[Congressional Record (Bound Edition), Volume 147 (2001), Part 7]
[House]
[Pages 10124-10125]
[From the U.S. Government Publishing Office, www.gpo.gov]



    CONGRESS MUST HOLD FORECASTERS ACCOUNTABLE FOR THEIR PROJECTIONS

  The SPEAKER pro tempore. Under a previous order of the House, the 
gentleman from Illinois (Mr. Kirk) is recognized for 5 minutes.
  Mr. KIRK. Madam Speaker, we must hold forecasters accountable for the 
accuracy of their projections. As we are asking for straight A 
performance out of our public schools, we must also ask that out of our 
budget forecasters. We want better and more efficient use of energy 
resources.
  As Secretary Rumsfeld is completing a comprehensive overall of our 
defense network, how can we expect anything less than continuous 
improvement

[[Page 10125]]

from the way that we prepare the Federal budget? And we have a long way 
to go.
  Everyone I talk to in Washington assumes that budget forecasts we use 
are setting priorities that are wrong; that they can be way off the 
mark; that we never are able to estimate correctly what our financial 
status is.
  In 1997, the Congressional Budget Office estimated a $145 billion 
deficit for fiscal year 1998. We had a surplus of $69 billion. In 1999, 
CBO predicted a $107 billion surplus for fiscal year 2000, $129 billion 
below the actual $236 billion achieved. You can see it here on chart 
number one, where CBO estimates a $211 billion deficit, it was only 
$107.

                              {time}  1400

  Then a $156 billion deficit, it was only 22. The biggest year they 
made a mistake was 1998; they forecast a $145 billion deficit. We ran a 
$69 billion surplus. And on and on the errors have gone.
  Mr. Speaker, this is no way to fill our elected mandate of keeping 
the economy strong. There is more at stake than the issue of whose 
numbers are right. Congress uses these estimates to make key decisions 
about tax policies that encourage economic growth, foster 
entrepreneurship, and reward individuals for seeking opportunities to 
work, learn and get ahead.
  Inaccurate forecasts end up crowding out uses of other Federal funds. 
If defense programs produce large cost overruns, then less money is 
left for new education projects. If the actual cost of Medicare part B 
programs often exceed preliminary estimates, it becomes harder to build 
support for new benefits such as a prescription drug benefit. Better 
forecasts should be a bipartisan initiative focused on the goal of 
making government more effective.
  Some errors of the past can be blamed on estimates that rely on 
status quo analysis, assuming that taxpayers will not change their 
actions in response to legislative changes that affect their 
pocketbook. Such a projection applies recent growth rates to baseline-
year figures, assuming that current trends will continue indefinitely. 
Common sense tells us when you increase taxes on something, such as 
saving and investment, you get less of it. A change in tax policy 
influences the decisions that individuals make, thereby affecting 
revenues.
  The recent history of the capital gains tax policy shows the 
shortcomings of status quo analysis. In 1984, Congress passed the 
Deficit Reduction Act, which temporarily reduced the long-term capital 
gains holding period from 12 months to 6 months, making it easier for 
investors to qualify for preferential tax treatment. Investors reacted, 
and quickly.
  Capital gains realizations in 1985 were twice the amount in 1984. 
However, investor euphoria was short-lived. Congress repealed the 
capital gains deduction as part of the Tax Reform Act of 1986. Our 
budget experts prepared status quo estimates that anticipated large 
Federal revenue gains from a higher capital gains tax. Quite the 
contrary happened. Capital gains realizations tumbled in 1987. Budget 
estimators were confounded by the fact that taxpayers acted to avoid 
taxes.
  Chart 2 shows the reaction.
  We projected as we raised taxes, that we would actually raise 
revenue. We did not. We lost it when we raised the tax on capital 
gains.
  The status quo then changed once again when we used the estimates and 
when we reduced capital gains charts. The status quo predicted a dismal 
drop in revenue. In actuality, capital gains realizations increased 
steadily and substantially, contributing to the surpluses we have now 
enjoyed, as you can see from this chart, where the realizations for 
fiscal year 2000, we projected $329 billion and we have $643 billion.
  In order to make the best decisions, Congress needs real-world 
estimates that account for the interaction between Federal taxes and 
Federal programs and individuals' behavior. We have just passed one of 
the largest tax relief packages in U.S. history without the benefit of 
real-world analysis that effectively forecasts the turning points that 
we can use.
  Under the current House rules, the chairman of the Committee on Ways 
and Means has the right to request real-world forecasts, and the Joint 
Committee on Taxation must provide them in a timely manner. This should 
be required, not optional, and should be used for all tax bills.
  The chairman of the Committee on Rules has introduced a capital gains 
tax reduction bill. Consider how a status quo analysis would misguide 
us on examining that legislation. Budget accuracy will be achieved with 
small steps, and we need it now.
  This is a job for innovators ready to meet the challenge of helping 
Congress spend taxpayers dollars wisely. As a start, we can improve 
budgeting accuracy by using projections that do not ignore changes in 
the behavior of individuals when taxes increase and decrease. next, we 
need to account for expenditure increases when the government 
establishes a program that ``pay for'' goods and services, thereby 
making them less expensive for individuals. The Joint Committee on 
Taxation and the Congressional Budget Office are developing models that 
incorporate certain ``real world'' assumptions to measure behavioral 
changes; however, we are just at the beginning of this process. As we 
move forward, it will be important to check ``projected'' against 
``actual'' results. By ``backcasting''--loading actual economic 
variables in models to determine how much the variability of particular 
assumptions affected the overall forecast--we can isolate the best of 
what we have and identify what areas of our forecast models need work. 
Third, we must give every federal agency the incentive to employ the 
assets they own to their highest and best uses. For example, the 
Defense Department owns major bands of Spectrum, but is unwilling to 
turn them over for commercial use; could this decision be based on the 
fact that it does not benefit from the sale of these assets?
  The next few years should be a time of testing new limits and 
learning from what does not work. In the end, our goal should be to 
``leave no Congress behind.'' The accuracy of the projections we work 
with will influence the quality of our policy decisions. Each Congress 
deserves the best it can get--and so do the American people. The right 
decisions will stand behind economic growth that benefits us all.

                          ____________________