting or those votes that were cast 
by methods other than the ES&S machine.
    So the question is as follows: Has the GAO reached any 
conclusions about whether the unusually elevated undervote rate 
was due to intentional undervoting or to unintentional human 
factors such as voter confusion caused by poor ballot design?
    Mr. Barkakati. Unfortunately, we didn't do anything to 
determine between those two what might be the reason, except 
for knowing the other explanations. People have stated that 
ballot design might be an issue. We are aware of those kind of 
things. But we did not really evaluate to kind of try to 
eliminate one versus the others. So, unfortunately, I don't 
think we have any conclusive statement about whether it was the 
ballot design that might have caused it or the intentional 
undervoting that caused it. That is probably all----
    I mean, I am aware of other--we have mentioned in a past 
statement that there were other, I think, experiments being 
done on humans--using humans to see--or human subjects to see 
whether they can miss it or not by, for instance, a professor 
at MIT that we know of. Ted Selker is the name. And then, for 
that one, we know that the work is proceeding on that area to 
find out whether ballot design might have been the reason, but 
he hasn't finished the whole results, I guess, yet. And he was 
working with ES&S machines actually, using the same ballot 
layout and all. So there may be some report results coming out 
from that in the future which might provide, you know, more 
light into this area.
    The Chairman. Because I think people still want some 
answers. And, of course, again, that wasn't the central 
allegation that we had to resolve here before the committee.
    One last observation, of course, is, this really does point 
out--it goes back to ballot design. It doesn't matter how 
sophisticated and reliable the voting machine may be. It could 
even have a paper trail. But at the end of this whole process 
is that ballot design many times can be confusing. And we know 
that Florida has more or less been the poster child, but they 
have moved forward. But still, ballot design could still 
present a real, real problem as to whether it was the butterfly 
ballot in 2000 or whether it was the 2006 Florida-13 election.
    So we understand the tremendous challenge that local 
election officials have; and we would just caution them to 
take, again, every precaution out there that they possibly 
could regarding how they designed that ballot for presentation.
    The last thing I want to say is that every candidate for 
office has a right, when they are running for the United States 
House of Representatives, to challenge the validity of the 
election if, obviously, they were not the victor. And that is 
their right. And then it is the constitutional duty of the 
United States Congress, then, to pass on whether someone is 
going to be seated and sworn in as a Member of the House. And I 
think that was our charge, that was our duty. We want to thank 
everyone that assisted us in performing that duty.
    At this time, I am going to recognize Ms. Lofgren for the 
purpose of making a motion.
    Ms. Lofgren. Mr. Chairman, I move that the chairman be 
authorized and directed to report to the committee that the 
task force has completed its investigation related to the 
election of a representative from the 13th Congressional 
District of Florida to the House of Representatives, and I move 
further that the chairman report to the committee the task 
force's recommendation that the election contest in the 13th 
District of Florida be dismissed.
    [The information follows:]

    [GRAPHIC] [TIFF OMITTED] T2709A.045
    
    The Chairman. And at this time, task force members, if they 
are in favor, will register by signaling aye. Opposed, nay.
    It's unanimous, the ayes; and we will proceed with this 
formal adoption of the motion and report it to the full 
committee for its consideration at a later date.
    And, with that, we stand adjourned.
    [Whereupon, at 10:49 a.m., the briefing was adjourned.]
