[House Hearing, 106 Congress]
[From the U.S. Government Publishing Office]




 KYOTO AND THE INTERNET: THE ENERGY IMPLICATIONS OF THE DIGITAL ECONOMY

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

                                HEARING

                               before the

               SUBCOMMITTEE ON NATIONAL ECONOMIC GROWTH,
               NATURAL RESOURCES, AND REGULATORY AFFAIRS

                                 of the

                              COMMITTEE ON
                           GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED SIXTH CONGRESS

                             SECOND SESSION

                               __________

                            FEBRUARY 2, 2000

                               __________

                           Serial No. 106-125

                               __________

       Printed for the use of the Committee on Government Reform


  Available via the World Wide Web: http://www.gpo.gov/congress/house
                      http://www.house.gov/reform



                                 ______
                     COMMITTEE ON GOVERNMENT REFORM

                     DAN BURTON, Indiana, Chairman
BENJAMIN A. GILMAN, New York         HENRY A. WAXMAN, California
CONSTANCE A. MORELLA, Maryland       TOM LANTOS, California
CHRISTOPHER SHAYS, Connecticut       ROBERT E. WISE, Jr., West Virginia
ILEANA ROS-LEHTINEN, Florida         MAJOR R. OWENS, New York
JOHN M. McHUGH, New York             EDOLPHUS TOWNS, New York
STEPHEN HORN, California             PAUL E. KANJORSKI, Pennsylvania
JOHN L. MICA, Florida                PATSY T. MINK, Hawaii
THOMAS M. DAVIS, Virginia            CAROLYN B. MALONEY, New York
DAVID M. McINTOSH, Indiana           ELEANOR HOLMES NORTON, Washington, 
MARK E. SOUDER, Indiana                  DC
JOE SCARBOROUGH, Florida             CHAKA FATTAH, Pennsylvania
STEVEN C. LaTOURETTE, Ohio           ELIJAH E. CUMMINGS, Maryland
MARSHALL ``MARK'' SANFORD, South     DENNIS J. KUCINICH, Ohio
    Carolina                         ROD R. BLAGOJEVICH, Illinois
BOB BARR, Georgia                    DANNY K. DAVIS, Illinois
DAN MILLER, Florida                  JOHN F. TIERNEY, Massachusetts
ASA HUTCHINSON, Arkansas             JIM TURNER, Texas
LEE TERRY, Nebraska                  THOMAS H. ALLEN, Maine
JUDY BIGGERT, Illinois               HAROLD E. FORD, Jr., Tennessee
GREG WALDEN, Oregon                  JANICE D. SCHAKOWSKY, Illinois
DOUG OSE, California                             ------
PAUL RYAN, Wisconsin                 BERNARD SANDERS, Vermont 
HELEN CHENOWETH-HAGE, Idaho              (Independent)
DAVID VITTER, Louisiana


                      Kevin Binger, Staff Director
                 Daniel R. Moll, Deputy Staff Director
           David A. Kass, Deputy Counsel and Parliamentarian
                    Lisa Smith Arafune, Chief Clerk
                 Phil Schiliro, Minority Staff Director
                                 ------                                

   Subcommittee on National Economic Growth, Natural Resources, and 
                           Regulatory Affairs

                  DAVID M. McINTOSH, Indiana, Chairman
PAUL RYAN, Wisconsin                 DENNIS J. KUCINICH, Ohio
BOB BARR, Georgia                    TOM LANTOS, California
LEE TERRY, Nebraska                  PAUL E. KANJORSKI, Pennsylvania
GREG WALDEN, Oregon                  BERNARD SANDERS, Vermont
HELEN CHENOWETH-HAGE, Idaho          HAROLD E. FORD, Jr., Tennessee
DAVID VITTER, Louisiana

                               Ex Officio

DAN BURTON, Indiana                  HENRY A. WAXMAN, California
                    Marlo Lewis, Jr., Staff Director
              Barbara F. Kahlow, Professional Staff Member
                       Gabriel Neil Rubin, Clerk
                 Elizabeth Mundinger, Minority Counsel
                            C O N T E N T S

                              ----------                              
                                                                   Page
Hearing held on February 2, 2000.................................     1
Statement of:
    Hakes, Jay, Administrator, Energy Information Administration, 
      Department of Energy; Joseph Romm, executive director, 
      Center for Energy and Climate Solutions, and former Acting 
      Assistant Secretary for Energy Efficiency and Renewable 
      Energy, Department of Energy; and Mark Mills, senior 
      fellow, Competitive Enterprise Institute, and scientific 
      advisor, Greening Earth Society............................    12
Letters, statements, et cetera, submitted for the record by:
    Chenoweth-Hage, Hon. Helen, a Representative in Congress from 
      the State of Idaho, prepared statement of..................     2
    Hakes, Jay, Administrator, Energy Information Administration, 
      Department of Energy, prepared statement of................    14
    Kucinich, Hon. Dennis J., a Representative in Congress from 
      the State of Ohio, prepared statement of...................     9
    McIntosh, Hon. David M., a Representative in Congress from 
      the State of Indiana, prepared statement of................     5
    Mills, Mark, senior fellow, Competitive Enterprise Institute, 
      and scientific advisor, Greening Earth Society, prepared 
      statement of...............................................    58
    Romm, Joseph, executive director, Center for Energy and 
      Climate Solutions, and former Acting Assistant Secretary 
      for Energy Efficiency and Renewable Energy, Department of 
      Energy, prepared statement of..............................    40

 
 KYOTO AND THE INTERNET: THE ENERGY IMPLICATIONS OF THE DIGITAL ECONOMY

                              ----------                              


                      WEDNESDAY, FEBRUARY 2, 2000

                  House of Representatives,
 Subcommittee on National Economic Growth, Natural 
                 Resources, and Regulatory Affairs,
                            Committee on Government Reform,
                                                    Washington, DC.
    The subcommittee met, pursuant to notice, at 10:30 a.m., in 
room 2247, Rayburn House Office Building, Hon. David M. 
McIntosh (chairman of the subcommittee) presiding.
    Present: Representatives McIntosh, Ryan, and Kucinich.
    Staff present: Marlo Lewis, Jr., staff director; Barbara F. 
Kahlow, professional staff member; Bill Waller and Heather 
Henderson, counsels; Gabriel Neil Rubin, clerk; Elizabeth 
Mundinger, minority counsel; and Jean Gosa, minority staff 
assistant.
    Mr. Ryan [presiding]. A quorum being present, the 
Subcommittee on National Economic Growth, Natural Resources, 
and Regulatory Affairs will come to order. I ask unanimous 
consent that all Members' and witnesses' written opening 
statements be included in the record. Without objection, so 
ordered.
    I ask unanimous consent that all articles, exhibits, 
extraneous or tabular material referred to be included in the 
record. Without objection, so ordered.
    I also ask unanimous consent that the record be left open 
for 10 days for any additional statements to be included in the 
record. Without objection, so ordered. I understand that my 
colleague, Ms. Chenoweth-Hage, is attending a mark up and may 
not be able to join us today. However, she has asked that a 
prepared statement of hers be entered into the record. Without 
objection, so ordered.
    [The prepared statement of Hon. Helen Chenoweth-Hage 
follows:]
[GRAPHIC] [TIFF OMITTED] T4876.001

    Mr. Ryan. Today's hearing will examine how the burgeoning 
digital economy, comprised of e-commerce and the information 
technology industries that make e-commerce possible may change 
energy trends in the U.S. economy and how such changes may 
affect the cost and feasibility of the Kyoto protocol.
    In March 1999, Vice President Gore stunned some of us and 
amused others by claiming to have created the Internet. 
``During my service in the U.S. Congress, I took the initiative 
in creating the Internet,'' Mr. Gore said during an interview 
with CNN's Wolf Blitzer. Now, whether or not Al Gore is the 
father of the Internet, he is unquestionably the father of the 
Kyoto protocol. Mr. Gore led the United States negotiating team 
at the December 1997 Climate Change Conference in Kyoto, Japan. 
During his service in the Senate and at the White House, Al 
Gore took the initiative in creating a legally binding 
international climate treaty.
    So, one might say that today's hearing will examine the 
relationship between the Vice President's brain children. Will 
the digital economy facilitate Kyoto-style energy restriction 
policies by decreasing the energy intensity of the U.S. 
economy? Or, will the digital economy sweep the Kyoto protocol 
into the dust bin of history by increasing United States and 
global demands for electric power?
    The digital economy is growing at a phenomenal pace. For 
example, during the past 2 years alone, the number of web users 
worldwide increased by 55 percent, the number of web servers 
increased by 128 percent, and the number of new web address 
registrations rose by an astounding 137 percent. The U.S. 
Department of Commerce's recent publication, the Emerging 
Digital Economy II, forecasts that by 2006, almost half the 
U.S. work force will be employed by industries that are either 
major producers or intensive users of information technology 
products and services. The Commerce Department study also 
reports that information technology industries contributed over 
one-third of the Nation's real economic growth from 1998 to 
1999.
    In thinking about this key driver of U.S. economic 
performance, I can't help noticing that the digital economy 
runs solely on electricity. More than half of U.S. electricity 
is produced from coal. Coal is the fuel source targeted for 
extinction by the Kyoto protocol. So, is there not a 
fundamental incompatibility between the energy requirements of 
a digital economy and the Kyoto protocol? Can we really wire 
the world and, at the same time, restrict United States and 
global access to abundant, affordable and reliable electric 
power?
    Our witnesses, I am aware, have different answers to those 
questions. We encourage this, and we are excited about hearing 
your testimony. Yet, there appear to be some broad areas of 
agreement. The experts agree that the digital economy is 
enormously important, enormously large, and growing 
exponentially. They also agree that Internet forecasting is 
fraught with uncertainty. As the Commerce Department has 
observed, this emerging digital economy regularly surprises 
those who study it most closely. What this suggests to me is 
that we should proceed with caution in considering any policy 
that may damage or restrict the electricity supply that powers 
the digital economy. Indeed, since the digital economy is still 
very much in its infancy, we should probably move more slowly 
and cautiously in considering a treaty like the Kyoto protocol 
than at any other time.
    Before turning to my colleague, Mr. Kucinich, for his 
opening statement, I would like to introduce our witnesses: The 
Honorable Jay Hakes, Administrator of the Energy Information 
Administration [EIA], of the U.S. Department of Energy. Jay, I 
believe, will be speaking first. Mr. Hakes will present EIA's 
analysis of energy trends and the digital economy. EIA 
testified before this subcommittee last year on a related 
issue. Welcome back, Mr. Hakes.
    Mr. Hakes. Thank you.
    Mr. Ryan. Dr. Joseph Romm, executive director of the Center 
for Climate and Energy Solutions and a former Clinton 
administration Acting Assistant Secretary for Energy Efficiency 
and Renewable Energy, will speak next. Welcome, Dr. Romm.
    Mr. Romm. Thank you.
    Mr. Ryan. Dr. Romm will present the case that the Internet 
economy is breaking the historic link between economic growth 
and energy demand growth. If I am not mistaken, Dr. Romm 
believes that Internet efficiencies will enable us to implement 
the Kyoto protocol without economic pain. Thank you for 
testifying, Dr. Romm.
    Our final witness is Mark Mills, senior fellow at the 
Competitive Enterprise Institute and scientific advisor to the 
Greening Earth Society. Mr. Mills will present the case that 
the digital economy is fueling a surge in demand for cheap, 
abundant and reliable electric power. He views the Kyoto 
protocol as a threat to the digital economy. To my knowledge, 
Mr. Mills was the first analyst to raise the topic of today's 
discussion. Thank you for participating in this hearing, Mr. 
Mills.
    Mr. Mills. Thank you.
    [The prepared statement of Hon. David M. McIntosh follows:]
    [GRAPHIC] [TIFF OMITTED] T4876.002
    
    [GRAPHIC] [TIFF OMITTED] T4876.003
    
    Mr. Ryan. Mr. Kucinich.
    Mr. Kucinich. Good morning, Mr. Chairman.
    Mr. Ryan. Good morning.
    Mr. Kucinich. Nice to see you. I appreciate the chance to 
participate with this panel, with you in the chair.
    As you know, my approach in this Congress has been 
bipartisan, and I think that is a good way to proceed in trying 
to get to the truth of all of the matters into which we 
inquire, and I think that one of the things that we may be able 
to do in future Congresses, anticipating that we are both going 
to be here, is to find a way to diminish the partisan rancor 
and find ways where we can work together.
    Now, we may have differences of opinion on this particular 
hearing. I would like to say that I consider Vice President 
Gore one of the most visionary leaders we have ever had in 
American government. He has corrected his misstatement about 
whether he was the--about being the creator of the Internet, 
but I think that everyone understands that his involvement in 
the development of new technologies cannot be disputed. I also 
know that as someone who is the author of a book called Earth 
in the Balance, it was one of the most visionary statements on 
the concern for the global and the U.S. environment.
    So I just want that said, because to let this hearing begin 
on a note where the Vice President is challenged I think 
requires a response, not simply in order to defend his record, 
which speaks for itself, it can stand without my humble 
efforts, but because I think that the nature of what we are 
discussing today is so significant that it shouldn't be 
minimized by setting a tone, which I think we are all capable 
of going beyond.
    I really appreciate having this hearing on the impact of 
the Internet on energy demand and our production of greenhouse 
gases.
    The Internet economy is growing at an explosive rate. 
Studies estimate that it may grow more than tenfold, from the 
tens of billions of dollars to perhaps trillions of dollars. 
Our economy is already reaping the benefits. The question 
remains, what does this mean for the environment?
    Historical patterns indicate that economic growth comes at 
the expense of the environment. Economic growth and energy use 
rose and fell in tandem. However, the rules seem to be 
changing. In 1997 and 1998, our economy soared, yet energy use 
and our emission of greenhouse gases rose at a much slower rate 
than was indicated by historical pattern. Apparently, our 
economy is becoming more energy efficient.
    I think President Clinton made that point in his State of 
the Union when he spoke about how perhaps in this next century 
we can finally put to rest this mythology that you cannot have 
economic growth and a cleaner environment at the same time. So 
we are challenged in this Congress to keep unfolding this new 
thinking, which I think we are capable of doing.
    I think there is good reason to believe the Internet will 
help us become even more energy efficient. E-businesses require 
less retail space, relying instead on warehouses. This saves 
energy otherwise spent on construction. Businesses and 
individuals do not need to store as much information on paper, 
which saves energy otherwise spent producing paper and building 
storage space. Businesses will likely obtain more accurate 
consumer information, which saves energy otherwise spent 
building storage space for unused inventory. Consumers save the 
energy used to drive to and from the mall when they order 
products on line, and telecommuting saves energy otherwise 
expended on driving to work and constructing office space.
    These benefits apparently come at little cost because the 
Internet is not energy intensive. According to the Energy 
Information Administration, personal computers in residences 
and business accounted for less than 2 percent of our 
electricity use in 1999. Much of this energy was used to power 
other aspects of computer use, not just the Internet. And we 
did not see a substantial increase in energy use in 1997 and 
1998 when Internet use rose dramatically. In fact, energy use 
increased at a much slower rate than expected. In his written 
testimony, Mr. Hakes testifies that ``it is clear that the size 
of Internet electricity use today is small compared to that of 
all other uses of electricity.''
    If the Internet does provide great energy benefits at 
little cost, we may need to rethink some of our earlier 
analyses of the cost of complying with the Kyoto protocol. A 
working paper prepared at the Environmental Protection Agency 
found that information and communication technology ``may 
prompt large structural changes that can reduce overall energy 
consumption.'' These changes could potentially significantly 
reduce mainstream estimates of energy use and carbon emissions, 
such as the Energy Information Administration's analysis of the 
cost of the Kyoto protocol. In other words, we may have greatly 
overestimated the cost of complying with the Kyoto protocol.
    I think that one of the things that we should also consider 
is that as we move toward this tremendous explosion of new 
thought in this new century, that we will most likely be 
ushering in alternative energy strategies. It is inevitable. We 
could not foresee at the beginning of the 19th century the 
great developments in transportation and energy and science. I 
think that we have an understanding of how this impulse to 
create is so powerful in our society, and I think we can 
anticipate that there will be dramatic developments in 
information technologies, which in some ways will make this 
debate itself hopefully obsolete. But it is not obsolete at the 
moment, I respect that, and I appreciate the opportunity to 
proceed. I think it is an important issue.
    I look forward to hearing from the witnesses today, and I 
would ask the Chair unanimous consent to hold the record open 
for relevant materials.
    Mr. Ryan. Without objection.
    [The prepared statement of Hon. Dennis J. Kucinich 
follows:]
[GRAPHIC] [TIFF OMITTED] T4876.004

[GRAPHIC] [TIFF OMITTED] T4876.005

[GRAPHIC] [TIFF OMITTED] T4876.006

    [Witnesses sworn.]
    Mr. Ryan. I will start with Mr. Hakes. Mr. Hakes, thank 
you.

  STATEMENTS OF JAY HAKES, ADMINISTRATOR, ENERGY INFORMATION 
 ADMINISTRATION, DEPARTMENT OF ENERGY; JOSEPH ROMM, EXECUTIVE 
 DIRECTOR, CENTER FOR ENERGY AND CLIMATE SOLUTIONS, AND FORMER 
ACTING ASSISTANT SECRETARY FOR ENERGY EFFICIENCY AND RENEWABLE 
 ENERGY, DEPARTMENT OF ENERGY; AND MARK MILLS, SENIOR FELLOW, 
   COMPETITIVE ENTERPRISE INSTITUTE, AND SCIENTIFIC ADVISOR, 
                     GREENING EARTH SOCIETY

    Mr. Hakes. Thank you, Mr. Chairman. It is a pleasure to be 
before the Congress talking about something other than high oil 
prices this week.
    The Energy Information Administration estimates that if 
current Federal policies remain in place, electricity use in 
the United States over the next two decades is likely to grow 
by about 1.5 percent a year. While no one can estimate growth 
in electricity with exactitude, this number does provide a 
useful benchmark for people who study energy issues.
    A lively debate has emerged over whether consumption of 
electricity is likely to rise more slowly or more rapidly than 
the EIA projects. One camp represented today argues the former 
and another the latter, and they can testify why they have 
reached these conclusions.
    The points raised by these analysts certainly have some 
merit. Most of them, however, are already incorporated in the 
EIA analysis, although not at a scale that would satisfy either 
of the camps. Several of our assumptions would seem to reflect 
the trends they are talking about; for example, a steady shift 
to less energy intensive industries, new equipment in the 
industrial sector that is much more efficient than the stock it 
replaces, penetration of more efficient equipment in the 
residential and commercial sectors, and a growth of 3.2 percent 
a year in a category called ``other uses of electricity,'' 
which would allow for considerable growth in electric load 
brought about by computers and the Internet.
    While it is certainly not possible to rule out the slower 
electronic growth or the faster electric growth arguments, we 
don't find either convincing enough to suggest that the EIA 
should be altered in either direction. In the first place, as 
seen from the list above, the arguments haven't been ignored in 
the EIA analysis and, in addition, to some extent, they offset 
each other.
    The slow energy growth camp must deal with expectations 
that lower prices for electricity and rapid growth in personal 
disposable income will create a climate in which demand for new 
energy services is likely to grow and the priority given to 
reducing fuel costs is not likely to be high. The rate of stock 
turnover limits the speed with which many more efficient pieces 
of equipment can enter the market and many of the changes 
suggest that they are less dramatic than they might appear.
    For instance, as web-based shopping grows, much of the 
displacement will be of catalogs and 800 telephone numbers. 
There will still be some reduction in energy use, but trucks 
will still be needed for their frequently long journey from the 
warehouse to the home. The warehouses the products come from 
are likely to be smaller, but the homes they go to are likely 
to be larger.
    The high energy growth camp has to deal with the fact that 
computer equipment of whatever type tends to have very low 
electric load compared to space heating and cooling and other 
more traditional equipment. This can be seen in the graph which 
is figure 1 of my prepared testimony. Those of you who have the 
testimony can find this as figure 1, but on the graph here, you 
can see that PCs are estimated to be 2 percent of the electric 
load in residential buildings. Now, as other witnesses will 
point out, there is some fuzziness to the data, and maybe that 
is not perhaps an exact number. But we do know a lot about 
these other uses and we know the total electric consumption. 
Say we are off by some magnitude here, we are still not talking 
about something that is a heavy load in the home. At home I now 
listen to more CDs, music on my computer, but that means I am 
using my stereo system less, which actually uses more watts.
    PCs and other electronics are growing rapidly, but still a 
small part of overall consumption, and this is shown in a later 
graph. This is all buildings, commercial and residential. Some 
of what Mr. Mills is talking about is in the PC category and 
some of it is in commercial office equipment. You can see that 
the rates of growth are very, very high there, much above 
average, but the BTUs growth is relatively small, because it is 
just a small part of the electrical load.
    I would also say that we do question some of the 
calculations in the high-growth camp on the load of these 
pieces of electrical equipment.
    Future trends in energy efficiency and expanded demand for 
energy services are particularly difficult to anticipate. It is 
hard often to get good data. But in summary, we continue to 
believe that our estimates on the growth of electricity are 
good base case projections under current policies for at least 
a couple of reasons. One, we are not seeing either of the 
alternate growth paths reflected in the current data. Second, 
the advances in technology that accelerate efficiency are often 
the same as those that accelerate demand for energy services. 
Just look at the automobile as an example of that.
    For instance, if technology progresses more rapidly than 
projected, it will likely spur both greater efficiency and 
greater demand for new energy services, so the two results 
tend, to some extent, to offset each other.
    I would also hope in the question period I will have an 
ample opportunity to discuss the testimony of my former 
colleague, Mr. Romm. His characterizations of EIA's previous 
work on Kyoto and other subjects I believe are highly 
misleading. These mistakes have been pointed out rather fully 
in the public record, and I am somewhat puzzled why they 
continue to appear in the public record. So I hope that in the 
question and answer period we will have a chance to fully 
discuss some of the points he raises in his testimony.
    [The prepared statement of Mr. Hakes follows:]
    [GRAPHIC] [TIFF OMITTED] T4876.007
    
    [GRAPHIC] [TIFF OMITTED] T4876.008
    
    [GRAPHIC] [TIFF OMITTED] T4876.009
    
    [GRAPHIC] [TIFF OMITTED] T4876.010
    
    [GRAPHIC] [TIFF OMITTED] T4876.011
    
    [GRAPHIC] [TIFF OMITTED] T4876.012
    
    [GRAPHIC] [TIFF OMITTED] T4876.013
    
    [GRAPHIC] [TIFF OMITTED] T4876.014
    
    [GRAPHIC] [TIFF OMITTED] T4876.015
    
    [GRAPHIC] [TIFF OMITTED] T4876.016
    
    [GRAPHIC] [TIFF OMITTED] T4876.017
    
    [GRAPHIC] [TIFF OMITTED] T4876.018
    
    [GRAPHIC] [TIFF OMITTED] T4876.019
    
    [GRAPHIC] [TIFF OMITTED] T4876.020
    
    [GRAPHIC] [TIFF OMITTED] T4876.021
    
    [GRAPHIC] [TIFF OMITTED] T4876.022
    
    [GRAPHIC] [TIFF OMITTED] T4876.023
    
    [GRAPHIC] [TIFF OMITTED] T4876.024
    
    [GRAPHIC] [TIFF OMITTED] T4876.025
    
    [GRAPHIC] [TIFF OMITTED] T4876.026
    
    [GRAPHIC] [TIFF OMITTED] T4876.027
    
    [GRAPHIC] [TIFF OMITTED] T4876.028
    
    [GRAPHIC] [TIFF OMITTED] T4876.029
    
    Mr. McIntosh [presiding]. Thank you, Mr. Hakes. We will 
certainly afford you that opportunity during that period.
    Our next witness is Mr. Romm. If you could summarize your 
written testimony for us, that would be great.
    Mr. Romm. Mr. Chairman and members of the subcommittee, I 
am delighted to appear before you today. Let me start with 
three things that we know for sure. Mr. Mills is wrong; EIA's 
key forecasts are often wrong; and something very unusual is 
happening with the way the United States uses energy. Let's 
start with Mr. Mills.
    It is rare that we know so conclusively that someone is as 
wrong as Mr. Mills. Let me present the results of a detailed 
study by five scientists at Lawrence Berkeley National 
Laboratory. On the left is current electricity used by the 
Internet according to Mr. Mills, the revision by Lawrence 
Berkeley National Laboratory on the right. Mills now repeatedly 
overestimates the electricity consumed by different components 
of the Internet. Mills is wrong by at least a factor of 8. He 
makes a variety of methodological and analytical errors that I 
would be happy to go into during the question and answer. He 
keeps ascribing all of the electricity used by a computer for 
all purposes, even if it only spends a few hours on the 
Internet. But the main reason we know Mr. Mills is wrong is, as 
Dr. Hakes said, electricity consumption hasn't exploded in the 
last few years, and I will return to this point later.
    On EIA, EIA is quite good at collecting and analyzing data. 
It is very bad at forecasting. In a 1996 Science Committee 
hearing, Dr. Hakes and I were witnesses. Chairman Dana 
Rohrabacher asked him, ``Dr. Hakes, you admit in your testimony 
that your forecasts have been off, and I would say they have 
been off not just a small amount, but they have been off the 
chart. How come your predictions have been so far off?'' Dr. 
Hakes replied, ``It is the price side where there has been the 
greatest error in areas like predicted demand and consumption. 
I think basically a lack of understanding of the impact of 
decontrol of the market. Second, I think a lot of people, 
including us, underestimated the impact of technologies.''
    Dr. Hakes testified that such errors were in the past, but 
that is not the case. Fourteen months ago, EIA projected that 
the world price of oil in the year 2000 would be $13.97 a 
barrel. That looks to be wrong by a factor of 2. EIA also 
predicted that total U.S. carbon emissions would rise by 105 
million tons from 1997 to 2000. That also looks to be wrong by 
more than a factor of 2. EIA's analysis of the Clean Air Act 
overestimated the cost of sulfur permits by a factor of 4. Why? 
They underestimated technology improvements, fuel switching and 
the benefits of railroad competition.
    I believe their estimates of the costs of Kyoto are also 
wrong by a factor of 4 since they make the exact same mistakes. 
EIA's model shuts out almost all industrial fuel-switching; EIA 
freezes electric utility restructuring and, as but one 
technology example, EIA projects that fuel cells will provide 
no electricity, not even 100 megawatts, through 2020, even 
under the most extreme scenario, coal prices up 600 percent, 
electricity prices doubled. These kind of obviously wrong 
assumptions have hurt their credibility.
    The last thing we know for sure and the most interesting is 
that something unusual has happened to the way the United 
States uses energy. I think this is the central chart. It 
compares 4 years of data, 1992 to 1996, which are the left-hand 
bars in each case, with the 3 years of 1996 through 1999. The 
left-hand bars could be called right before the Internet 
economy took off, and the right-hand bars, the Internet 
economy. What is fascinating is that as you can see, GDP before 
the Internet had about 3.2 percent growth per year. GDP growth 
since the Internet was accelerated by 1 full percentage point. 
But what is fascinating is at the same time that the GDP has 
taken off, electricity consumption is down, which is the main 
reason we know that Mr. Mills is wrong. We are actually using 
less electricity, energy use is way down, and carbon dioxide 
emissions are also way down. This is stunning.
    Mr. McIntosh. Mr. Romm, just to interject quickly, your 
chart doesn't really say energy use is down--just the growth in 
energy use----
    Mr. Romm. Yes, you are right. What we have here is the 
historical rate relationship between GDP and energy use. As you 
can see, if you compare these 2 numbers, energy use grew just 
slightly less than GDP from 1992 to 1996, but from 1997 through 
1999, 3 years of data, what we see is bigger GDP growth, and I 
appreciate the correction, and slower energy growth and slower 
electricity growth and slower CO2 growth. Obviously, 
the key question is whether this is an anomaly or a trend. I 
believe that this is, in fact, a trend, and I think that is the 
central point.
    Let me, if I may, just briefly explain why I do think this 
is a trend, because after all, if this were to continue, it 
would be the biggest trend to hit the U.S. energy economy in 50 
years.
    First, technology. The information technology sector, as 
Dr. Hakes has said, which includes computer manufacturing and 
software, is just not very energy-intensive. A new EPA study 
that Mr. Kucinich cited projected that if this sector continues 
to generate a large fraction of our economic growth, this alone 
will reduce energy consumption in 2010 by 5 percent, compared 
to current EIA forecasts.
    Second, the Internet economy also makes the economy more 
efficient. I won't go into this at length. I wrote an 80-page 
report on this and we can get into this in the question and 
answer. Let me say what I think is the most important aspect of 
the Internet economy. As more companies put their supply chain 
on the Internet and reduce inventories, overproduction, 
unnecessary capital purchases and mistaken orders, they achieve 
greater output with less energy consumption. Federal Reserve 
Board Chairman Alan Greenspan testified in front of Congress in 
June, ``Newer technologies and foreshortened lead times have, 
thus, apparently made capital investment distinctly more 
profitable, enabling firms to substitute capital for labor and 
other inputs far more productively than they could have a 
decade or 2 ago.''
    What I am basically positing is that if you believe that 
the Internet is increasing labor productivity and reducing 
inflation, that it is also improving total factor productivity, 
including the Nation's energy productivity.
    I think there is another reason why emissions may be 
slowing, and that is that a lot of companies are voluntarily 
cutting their emissions. The Wall Street Journal article noted 
in October 1999, ``in major corners of corporate America, it is 
suddenly becoming cool to fight global warming.'' Mr. Hakes 
collects data on this subject and their latest report, which 
came out just last week, said that since 1994, the quantity of 
emissions reductions reported each year has roughly tripled. 
Total reported savings in 1998 alone exceeded 3 percent of U.S. 
emissions, no small amount. I also believe that electricity 
competition will slow emissions. We can get to that in the 
question and answer.
    Let me just sum up. In conclusion, we know Mr. Mills is 
completely wrong. Indeed, his argument is the reverse of the 
truth. The Internet almost certainly saves energy. I personally 
believe that the Internet may be one of the greatest systems 
ever devised for minimizing energy use and greenhouse gas 
emissions.
    The final point. Again, something big is happening to the 
U.S. economy. Faster economic growth and slower energy and 
greenhouse gas growth. I believe this is a major trend, a new 
energy economy, and that current EIA forecasts of high growth 
in greenhouse gas emissions for the next decade are, like many 
of their forecasts, wrong, perhaps by a factor of 2.
    Thank you very much.
    [The prepared statement of Mr. Romm follows:]
    [GRAPHIC] [TIFF OMITTED] T4876.030
    
    [GRAPHIC] [TIFF OMITTED] T4876.031
    
    [GRAPHIC] [TIFF OMITTED] T4876.032
    
    [GRAPHIC] [TIFF OMITTED] T4876.033
    
    [GRAPHIC] [TIFF OMITTED] T4876.034
    
    [GRAPHIC] [TIFF OMITTED] T4876.035
    
    [GRAPHIC] [TIFF OMITTED] T4876.036
    
    [GRAPHIC] [TIFF OMITTED] T4876.037
    
    [GRAPHIC] [TIFF OMITTED] T4876.038
    
    [GRAPHIC] [TIFF OMITTED] T4876.039
    
    [GRAPHIC] [TIFF OMITTED] T4876.040
    
    [GRAPHIC] [TIFF OMITTED] T4876.041
    
    [GRAPHIC] [TIFF OMITTED] T4876.042
    
    [GRAPHIC] [TIFF OMITTED] T4876.043
    
    [GRAPHIC] [TIFF OMITTED] T4876.044
    
    Mr. McIntosh. Thank you, Mr. Romm. I look forward to our 
questioning session. Actually, this is going to be a good 
hearing.
    Our final witness is Mr. Mills, who has already been much 
talked about. Share with us a summary of your written 
testimony, please.
    Mr. Mills. Thank you, Mr. Chairman and distinguished 
members of the subcommittee. I really am delighted that you are 
having this hearing, and I thank you for doing this. It is a 
fascinating area, and there is one thing that Mr. Romm and I 
agree about; perhaps only one, but it is that the Internet is a 
very big deal. As your opening remarks indicated, Mr. Chairman, 
it is a very big deal and it is going to be a bigger deal.
    This subject is particularly fascinating for me for two 
reasons. My testimony I will summarize very quickly, but I 
should start with a preamble. As a personal reason it is 
fascinating. My career started in integrated circuits, 
telecommunications, and fiberoptics. In fact, I have patents in 
those areas. I was intimately involved in that business in my 
early career. I will confess that I had no idea that the field 
of telecommunications, nor did anybody I worked with at that 
time in research and development, have any idea that the 
telecom industry was going to become what it is today, which is 
the driver of our economy. Nor did I anticipate that that 
industry, that profession would intersect with my current 
profession, which is the study of energy and electricity. It is 
utterly fascinating, and it is delightful to be working with my 
colleagues again in the telecom industry.
    The second point I want to make is that what you have heard 
is actually dramatized in what I learned in doing this 
analysis, is that the folks in the energy business don't 
understand telecom. They have profound misunderstandings and in 
some cases ignorance of the telecom industry and the 
technologies of that industry. Not surprising, because the 
energy industry and the environmental community and analysts 
that have grown up around it have basically been brought up on 
the culture of the oil era that started in 1973, and telecom is 
quite different.
    We have been told for 20 years by forecasters like Mr. Romm 
that efficiency measures were going to stop electric load 
growth. I published a long monograph in fact for CEI 
documenting dozens and, in fact, hundreds of forecasts for 20 
years that we were going to stop seeing electric demand grow 
because of efficiencies of all kinds. I should like to point 
out for the record that EIA's data does show that in the last 
digital decade, and I am a firm believer as a physicist that 
you can't tell a trend from one data point; 1 year doesn't help 
a lot. This past decade I have called the digital decade, and 
in the past digital decade where there has been an explosion in 
the purchase and installation of telecommunications equipment 
of all kinds, we have not only seen our economy grow, but the 
consumption of electricity has grown by 650 billion kilowatt 
hours. Just for perspective that would have required the 
additional electrical capacity equal to all of the electric 
output of Central and South America. So it is a fairly 
significant addition to the U.S. electric supply system driven, 
I would submit, primarily by our digital economy. Not 
specifically by the Internet, but by our digital economy.
    The principal objection, and we will do this in the 
question and answer, there are two principal objections to the 
analysis that I detect. One is the efficiency sufficiency 
argument, that I didn't count it, or my colleagues and I 
ignored it. We didn't. We weren't seeking to look for the 
efficiency information. We were seeking the answer to a simple 
question, and I will put it to you this way. It is based--and, 
in fact, this whole analysis started with a very simple 
premise. Every single piece of equipment in the digital 
economy, every piece of information technology equipment has 
two plugs, one for bits, one for power. All of them have power 
plugs.
    The purchase rate of hardware in the information economy 
today is running at $400 billion a year. In the last several 
years we have added $1 trillion of telecom hardware to the U.S. 
economy. We have added trillions of dollars of hardware that 
has been plugged in, net-new hardware in the last digital 
decade. In looking at this, it sort of begged the question, 
wouldn't it be reasonable that all of this hardware at some 
point would begin to consume a reasonable or significant amount 
of electricity?
    What we found in trying to seek an answer to this question 
was that the data collection mechanisms aren't up to the task. 
I am a personal fan of EIA; I think their data is the finest, 
in fact, in the world. I study data from all kinds of 
countries. EIA does a great job, but the problem is they don't 
have a collection mechanism to find what we need to find.
    Let me just give you a real quick example of the 
profoundness of the misunderstanding of the Internet and one 
example of the very disappointing analysis that the Lawrence 
Berkeley folks undertook that Mr. Romm has summarized. It is 
profoundly disappointing and misleading as well as being 
factually wrong, which is a problem for scientists, but it does 
happen. We all make mistakes.
    The first problem is that folks keep interpreting the 
Internet as the PC on your desktop. The PC on your desktop 
constitutes a tiny fraction of the portal to the Internet. The 
Internet comprises the hundreds of millions of pieces of 
equipment that are in the network that create, shape, move, 
route and feed bits into the entire network. In fact, when one 
looks at the analysis, you not only find that the PC itself 
constitutes about 20 percent of a total energy appetite of the 
Internet, the critical fact is that you can take the PC out of 
the picture totally, and you still have electric use growing 
for the Internet.
    The folks at Lawrence Berkeley did an analysis that 
rebutted themselves. Their own rebuttal of my study contains 
statements of fact which contradict their own analyses from 5 
years ago. They also contradict analyses from the National 
Academy of Sciences, from the Environmental Protection Agency 
and other organizations. It is an embarrassingly shoddy piece 
of analysis, frankly, and I was surprised that they would not 
do better research. I will be happy to give you more examples 
in the question and answer.
    Let me just finish with an observation about efficiencies. 
Quite obviously, the Internet is driving economic efficiencies, 
and it is certainly driving some energy efficiencies. Where it 
is driving efficiency largely is in the transportation sector. 
I would submit to you that that means that electricity is 
substituting for oil. This has been going on in the American 
economy for about 30 years; in fact, for a century. This is a 
good trend, and, in fact, it is a trend that I analyzed a 
decade ago and published widely on. I think it is continuing. 
However, I do believe that what we are seeing is the thin edge 
of a wedge that, in fact, I don't believe you can see in EIA 
data yet.
    Let me give you one specific example of why you can't. The 
data you see right here for the commercial sector shows you 
that the largest category after lighting for a commercial 
building energy use is ``other.'' ``Other,'' if I understand 
the data correctly, and Mr. Hakes will correct me if I'm 
wrong--but ``other,'' if I understand it, includes telecom 
equipment. I will remind you again, telecom equipment is the 
Internet.
    Now, if we have a disagreement about how much electricity 
the Internet uses, I submit to the Lawrence Berkeley scientists 
and I submit to my other colleagues, let's start then with the 
telecom industry, and then let's ask telecom experts, what 
share of the telecom industry is being driven by the Internet. 
I think you will find that they will say that it is the driving 
force. But we can at least agree to start with a broader 
definition to begin to understand the broader electrical 
appetite of a digital age. It is enormous, it is growing.
    When I started this analysis, it was not to shoot holes 
into Kyoto or EIA or forecasters, but to answer a very 
important question that is still not adequately answered by 
EIA, Lawrence Berkeley, by Mr. Romm, or by anybody else. How 
much electricity is the digital economy using specifically, 
and, as a subset, how much electricity is the Internet using?
    I will tell you since I published my study, I have engaged 
in substantial additional research with my colleagues in the 
telecommunications industry. All of the data strengthens my 
original conclusions, it does not diminish it. Thank you, Mr. 
Chairman.
    [The prepared statement of Mr. Mills follows:]
    [GRAPHIC] [TIFF OMITTED] T4876.045
    
    [GRAPHIC] [TIFF OMITTED] T4876.046
    
    [GRAPHIC] [TIFF OMITTED] T4876.047
    
    [GRAPHIC] [TIFF OMITTED] T4876.048
    
    [GRAPHIC] [TIFF OMITTED] T4876.049
    
    [GRAPHIC] [TIFF OMITTED] T4876.050
    
    [GRAPHIC] [TIFF OMITTED] T4876.051
    
    [GRAPHIC] [TIFF OMITTED] T4876.052
    
    [GRAPHIC] [TIFF OMITTED] T4876.053
    
    [GRAPHIC] [TIFF OMITTED] T4876.054
    
    [GRAPHIC] [TIFF OMITTED] T4876.055
    
    [GRAPHIC] [TIFF OMITTED] T4876.056
    
    [GRAPHIC] [TIFF OMITTED] T4876.057
    
    Mr. McIntosh. I thank all of you.
    Mr. Kucinich has to go to the floor, but let me ask 
unanimous consent that the record be held open, and if Mr. 
Kucinich has some questions that he would like to pose to any 
of you, I ask that you answer them in writing.
    Mr. Kucinich. I just want to thank the Chair for doing 
that. Unfortunately, I have to go to the floor right now. A 
bill that I am involved in is in debate.
    So I want to tell you, I am very impressed with this panel. 
There are differences of opinion, and I wish I could stay. But 
if we have another Member come, they might want to ask the 
questions that I have. But if they don't show up, I would like 
them inserted in the record.
    Mr. McIntosh. In fact, let me ask unanimous consent for 
questions from any of the members of the subcommittee.
    Mr. Kucinich. Mr. Chairman, thank you.
    Mr. McIntosh. Thank you, Dennis.
    Mr. Mills, I called this hearing because some of your data 
made its way into one of our hearings on Kyoto. Today, you made 
some pretty conclusory statements that you are right and they 
are wrong. It appears from this chart that we were given that 
there is a factual question on the magnitude of Internet-
related electricity demand, and that you and your critics are 
using the same definitions, or is there an explanation for why 
the Lawrence Berkeley people using your definitions came up 
with lower numbers?
    Mr. Mills. Well, let me, Mr. Chairman, give you an example. 
In their analysis to adjust my data down, they looked at web 
servers. This is one representative example. There are many, 
and it would take far too long, but this is typical of what 
they did wrong. They said that I counted the number of servers 
incorrectly.
    One of the things you said in your opening remarks is that 
we are having a hard time measuring the Internet right now. In 
fact, there is a great debate figuring out what the metrics are 
for it. I wasn't looking at how much e-commerce was going on, 
but how much hardware exists. A server is the computer box that 
contains the website, and it can contain one website, or it can 
contain 100 websites.
    It turns out that there are organizations that do something 
called ping the Internet. They send signals up, and they ask 
computers, are you there, are you a server. Every server has an 
identity, and it says, yes, hello, I am here.
    In my original report, I estimated the number of servers by 
looking at the number of websites and doing an extrapolation. I 
came up with the number 4 million servers in the United States. 
By the way, that number is growing geometrically, but that 
number was 4 million last year. Lawrence Berkeley said, no, no, 
Mills' assumptions were wrong, it should be reduced by 80-some 
percent to 1 million, because there are more websites per box.
    What I have done subsequently and they obviously decided 
not to do is to find out what the actual hard data is for the 
number of servers; not websites, but the number of physical 
boxes. The number is 4 million. So their adjustment now of my 
web servers by a factor of 80 percent is simply completely 
wrong.
    Now, they have done this over and over again in other 
areas. It is that kind of methodological failure on their part 
that made me wonder if they were seeking their own conclusion 
rather than seeking to get the facts, because they clearly did 
not look into this data properly.
    Mr. McIntosh. OK. Would you for the committee go through an 
analysis of the different ways in which they have done that? 
Frankly, we are going to ask them to justify their work as 
well.
    Let me ask this question to Mr. Romm, because I actually 
think that it is possible that Mr. Mills or Mr. Romm are 
actually looking at the same thing from 2 different 
perspectives.
    Part of what your chart there demonstrates is this. As we 
have seen growth in the economy, acknowledged by all of us to 
be driven by the technology, computers and telecommunications, 
we are seeing a reduction in the growth of energy and 
electricity use in particular, but energy overall, and 
therefore the growth in emissions from energy production. That 
means, essentially, that there are jobs being added into the 
economy driven by that technology sector. The flip side of 
that, and we see this in my district, there are jobs leaving 
the economy that are the old manufacturing jobs that were more 
electricity-intensive. So you have a shift going on from the 
older economy to the new.
    Now, that doesn't necessarily mean that you don't also have 
an increase of use in electricity, because if you assume that 
that decrease in manufacturing jobs were to occur regardless of 
the technology revolution, what could be happening is that we 
might, had we not seen the technology revolution, have seen the 
growth in electricity be even lower. Is that not right?
    Mr. Romm. Well, you ask a very complicated question, 
because one has to try to figure out exactly what is going on 
in every aspect of the economy. I think that part of what you 
are saying is certainly correct, although it is worth saying 
that U.S. manufacturing output has not declined. As I am sure 
you know, Bethlehem Steel and other companies in Indiana have 
gotten more productive, far more productive in the last few 
years, so they have been able to generate more output with the 
same or fewer workers.
    I think the main point is that a very large fraction of our 
growth is coming from the computer industry, the software 
industry, the telecom industry, industries that do not consume 
as much energy as the steel industry would. And so----
    Mr. McIntosh. Let me ask you, what does that mean, 
``consume as much''?
    Mr. Romm. In other words, let's say----
    Mr. McIntosh. Is that per dollar of GDP?
    Mr. Romm. Per dollar of GDP. In other words, what we are 
seeing here is, in some sense, the measure that is used of 
overall U.S. economic efficiency is to divide the total energy 
number each year, the growth in energy, by the growth in GDP, 
to come up with a metric called energy intensity. You can 
clearly see that the Nation became a lot more--a lot less 
energy-intensive in the last 3 years because we had big growth 
in GDP, small growth in energy. What this suggests is that in 
part, we might get $1 trillion in economic growth from the 
information technology sector, but if we had gotten $1 trillion 
in economic growth from the steel industry, that would have 
used, you know, easily that much energy.
    Mr. McIntosh. So you are seeing a shift from manufacturing 
and therefore manufacturing jobs into technology and technology 
jobs, which use less electricity?
    Mr. Romm. By and large, yes. I think--there has been a 
study done by the EPA and Argonne which roughly says that that 
structural change which you have just described is about a 
third to a half of the explanation for this trend. There is 
another trend--there are two ways the economy can become more 
efficient at using energy or less energy-intensive. One is the 
structural change that you have described. The other is to put 
in energy-efficient light bulbs, to replace trips to the mall 
with working at your desktop. That is becoming more efficient.
    Mr. McIntosh. Exactly. Which leads me to my next question, 
and then I will come back for some of the other witnesses.
    If--and as I understand by your testimony, that structural 
shift can account for a savings of about 300 million metric 
tons by 2010, and that that is approximately two-thirds of what 
the obligation under the Kyoto protocol would be for the United 
States, my question is, wouldn't we be better off allowing that 
invisible hand of the marketplace to work, rather than bringing 
in a regulatory structure when you are going to be able to see 
those reductions and savings because of the structural shifts 
in the economy?
    Mr. Romm. Sure. Well, I know that we probably take 
different views of some aspects of this issue. Let me try to 
answer the question this way: I believe that greenhouse gas 
emissions are going to grow more slowly because of the Internet 
and other factors than EIA does, and that will certainly make 
it easier to meet the Kyoto targets. Those of us who think that 
it is important for the country and the world to reduce 
greenhouse gas emissions still feel strongly that some market-
based signals help show businesses that we don't want to reduce 
energy consumption per se. What we want to do is pull 
CO2, greenhouse gases, out of the U.S. economy the 
way we pulled sulfur out of the electricity grid and lead out 
of gasoline. It is my belief that a relatively low market----
    Mr. McIntosh. Unless you move to hydroelectricity or 
nuclear fuels, you can't do that.
    Mr. Romm. No, that is not true. The utility system right 
now is only 28 percent efficient in converting fossil fuels to 
electricity. It is very inefficient, in part because we have 
had a regulated monopoly.
    The average coal plant is only 30 percent efficient. There 
are gas-fired, combined-cycle units which are 55 percent 
efficient. There is cogeneration, the simultaneous generation 
of electricity and steam, which is 90 percent efficient. We can 
have, you know, what is called fuel-switching or more efficient 
conversion. We can have some renewables. I don't think----
    Mr. McIntosh. All of that depends on the price signals.
    Mr. Romm. Well, what is very clear is that there are a lot 
of low-cost measures, and it appears, from Mr. Hakes's report 
on voluntary actions by companies, that a lot of companies are 
taking voluntary actions to reduce their greenhouse gas 
emissions. Most of those, I believe, are utilities.
    I think it is quite interesting that in 1998, U.S. global 
greenhouse gas emissions dropped, and total energy use in the 
world actually dropped for the first time since 1982.
    I think you will find that there are a lot of steps that 
companies and governments can take that do not require a very 
big price signal, and that are very--obviously, we have shown 
that you can have low CO2 growth and high GDP growth 
at the same time, so it is certainly difficult to argue that 
having lower CO2 growth harms GDP.
    Mr. McIntosh. Let me let Mr. Ryan ask questions, and then, 
Mr. Hakes, I will give you your opportunity which you asked for 
in your testimony.
    Do you have a set of questions?
    Mr. Ryan. Sure.
    Actually, Mr. Hakes, you haven't had a chance to talk about 
any of this. I would like to get you involved right now, if I 
may.
    I was intrigued with what Dr. Romm said about EIA's 
overestimation of Kyoto's costs. EIA estimates that the Kyoto 
protocol would increase average electricity prices from 20 
percent to 86 percent in 2010.
    Dr. Romm, I think you said in your testimony that that is 
wholly irrelevant to the real world.
    Could you comment on that, your assessment of Dr. Romm's 
comments?
    Mr. Hakes. Well, this is a very detailed study that we did 
for the House Science Committee, and my view is that it has 
held up well over time. I think the analysis is quite good; I 
think there is material in there that somewhat supports the 
view of Kyoto critics, and I think there is some conclusions 
that do not support the views of Kyoto critics.
    I think that we have suggested in there that carbon 
emissions would grow a little bit over 1 percent a year; you 
know, a little bit larger than the population growth.
    If you look at Mr. Romm's chart for CO2, you can 
see in the last couple of years it grew at a slower pace. Why 
did it grow at a slower pace? The main reason was the mild 
weather; it wasn't because of some new efficiency gain. People 
have to use less heating oil, less heating from their space 
heaters and less natural gas. The other reason was there was an 
economic decline in Asia that reduced the demand for a lot of 
products that are energy-intensive.
    If you look at the preliminary data for 1999, which is not 
official yet, it looks like carbon emissions will grow about 1 
percent a year. In 2000, we are having a very cold winter, 
maybe they will grow more.
    On the issue of Mr. Romm's attacks on our forecasts, he has 
launched a lot of grenades here, but let me mention one that he 
suggests in his written testimony--that EIA doesn't have 
foresight on the Kyoto treaty. Critics like Mr. Romm have said 
we have overestimated costs because we don't allow people to 
adjust before 2005. It is just flat wrong, and it has been 
corrected in the public record.
    We allow perfect foresight by the electric utilities. In 
other words, in our model we assume that it started at the date 
of our study, which is now more than a year old, and is fully 
operational now, which is I think very overly optimistic to his 
side. We have also assumed perfect foresight in the refining 
industries, so the voluntary programs by BP and others that 
tend to focus on refining are already fully incorporated and 
obviously a lot of companies are not doing this. So if there 
are other industrial sectors that don't have foresight, I think 
it all adds up.
    We treat, I think, foresight in our study very, very fairly 
and very balanced, and maybe even perhaps tilted a little bit 
to Mr. Romm's point of view, and yet he continues to repeat 
over and over again, even when it is shown by us to be very 
clear in the document itself, if you actually read it, that we 
allow foresight rather amply throughout our system.
    His attacks on our oil price forecasts; our forecasts are 
very highly respected among energy experts. There was something 
that happened last March. OPEC cut world production by about 3 
million barrels. We do not have a pipeline by which OPEC tells 
us whether they are going to cut production or not, but that 
was certainly the main reason that everybody's signals on oil 
prices were reversed. And once they made that policy decision, 
we said that prices would be going up into the basic range that 
they are now.
    Of the big errors that he mentioned, and I certainly admit 
to, one of the things that EIA does as a Federal agency, as a 
public agency, is to bend over backward to point out ourselves 
any previous mistakes we have made. But the one he was talking 
about in the Science Committee was talking about the early 
1980's, the first one or two forecasts we did at the time of 
deregulation.
    I think our forecasts in the 1990's have stood up extremely 
well. We had some arguments. The administration had a study in 
1993 saying that voluntary programs would cause carbon 
emissions in the year 2000 to be what they were in 1990. We 
said, ``No, we don't think that will work, they are going to 
grow in the 1990's.'' I think we were basically correct, and I 
think we have been more correct than some of the optimists like 
Mr. Romm.
    One other point on electricity that is very interesting. In 
our forecasts, consistently in the 1980's, we have 
underestimated the growth of electricity because we had too 
much built in for appliance efficiency. So EIA, who Mr. Romm 
you seem to portray as always underestimating these efficiency 
gains, actually consistently overestimated them in the 
electricity sector in the 1990's.
    We try to do a balanced job. I think we don't always get it 
exactly right. I don't think there is a systematic bias, 
particularly since we have put more attention on technology 
issues in recent forecasts.
    Mr. Romm. Could I just--very little--many things that Mr. 
Hakes said are just factually incorrect. In their latest annual 
energy review, which they just released in December, they 
predict for the next 7 years that CO2 emissions will 
grow 1.8 percent for 7 years, not 1 percent, as Mr. Hakes said.
    I think the key point is there is no question that the EIA 
occasionally stumbles onto the truth. If they would be more 
humble about the fact that they are often wrong and not release 
reports that are called Impacts of the Kyoto protocol on U.S. 
Energy Markets and Economic Activities, but even Possible 
Impacts, or A Scenario of Impacts, but they continue to believe 
over and over again that they are correct, and over and over 
again they are proved not to be correct. I will leave the rest 
of the corrections for the record.
    Mr. McIntosh. Let me actually have a followup, because 
while Mr. Romm is critiquing them from one direction, Mr. 
Mills, in your testimony you seem to think that they had 
underestimated the impacts of the digital economy on 
electricity demand. Can you explain your viewpoint on that?
    Mr. Mills. Well, let me explain it from two perspectives. 
One is forecasting, and one is from the history. Actually, I 
think that--well, for the record, I actually believe that EIA 
is--not that I want to get into the middle of a battle between 
the lovers and haters of EIA. I actually think that their track 
record is remarkable. Mr. Hakes is absolutely correct. The 
track record in the 1980's was pretty abysmal. After that it 
has gotten very good. They have gotten very sophisticated in 
understanding the markets and have overestimated the impact, as 
Mr. Hakes said, of efficiency measures in the economy, partly 
because it is an engineer's dilemma. When more efficient 
refrigeration exist, people tend to use more of it. The markets 
are a little more complicated than that, as Mr. Hakes' analysts 
know.
    I think what we have seen in the last digital decade is, in 
fact, the information economy effect. In the last decade there 
has been an enormous increase in the efficiency of lighting, in 
refrigerators, in motors. We have spent in this country tens of 
billions of dollars on these so-called demand-side management 
programs for the last 20 years, and the fruit was born in the 
1990's. If you go by sector, you will find, yes, they are 
incredibly more energy-efficient in lighting of buildings and 
air conditioning of buildings, and yet demand went up. The 
growth rate wasn't as high as the 1970's and 1980's, but it 
still went up. It went up in aggregate by an enormous amount.
    The most fascinating thing to me is that it went up so much 
in the category that used to be a grab bag, and this is--not 
that I am pleading for funding for EIA, but it is a funding 
problem.
    In the old days, you threw things that you couldn't count 
into the ``other'' category, and everything else took up 90 
percent of electricity demands. So the 10 percent in ``other,'' 
who cares? If you look at commercial building electric use in 
the EIA data now, ``other'' is a third of all electricity used 
in commercial buildings. ``Other'' is growing faster than 
anything else, and ``other'' is exceeded only by lighting. It 
is 300 billion kilowatt hours of ``other'' stuff, which was 
where I contend a lot of the Internet hardware is showing up, 
because it is not tracked. In fact, many of the organizations 
that track data like refrigerators and motors don't even track 
servers.
    Cisco is everybody's pick. They hope they bought into Cisco 
5 years ago. Cisco is selling a million routers a year. Cisco's 
routers don't show up in information technology industry data 
books because they didn't exist when they created the data 
books. They don't track them. Neither does EIA track them. 
There is a methodological problem which is inherent in the new 
economy.
    Mr. McIntosh. And so if you are correct and the energy use 
is going to grow dramatically because of the electricity 
demands of the equipment that drives the Internet, what 
explains the decrease in electricity growth with the increase 
in economic growth? Is it that we are losing manufacturing jobs 
faster than we are replacing them with the demand for the 
Internet equipment?
    Mr. Mills. We found one other area where Mr. Romm and I 
agree. There is more manufacturing in America. There are fewer 
jobs, but there is actually more manufacturing output.
    Just as an aside, the manufacturing sector in the last 
decade has decreased its consumption of combustible fuels and 
increased its use of electricity. It has become more 
electrified. Steel mills are a great example. Five percent of 
steel mills in 1970 were electric; now it is 40 percent. So 
they are very electric intensive, very efficient. They are 
computerized, and they have fewer jobs. The jobs have gone to 
the information and technology and service sectors.
    Let me explain more clearly my forecast is that I think 
that EIA is close to right and, in fact, Mr. Hakes's testimony 
says he thinks the Internet could take it either way, lower or 
higher. My guess is that it will take it higher. But a 1 or 2 
percent range in load growth in the electric sector as big as 
ours is a very big number. We are not a small country. When we 
move electric demand a few tenths of a percent it makes a huge 
difference.
    The explanation for the efficiency numbers is something 
like that, and I don't mean to be facetious, but I call it the 
Bill Gates effect. We have a tendency, energy analysts, to 
measure energy efficiency the way Mr. Romm was showing you: 
BTUs per dollar of GDP. I would submit to you that this is 
interesting but probably totally irrelevant and indeed 
enormously misleading. If we have an economy like our economy 
where the Internet is consuming more energy, not enormous 
amounts, but more energy, not less, but it is bootstrapping 
economic growth even faster, which is the effect we see from an 
Amazon.com, if you like, what you get from that is modest 
continued energy growth from a huge base. Also huge GDP growth. 
So the energy efficiency improves.
    Let me put it to you this way. Who would be the most energy 
efficient person in the world today by this measure? The answer 
is Bill Gates.
    Mr. McIntosh. Let me take it one step further. If you have 
a public policy, such as the Kyoto protocol, that says we are 
going to increase the cost of electricity for the social good 
of eliminating or reducing carbon dioxide, what is the impact 
on the new emerging computer and communications industry 
compared to the other sectors of the economy?
    Mr. Mills. It is a serious impact, and it would have been 
modest, if not irrelevant, 20 years ago. But once you take--
once you state the premise that the information economy is at 
the core of our economy as the Commerce Department has 
estimated a third of GDP, real growth, once you know that it is 
no longer trivial, web-serving firms are huge, electric-
consuming beasts, the price, supply and reliability of 
electricity become the central factor in our economy, not 
secondary. They become, I would argue, more important than oil.
    It is now--oil is still extremely important, but the price, 
supply and reliability of electricity are now more important 
than oil to our economic future. To me that suggests that 
things like Kyoto and the Kyoto protocol that create enormous 
risk in the electric supply sector are dangerous to move 
quickly on for that simple reason. It is so important now.
    Mr. McIntosh. I will share with you just one example from 
my home State in Indiana where EPA has filed a ridiculous 
lawsuit against the electric utilities saying that if you try 
to maintain your plants you are going to be in violation of the 
Clean Air Act. What is driving that policy is a desire to stop 
the utilities from using coal. If you remove coal as an 
electric power source, that does a huge amount of damage to the 
economy in the Midwest, driven by these other social goals.
    Mr. Mills. One of the things that we are seeing here is 
that EIA, regardless of whether we think their forecast is 
accurate or not, I think it is pretty good, sees the demand for 
electricity rising. So we have, as Mr. Hakes said, about half 
of our electricity--a little over half comes from coal in 
America. In the future, we will need more electricity and not 
less. If we implement policies that reduce the base while we 
still have an increased demand, we create enormous risks.
    Mr. McIntosh. I have one question for each of you, 
actually. Are there any forecasts out there if you double the 
cost of electricity, what that does to impede the growth in the 
technology sector?
    Mr. Mills. Well, I did an analysis like that 2 years ago on 
the commodities basis, but I will be happy to defer to Mr. 
Hakes first if you have looked at this.
    Mr. Hakes. If you are talking about a Kyoto scenario, it 
depends on how you would implement the Kyoto scenario. One of 
the scenarios that economists have used is that you auction off 
the carbon credits which are then recycled back into the 
economy, perhaps through a Social Security tax cut. If that was 
done, your energy-intensive industry, your coal industry would 
be net losers, but some of your low intensity users might 
actually be winners. But there is a whole range of policies 
there that have not been addressed in any systematic way, so 
without those kinds of considerations, it is hard to assess the 
total impact on an industry.
    Mr. McIntosh. You mentioned, Mr. Hakes, recycling through a 
carbon tax. At least as far as I can tell, there are no 
proponents for that, either in the administration or here in 
Congress. How about some of the other scenarios where it is 
just done through a command and control pricing increase 
policy?
    Mr. Hakes. Well, everything that I have seen has been 
proposed to cap carbon in some way. You can do that by either 
grandfathering people in or you can auction credits. There are 
different ways to do that. I don't think that those types of 
issues are being discussed very much, because they tend to be 
somewhat sensitive. There are large economic impacts. But you 
couldn't implement the Kyoto treaty without making some 
judgments in those areas. You probably would have to use some 
revenues to compensate the coal industry in some way, since 
they would be such obvious losers in a transition.
    Mr. McIntosh. But if you grandfathered people in and you 
see the shift from oil to electric generation, which is 
generally done through coal, then you could cause some 
distortions in the growth in the economy.
    Mr. Hakes. It would probably help some areas and hurt 
others.
    Mr. McIntosh. I guess my question is, on which side of that 
equation would technology and communications fall?
    Mr. Hakes. Depending on how it was recycled, they might be 
net beneficiaries.
    Mr. McIntosh. But that is assuming there is a tax. In the 
absence of a tax.
    Mr. Hakes. Without it, they would be less hurt than, say, 
the aluminum industry or the refining industry or the auto 
manufacturing industry, who are very energy intensive 
industries, or the airlines industry, who have high energy 
costs and therefore it would affect their bottom line a lot 
more. I think for the computer companies, this is not as big an 
issue for them, although, clearly, they do have energy costs 
and have a great interest in the reliability of the electric 
system.
    Mr. Romm. Let me make a few points.
    First of all, certainly the worst-case scenario, 
electricity prices aren't going to double under Kyoto.
    B, even in the most pessimistic analysis, which is Mr. 
Hakes, he says that we could comply with Kyoto with ``no 
appreciable change'' in the long-term GDP growth rate. So the 
fact is that it would not harm the economy.
    The third is that Mr. Mills--Mr. Mills is testifying that, 
on the basis of his study, he can tell you that Kyoto would be 
bad for the U.S. economy, and yet he has only looked at the 
electricity used by the Internet. He has not looked at the 
electricity saved by the use of the Internet in the economy. 
Therefore, his study is wholly inadequate to draw that 
conclusion.
    I am not going to testify here today that I know for a 
certainty that everything that I wrote is true. That is why we 
called our work a scenario. But it is very clear that if you 
look at the data, electricity, the rate of growth of 
electricity consumption has dropped, energy growth has dropped, 
CO2 growth has dropped, and GDP growth has grown. So 
I don't see how anyone can argue with the premise that we can 
have lower CO2 growth and higher GDP growth.
    Just to correct one other misstatement by Mr. Hakes, we 
have done a weather correction. It is very clear that the 
United States is getting warmer. NOAA has--you know, the 
National Oceanic and Atmospheric Administration has 
acknowledged that, EIA has acknowledged that. When you do the 
weather correction, we have had a slightly warm 1997, very warm 
1998, not a particularly warm 1999 compared to 1998. The 
weather correction brings annual energy growth up to slightly 
under 1 percent and the same for CO2. It explains 
about one-quarter of this remarkable graph here.
    Mr. McIntosh. And I think I am going to have to leave; but, 
Mr. Hakes, you have a comment you want to make.
    Mr. Hakes. Mr. Romm talks about this remarkable graph. I 
would like to introduce into the record the intensity gains and 
losses in the U.S. economy going back to 1960. You look at 
1981, the intensity gain was 4.6. If you look at 1983, the gain 
was 4 percent, greater than we have seen in the last 2 years. 
So in 1984, you have Amory Lovins saying, we see electricity 
demand ratcheting downward over the medium and long term, much 
as Mr. Romm is saying today, and he actually had more 
convincing data at that time than Mr. Romm has today.
    So what happened in the next few years? The pattern 
reversed itself. Because things go in cycles. Weather goes in 
cycles. The economy in Asia goes in cycles. If we were to use 
Mr. Romm's approach, if you take the 4 years before 1997, the 
intensity gains in 3 of those 4 years rounded out to zero. We 
didn't change our optimism about continuing intensity gains 
because we thought those statistics were aberrations.
    Mr. Romm didn't come to us at that time and say lower your 
intensity gains because we got these very low intensity 
numbers. So he picks two numbers in 2 recent years that raise 
some interesting questions but are clearly highly relevant to 
weather, where you don't see energy use going down, you don't 
see carbon going down, and tries to argue that it is a brand-
new trend. I think the data are much less convincing than the 
data from 1981 and 1983.
    Mr. Ryan. Mr. Mills, could you comment on that as well, 
please?
    Mr. Mills. Well, the energy intensity trends of the 
economy, it again masks two things. One is, are we getting more 
energy efficient by this measure? Which again I contend is a 
misleading measure. What you see in these graphs is, regardless 
of what the growth rate is, the operative word is growth. The 
economy is growing. Consumption of energy is growing. The 
emissions of carbon dioxide are growing. None of it is 
declining.
    So my conclusion, which Mr. Romm misstated--misrepresented, 
was not that the Kyoto protocol harms us. It is that the kind 
of policies that would tamper with the electric supply system 
of this country, which is the single most important part of our 
economic infrastructure at the root, are dangerous given the 
rising importance of electricity, not its declining importance.
    Let me add one other point just as an overlay on this to 
get back to an earlier remark I said about the challenge we 
have here is understanding this new economy. Mr. Romm, I don't 
believe, understands it. Mr. Hakes knows that it is complicated 
and has said--in fact, I have seen in their data their analysts 
are struggling to figure out how to fit it in, and we don't 
have the data collection mechanisms.
    But one particular conclusion in Mr. Romm's study that is 
breathtaking in its erroneousness I have to just put on the 
record. His conclusion about how much energy the Internet uses 
states that it is not a significant energy user because it uses 
the existing telecommunications infrastructure.
    I have to say, this is laughable if you talk to people in 
the telecom industry and go to any hearings in the Congress on 
telecom. The telecom industry has been turned upside down by 
the Internet. We have had trillions of dollars of new 
infrastructure built because of the Internet. There was no 
existing infrastructure. It was created for the digital economy 
and for the Internet. This would be like saying in 1950 that 
the interstate highway system existed and was an existing 
infrastructure, did not require energy or materials to build 
out over the next 40 years. It didn't exist. It had to be 
built. We are in the build-out phase. This is 1950 for the 
digital highway right now.
    Mr. Ryan. Mr. Romm, I am very intrigued at his confidence 
in this chart and the confidence in these 2-year sets of data--
--
    Mr. Romm. Three.
    Mr. Ryan [continuing]. 3-year sets of data indicate a 
trend. Let me ask just left from right, Mr. Mills and then Mr. 
Hakes and Mr. Romm at the end, to comment on that. Does 3 years 
of data indicate a trend? Have we severed the tie?
    Mr. Mills. I will make this answer short. Three years does 
not set a trend. A decade, 5 to 10 years is useful. I do not 
believe we have severed any ties of any kind. I think there are 
some important structural changes, but, no, 3 years is 
interesting but not a trend.
    Mr. Ryan. Mr. Hakes.
    Mr. Hakes. He has averaged the 3 years, but you can hide a 
lot of things with averaging. The intensity gain in 1996 was 
0.2 percent. That happened to be a cold period, so that is not 
surprising. So the real intensity gains that he is talking 
about are only in 1997 and 1998, 3.4 and 3.9, and because of 
the unusual weather and economic conditions, as I have said 
before. There are a lot of people around town, economists and 
others, in the administration and others, and I think most of 
them agree with us that weather was a big factor. This is 
certainly worth watching but does not indicate a significant 
new trend.
    Mr. Romm. First, to just correct Mr. Mills, the paper makes 
very clear one of the things the Internet does is take 
advantage of the fact that everybody had computers or a lot of 
people had computers before the Internet and that consumed a 
great deal of electricity. For Mr. Mills, if the economy was 
consuming a lot of electricity, if someone had a PC and it 
consumed electricity, they spent any time on the Internet, that 
is all Internet electricity, even though the economy was using 
it for other purposes beforehand.
    Let me make a few things clear. In Mr. Hakes's testimony, 
he believes I am talking about 1997 and 1998. We are, in fact, 
averaging in the first 10 months of the 1990 data--the 1999 
data. What is interesting about the 1999 data is that it has 
nothing to do with weather, because 1999 was not as mild as 
1998. We have talked to EIA's analysts. We have done the 
weather correction. We have talked to EPA. The weather 
correction is only 25 percent of this effect.
    Mr. Ryan. Let me interject at this point. Mr. Hakes, what 
about the Asian meltdown? How much of that is significant to 
this data?
    Mr. Hakes. We think that it is part of the drop in natural 
gas usage in 1998. I think the chairman was getting at this 
point earlier. You can talk about intensity gains, but if you 
are relating this to Kyoto, you are not getting Kyoto gains if 
the economy grows faster and energy consumption stays the same. 
So we are sort of talking about apples and oranges here.
    Intensity is a little bit irrelevant--it is a factor, but 
the real issue is how much energy is being consumed and how 
much carbon is being emitted. So the fact that we have had this 
incredible economic growth in the last few years and its 
improved intensity, doesn't necessarily mean energy use will go 
down.
    One other point I think the committee might be interested 
in is that the Commerce Department has changed the way we 
calculate GDP and that is it gives us statistics that are 
higher not because the economy is growing faster but because 
the statistical calculations are different, which means that 
all of these intensity numbers are going to be adjusted upward 
in the next year or two, so that you will see more improvements 
of intensity because there is a better measurement system now 
for GDP. You can sit around and celebrate when that statistical 
adjustment is made and say that this makes Kyoto easy, but, the 
bottom line, it doesn't change energy consumption; it just 
changes the ratio between energy and the economy.
    Mr. Romm. What I would like to end up by saying is as 
follows: We put out our scenario on the table because it looks 
like something very interesting is happening in the economy. 
Mr. Hakes is supposed to be representing an objective, 
independent, energy analytical agency. They ought to be very 
interested in what is going on in the economy. And yet he comes 
here and he tries to tell you nothing is going on, I can 
explain it all, even though--and I would like to introduce into 
the record a chart that we have done that shows that, in fact, 
energy intensity in every sector has improved, including the 
transportation sector. In fact, the transportation sector, 
which is completely--almost completely independent of the 
weather, has seen the biggest 2-year drop in the intensity of 
travel per GDP in 30 years of data.
    The fact of the matter is that something big is going on in 
the economy, and I just challenge EIA to be curious about it.
    To give you an idea of what they did in their last annual 
energy outlook 1 year ago, they projected 8 years of 1.6 
percent CO2 growth. We had no growth in 
CO2, 0.3 percent, in 1998. And instead of saying, 
oh, something big may be happening in the economy, we should 
lower our CO2 forecasts, they actually jacked them 
up to 1.8 percent, which suggests to me that they are not 
sufficiently curious as to whether something big is going on in 
the economy that might affect forecasts that people make very 
important decisions on.
    Mr. Ryan. OK. Before I go with a final question, Mr. Mills, 
would you like to respond to that?
    Mr. Mills. I just want to make this observation about the 
forecasting challenges and energy intensity. One of the things 
that is very clear and that is important to the subject at hand 
is that we haven't stopped building out the Internet. 
Regardless of what we decide that the consumption of 
electricity is that is ascribed to the Internet, it is still 
growing exponentially.
    We are at the buildup phase. We are building up everywhere, 
not just .coms and servers, but homes, multiple PCs in home, 
cable modems, wireless telephony, wireless data access, 
wireless palms, all of these things use electricity and they 
are growing at exponential rates. They will increase the 
consumption of electricity reasonably, and it will bootstrap 
the GDP even faster.
    So when we combine that effect with the readjustment in how 
we measure the GDP, we are going to get an even bigger 
appearance of improved energy efficiency by this fallacious 
measure which is BTUs per dollar of GDP, but we will still use 
more energy, which is a critical thing to keep in mind.
    Mr. Ryan. How important is the comparisons of growth rates 
versus nominal mall growth?
    Mr. Mills. Well, what matters from the reality of physics 
and materials is growth, not the rate. And from a fundamental 
perspective, we want to know how much stuff are we using, where 
is oil, gas, coal, steel, aluminum, are we going to use more 
stuff next year. The rate can change, depending on how you 
measure the rate, per person, per house, per dollar. You can 
get very odd results with statistics. There is that old Mark 
Twain saying: there are lies, damn lies, and then there are 
statistics. Rates are tough things. But the absolute growth 
tells you a lot about the materials.
    And we have used more stuff, and we will use more stuff and 
the biggest increase in stuff will be kilowatt hours. That is 
where the risk is in Kyoto. It is not this rate discussion and 
forecast into the future. The one forecast I can make 
confidently is that we will use more. We don't know exactly how 
much more. That is the challenge.
    Mr. Ryan. I was a fan of the Presidential debates which we 
have seen over the last few weeks, and I think we had some 
interesting dynamics in those Presidential debates and that was 
where the candidates asked each other a question. I think this 
has been a very informative hearing, it has been interesting. 
We have some unique personalities assembled here today. I would 
like to interject that little trend here, so to speak.
    Let me do this, how about in no particular order or reason, 
Mr. Mills, why don't you ask Mr. Romm a question; Mr. Romm why 
don't you ask Mr. Hakes a question; and Mr. Hakes, why don't 
you ask Mr. Romm a question. We have been asking questions, but 
I would be intrigued to watch a dialog between the three of 
you, each of you asking each other a question, and then we will 
wrap it up. Mr. Mills.
    Mr. Mills. Thank you. I think the critical analytic 
question to ask Mr. Romm is that while it is clear that there 
are efficiency effects from the use of Internet, it is A minus 
B equals C, C being the net result. You have to know what A is 
first. You have to know how much energy the Internet uses to 
calculate how much energy it saves.
    So I guess my question is really simple. Does your 
organization intend to try to figure out on their own, 
independent of my analysis, through their own analysis, perhaps 
with Lawrence Berkeley, to figure out how much electricity the 
Internet uses?
    Mr. Romm. Well, as you know, we have had many e-mail 
exchanges on this very point. I think Lawrence Berkeley labs is 
one of the most recognized authorities on how the economy uses 
electricity and energy. EIA uses them, everybody uses them. The 
international energy bodies use them.
    Their five scientists did a very comprehensive analysis, in 
an admittedly difficult area. They came to the conclusion that 
you were wrong by a factor of 8 and that they won't say 
specifically, but I am certainly prepared to say that the 
Internet, specifically the Internet uses about 1 percent of 
U.S. electricity maximum.
    However, you know, I think this is an important point for 
the committee. If Mr. Mills were right and his numbers were 
correct, which is that electricity consumption has soared since 
1997 because of the Internet, these numbers couldn't be true. 
So the fact that electricity consumption rates have slowed 
since the Internet took off is, in my mind, prima facie 
evidence that he can't possibly be right. So he can engage in a 
very complicated and elaborate analysis to show that the Sun 
goes around the Earth, but the Sun doesn't go around the Earth, 
and so his analysis is pointless. We have put on the table our 
explanation of why his methodology is wrong and why he comes up 
with the wrong answer, which the macroeconomic data clearly 
shows is wrong. So I don't think we need to do any more work.
    Mr. Ryan. In keeping with the Iowa and the New Hampshire 
tradition, we will give the questioner a 30 second rebuttal.
    Mr. Mills. I think 2 things are relevant to this. The 
adjective ``soar'' is an adjective. The electric use went up 
when all of the forecasters in Mr. Romm's camp said it would 
stop going up. More to the point, I will take the 1 percent, 
that is fine. Given the growth rates on the Internet and let's 
say it is starting at 1 percent today, I will tell you that we 
just have to wait a few years, and it will be 8 percent and 
then more. Because the growth rate is astonishing.
    So the real issue is not so much whether it is 1 or 2 or 3 
or 4, but that it is a positive integer. I am glad to hear Mr. 
Romm recant the ``0'' that is in his study.
    Mr. Ryan. Mr. Romm, would you like to ask Mr. Hakes a 
question?
    Mr. Romm. I realize, to use the full New Hampshire format, 
he asked for the specific number of the Internet use of 
electricity: 1 percent. The net impact of the Internet on 
electricity which would include not only the energy used by the 
components of the Internet, but what the impact of the Internet 
is on the economy, I believe the Internet saves electricity and 
far more than electricity, it saves energy. So that having the 
Internet is why the electricity growth has slowed.
    The question that I would ask Mr. Hakes is, you know, first 
of all--let me think about this for a second.
    If it were the case--well, let me ask you this question. If 
it were the case that the energy intensity were, in fact, 
changing in the last 3 years, and that, in fact, energy 
intensity has averaged from 1997, 1998, and 1999 about 3.5 
percent, although you project that it is going to improve 1.1 
percent, so we think, you know, currently you have been wrong 3 
years in a row by a factor of 3, if it were the case that that 
was happening, would EIA--and we could convince you that that 
was happening, would EIA be willing to modify its forecasts?
    I take your point that you have cited, very high energy 
intensity declined in the 1980's, but that occurred when oil 
prices were doubling and tripling. We have never had energy 
gains this big when energy prices were low and even declining. 
So that is my question. Is EIA open to exploring this very 
important issue and perhaps changing its forecasts?
    Mr. Hakes. We have extensive internal dialogs about these 
issues and external dialogs, and we certainly would take this 
into consideration. But I think again you are mixing apples and 
oranges. It makes a big difference if the intensity improves 
because the economy is going up or it improves because energy 
use is going down. Because if, for instance, you take the 
scenario that energy use stays about the same and the economy 
goes up, then you would be asking us to adjust our intensity 
estimates, but you would also be asking us to adjust our GDP 
estimates. We estimate GDP to grow at about 2.3 percent a year. 
So if Mr. Romm wants us to take the last year as the 
statistical call trend, we would then have to raise our GDP 
rates to about 4 percent a year, which might conceivably make 
it harder to meet the Kyoto protocol.
    Now, if there was a comparable gain in intensity, it would 
be a wash.
    So you can't just change one part of the equation and take 
the part you like and not add in the other point you don't 
like. So are you arguing, Joe, that we have too low a growth 
rate for the economy and we should jack up economic growth 
which will create more new energy usage than we have in our 
model?
    Mr. Romm. Actually I am arguing that EIA has done what you 
just accused me of doing. In fact, if you look at EIA's 
forecasts this year and last year, they jacked up the GDP level 
in 2005, but they didn't change the energy intensity level. So 
they actually predict, even though we have had 3 amazing years, 
they have predicted higher GDP growth, but no improvement in 
energy intensity, which is why this year's forecast, even 
though oil prices are higher than they were a year ago and even 
though they have another year's data that something is going on 
in the economy, they actually raised GDP, but they don't raise 
the energy intensity, which is why they have higher 
CO2 levels predicted this year for 2005 than they 
did last year, which I would argue suggests that they are 
defending a perspective which is to say higher CO2 
growth rates, as opposed to saying gee, maybe something is 
happening in the economy.
    Mr. Ryan. Since we are talking about your chart so much, 
why don't we have Mr. Hakes ask Mr. Romm a question and then we 
will conclude.
    Mr. Hakes. Back in 1996, carbon emissions went up about 3 
percent and EIA took the position that a lot of that was 
weather related, and we didn't jack up our carbon emissions 
growth-rates. In fact, we, I think, came down a little bit, 
because we thought that was a specific aberration. Why didn't 
you come to us in 1996 and express your concerns that we were 
underestimating carbon growth because the data for that year 
showed it was much higher than were in the EIA estimates?
    Mr. Romm. Well, I take your point that 1 year's worth of 
data is not something that I would change your forecasts on. We 
started this report when we had about 2\1/2\ years worth of 
data. I think we now have 3 years worth of data.
    You can certainly say that OK, we have to wait for a 4th 
year and then a 5th year and a 6th year. I would pose that 3 
years worth of data is very impressive. We have never in U.S. 
history seen this improvement in how the Nation uses energy at 
a time of low energy prices. The only data that he could cite 
was 1982. Oil prices doubled in the early 1970's, they doubled 
again in the late 1970's, this is why energy intensity 
improved. He has to explain how it is that we had GDP go up and 
energy growth rates go down.
    As I say, we have talked to EIA to understand how they do 
weather analysis. When you do the weather correction, I will be 
happy to go over this with you, because we used your numbers in 
our second case, it only gets you up to 1 percent. The fact is 
that over the last 3 years, weather-adjusted energy growth has 
been under half of what the weather-adjusted energy growth was 
in the previous 4 years.
    Mr. Hakes. I would point out that the period you cover just 
coincidentally happens to be a period that was highly unusual 
historically in not having a heavy cold snap. It will be very 
interesting to look at the 2000 data where we obviously have 
had some months here where we have had heavy demands for 
heating oil and natural gas that we haven't seen for several 
years, and I think that will balance out our perspective quite 
a bit.
    Mr. Ryan. That was going to be my last question. Mr. Romm, 
you said something that intrigued me. You made an assumption 
that Internet efficiencies are going to lead to less energy 
consumption. I would like to ask the other 2 gentlemen, Mr. 
Mills to start with, if the digital economy increases wealth, 
will it or will it not increase demand? If the digital economy 
makes us wealthier, won't this increase the demand for TVs, 
computers, cars, air travel, energy-producing products?
    I represent the First Congressional District of Wisconsin. 
We produce the Chevy Tahoe, Suburban, medium-duty trucks, the 
Jeep Cherokee, the Wrangler, SUVs which are gas-guzzling 
vehicles. We are selling them like hot cakes. It is producing a 
lot of jobs where I live. Could you comment on that?
    Mr. Mills. Sure, and really there are two issues. The one 
issue is the wealth effect. I think there is no question that 
the digital economy is driving the wealth effect; and as Mr. 
Hakes has said, that is one of the complications. As you push 
the GDP up because of the lubrication, if you like, of the 
information economy, you get more purchases of SUVs, bigger 
houses, more renovations, more travel. In fact airline travel 
is up, driving is up, everything is up.
    So the measures of efficiency are really misleading. Yes, 
there is so much more money which drives up the energy 
efficiency metric. The efficiency of driving measured in miles 
per dollar is better, but that doesn't matter. Driving is up.
    The narrow point, of course, is that in the electric area, 
that as you keep adding this infrastructure, you get net more 
demand for electricity. Yes, it drives efficiencies in oil and 
transportation; yes, it drives efficiency, it controls lights 
better. But what we already see happening is that the growth 
rate is still a growth. So that the fundamental problem we had, 
and what I heard in this exchange just now between Mr. Hakes 
and Mr. Romm, is measuring the growth, in fact the reason we 
started our analysis.
    Much of the energy discussion that is going on triggered by 
Kyoto is locked in a historical way of looking at our energy 
economy and has not fully accommodated the profound structural 
changes that the information age has brought. It specifically 
has not accommodated the demand side of it. Not because it is 
going to necessarily soar--I used the word ``soar'' because I 
believe that is a big increase over what would have been 
claimed to be zero. But for me, a growth equal to all the 
electricity of Central and South America, that is soaring. But 
I use that phrase to point out this is an enormously important 
growth held against the desire to back out the coal industry.
    So the answer to the question is I am very confident in 
saying that the wealth effect will keep driving electric 
growth, that I am confident the telecom sector will be the 
recipient.
    Mr. Ryan. Mr. Hakes.
    Mr. Hakes. Well, I think motor vehicles are a very 
interesting issue, because it is something we all deal with 
every day. Mr. Romm was pointing out that he sees some tendency 
here of improved intensity. We are still taking some pre-1988 
vehicles off the road which does create some efficiency gains. 
But the fact is that new vehicles today are less efficient than 
ones from the early 1990's.
    And what happened is we have had a lot of advanced 
technologies go into these vehicles, but we have had a lot of 
service requirements added on to the vehicles and those seem to 
be offsetting these new technologies. So where this is going to 
come from, it certainly is not showing up in the data at this 
point. The new vehicles today are less efficient than the new 
vehicles in the early 1990's.
    Mr. Romm. I this--I--the Internet can't affect the 
efficiency of vehicles, but it can affect how they are used. 
EIA in October--excuse me, in November, acknowledged that there 
has been a break-off in the historical relationship between 
economic growth and transportation. It is very clear, although 
it is too early to say it is a trend, because it is only 1 or 2 
years' data, that vehicle miles traveled in the last 2 years 
have slowed noticeably.
    I won't repeat all of the ways that the Internet economy 
increases efficiency. I had the 30,000 word report on that 
subject. Compared to traditional companies, Internet firms 
require less square footage and under one-tenth of the building 
energy consumption per dollar of sales; companies are using the 
Internet to cut inventories 25 to 50 percent; and more or more 
firms like IBM and AT&T are reducing square footage for their 
mobile workers because of the Internet. Some firms are even 
auctioning off empty space on cargo trucks, I am sure you read 
about that, making the freight system more efficient.
    So I believe that we have only scratched the surface in 
understanding all of the ways that the Internet is making the 
economy more efficient. But I put it to anyone else to explain 
what is going on in the economy, if the Internet isn't playing 
a role in making it more efficient.
    Mr. Ryan. Those are interesting anecdotes, and I wonder if 
they encapsulate the whole picture. I was just wondering, you 
said 1 or 2 years of data doesn't indicate a trend in the 
automobile industry, but it does in your chart here. I find 
that to be quite a contradiction.
    Mr. Romm. I think 3 years worth of data in the entire U.S. 
economy suggests something big is going on. We titled our 
report a scenario, because I don't think anyone can testify 
that they know exactly what is going on. I mean I think we know 
enough to know that Mr. Mills has to be wrong, and that Mr. 
Hakes's forecasts are probably wrong. We have offered the best 
explanation for something big going on in the economy. I am not 
certain that--and we just want people to say gee, we better 
look at this because it is important.
    Mr. Ryan. Well, I appreciate your candor.
    Mr. Hakes. His saying that I am probably wrong is the 
nicest thing he has said all day.
    Mr. Ryan. I will let Mr. Mills have the final say because 
of that salvo.
    Mr. Mills. Let me just end with a Lawrence Berkeley number 
to put the whole thing into perspective. I agree that the 
Internet drives efficiency. That is fine. But Lawrence Berkeley 
guys said that the commercial sector's use of computers: PCs, 
monitors, printers collectively in 1995 was 50 billion kilowatt 
hours. That was Lawrence Berkeley in 1995 for 1993 consumption 
of that whole class, not just PCs. Seven Internet years ago. My 
study says that that class of devices not just nor the Internet 
is up in the 300 to 400 billion kilowatt hour range. I would 
just submit to you that in that 7 Internet years, I don't think 
it is obvious that I am obviously wrong.
    Mr. Ryan. Point taken.
    Gentlemen, thank you very much for very interesting 
testimony. This hearing is adjourned.
    [Whereupon, at 12 noon, the subcommittee was adjourned.]
    [Additional information submitted for the hearing record 
follows:]
[GRAPHIC] [TIFF OMITTED] T4876.058

[GRAPHIC] [TIFF OMITTED] T4876.059

[GRAPHIC] [TIFF OMITTED] T4876.060

[GRAPHIC] [TIFF OMITTED] T4876.061

[GRAPHIC] [TIFF OMITTED] T4876.062

[GRAPHIC] [TIFF OMITTED] T4876.063

[GRAPHIC] [TIFF OMITTED] T4876.064

[GRAPHIC] [TIFF OMITTED] T4876.065

[GRAPHIC] [TIFF OMITTED] T4876.066

[GRAPHIC] [TIFF OMITTED] T4876.067

[GRAPHIC] [TIFF OMITTED] T4876.068

[GRAPHIC] [TIFF OMITTED] T4876.069

[GRAPHIC] [TIFF OMITTED] T4876.070

[GRAPHIC] [TIFF OMITTED] T4876.071

[GRAPHIC] [TIFF OMITTED] T4876.072

[GRAPHIC] [TIFF OMITTED] T4876.073

[GRAPHIC] [TIFF OMITTED] T4876.074

[GRAPHIC] [TIFF OMITTED] T4876.075

[GRAPHIC] [TIFF OMITTED] T4876.076

[GRAPHIC] [TIFF OMITTED] T4876.077

[GRAPHIC] [TIFF OMITTED] T4876.078

[GRAPHIC] [TIFF OMITTED] T4876.079

[GRAPHIC] [TIFF OMITTED] T4876.080

[GRAPHIC] [TIFF OMITTED] T4876.081

[GRAPHIC] [TIFF OMITTED] T4876.082

[GRAPHIC] [TIFF OMITTED] T4876.083

[GRAPHIC] [TIFF OMITTED] T4876.084

[GRAPHIC] [TIFF OMITTED] T4876.085

[GRAPHIC] [TIFF OMITTED] T4876.086

[GRAPHIC] [TIFF OMITTED] T4876.087

[GRAPHIC] [TIFF OMITTED] T4876.088

[GRAPHIC] [TIFF OMITTED] T4876.089

[GRAPHIC] [TIFF OMITTED] T4876.090

[GRAPHIC] [TIFF OMITTED] T4876.091

[GRAPHIC] [TIFF OMITTED] T4876.092

[GRAPHIC] [TIFF OMITTED] T4876.093

[GRAPHIC] [TIFF OMITTED] T4876.094

[GRAPHIC] [TIFF OMITTED] T4876.095

[GRAPHIC] [TIFF OMITTED] T4876.096

[GRAPHIC] [TIFF OMITTED] T4876.097

[GRAPHIC] [TIFF OMITTED] T4876.098

[GRAPHIC] [TIFF OMITTED] T4876.099

[GRAPHIC] [TIFF OMITTED] T4876.100

[GRAPHIC] [TIFF OMITTED] T4876.101

[GRAPHIC] [TIFF OMITTED] T4876.102

[GRAPHIC] [TIFF OMITTED] T4876.103

[GRAPHIC] [TIFF OMITTED] T4876.104

[GRAPHIC] [TIFF OMITTED] T4876.105

[GRAPHIC] [TIFF OMITTED] T4876.106

[GRAPHIC] [TIFF OMITTED] T4876.107

[GRAPHIC] [TIFF OMITTED] T4876.108

[GRAPHIC] [TIFF OMITTED] T4876.109

[GRAPHIC] [TIFF OMITTED] T4876.110

[GRAPHIC] [TIFF OMITTED] T4876.111

[GRAPHIC] [TIFF OMITTED] T4876.112

[GRAPHIC] [TIFF OMITTED] T4876.113

[GRAPHIC] [TIFF OMITTED] T4876.114

[GRAPHIC] [TIFF OMITTED] T4876.115

[GRAPHIC] [TIFF OMITTED] T4876.116

[GRAPHIC] [TIFF OMITTED] T4876.117

[GRAPHIC] [TIFF OMITTED] T4876.118

[GRAPHIC] [TIFF OMITTED] T4876.119

[GRAPHIC] [TIFF OMITTED] T4876.120

[GRAPHIC] [TIFF OMITTED] T4876.121

[GRAPHIC] [TIFF OMITTED] T4876.122

[GRAPHIC] [TIFF OMITTED] T4876.123

[GRAPHIC] [TIFF OMITTED] T4876.124

[GRAPHIC] [TIFF OMITTED] T4876.125

[GRAPHIC] [TIFF OMITTED] T4876.126

[GRAPHIC] [TIFF OMITTED] T4876.127

[GRAPHIC] [TIFF OMITTED] T4876.128

[GRAPHIC] [TIFF OMITTED] T4876.129

[GRAPHIC] [TIFF OMITTED] T4876.130

[GRAPHIC] [TIFF OMITTED] T4876.131

[GRAPHIC] [TIFF OMITTED] T4876.132

[GRAPHIC] [TIFF OMITTED] T4876.133

[GRAPHIC] [TIFF OMITTED] T4876.134

[GRAPHIC] [TIFF OMITTED] T4876.135

[GRAPHIC] [TIFF OMITTED] T4876.136

[GRAPHIC] [TIFF OMITTED] T4876.137

[GRAPHIC] [TIFF OMITTED] T4876.138

[GRAPHIC] [TIFF OMITTED] T4876.139

[GRAPHIC] [TIFF OMITTED] T4876.140

[GRAPHIC] [TIFF OMITTED] T4876.141

[GRAPHIC] [TIFF OMITTED] T4876.142

[GRAPHIC] [TIFF OMITTED] T4876.143

[GRAPHIC] [TIFF OMITTED] T4876.144

[GRAPHIC] [TIFF OMITTED] T4876.145

[GRAPHIC] [TIFF OMITTED] T4876.146

[GRAPHIC] [TIFF OMITTED] T4876.147

[GRAPHIC] [TIFF OMITTED] T4876.148

[GRAPHIC] [TIFF OMITTED] T4876.149

[GRAPHIC] [TIFF OMITTED] T4876.150

[GRAPHIC] [TIFF OMITTED] T4876.151

[GRAPHIC] [TIFF OMITTED] T4876.152

[GRAPHIC] [TIFF OMITTED] T4876.153

[GRAPHIC] [TIFF OMITTED] T4876.154

[GRAPHIC] [TIFF OMITTED] T4876.155

[GRAPHIC] [TIFF OMITTED] T4876.156

[GRAPHIC] [TIFF OMITTED] T4876.157

[GRAPHIC] [TIFF OMITTED] T4876.158

[GRAPHIC] [TIFF OMITTED] T4876.159

[GRAPHIC] [TIFF OMITTED] T4876.160

[GRAPHIC] [TIFF OMITTED] T4876.161

[GRAPHIC] [TIFF OMITTED] T4876.162

[GRAPHIC] [TIFF OMITTED] T4876.163

                                   - 
