Chairman Dreier, Representative Moakley, Members of the Committee:

I appreciate the opportunity to testify before the Committee today on the subject of biennial budgeting. I am Jim Horney, Senior Fellow at the Center on Budget and Policy Priorities here in Washington, D.C. The Center is a nonprofit policy institute that works on an array of public policy issues, with a particular interest in matters of fiscal policy and policy impacts on low- and moderate-income families. The Center receives no federal grants or contracts.

Along with this testimony I would like to submit a paper on biennial budgeting written by Bob Greenstein, the Executive Director of the Center on Budget and Policy Priorities. That paper is a slightly revised version of a paper published last March, before I joined the Center. I am in complete accord with the points made in that paper — many of which have also been made by previous witnesses before this Committee — and will briefly outline the major ones here:

• Biennial budgets will probably have to be revised frequently. This will both reduce the likelihood that more time will be available for the executive branch and the Congress to devote to other activities, and increase the extent of budget decisions that are made on an ad hoc basis rather than as part of the more comprehensive examination of budget priorities that the regular budget process provides.

   

• Biennial budgeting is unlikely to increase effective Congressional oversight. Congressional authorizing committees are unlikely to substantially increase their oversight efforts since the time they can devote to such activities now is not significantly limited by the current annual budget process. Furthermore, the important oversight functions carried out by the appropriations committees in their annual review of the budget would be reduced.

   

• The experience in the states indicates that biennial budgeting does not work well for entities with large, complex budgets. Since 1940, the number of states using biennial budgeting has been cut by more than half. Most of the states still using biennial budgeting are small or medium states, which do not represent a good model for the federal government.

   

• The disadvantages of biennial budgeting are likely to outweigh the advantages.

Rather than going through all of these arguments against biennial budgeting in more detail, I want to focus on one particular problem with a two-year budget cycle that I am very familiar with — the likelihood that the budget outlook will change dramatically from the time that Congressional consideration of a two-year budget begins until the second year of that budget cycle has been completed.

Prior to joining the Center last July, I spent more than seven years at the Congressional Budget Office. During that time, I was in charge of the unit responsible for coordinating CBO's baseline budget projections. In that position, I could hardly fail to be struck by the dramatic changes that were occurring in the budget outlook. Each time CBO updated its baseline projections, I had the task of explaining how and why the projections of revenues, outlays, and surpluses or deficits had changed by scores, or even hundreds, of billions of dollars in just a few months. I firmly believe that those changes occurred not because CBO was failing to do its job properly, but because the United States economy and the federal budget are so large, so complicated, and so dynamic that no person or organization will ever be able to project economic or budget outcomes with any degree of certainty.

For instance, just since last March CBO has increased the projected surplus for fiscal year 2000 by $84 billion and for fiscal year 2001 by $105 billion — not including the effects of legislation enacted during the last year. Most of this change stems from an increase in revenues projected those years, which is largely the result of the economy growing more rapidly than anticipated. In light of recent experience, it might seem that budget projections always get better. But that is clearly not the case — as recently as the early 1990s, there was a long period during which the budget outlook got progressively worse. For instance, from March 1990 to March 1991, CBO increased its projection of the total budget deficit for fiscal year 1991 by $181 billion and for 1992 by $238 billion.

Those changes in estimates are not unique. In a very revealing chapter on the "Uncertainties of Budget Projections" in the Budget and Economic Outlook: Fiscal Years 2001-2010 it released in January, CBO analyzed the errors in the budget projections it has produced each winter since 1986. CBO first calculated the difference between the deficits projected each year for six fiscal years (the then-current fiscal year, the fiscal year beginning the following October, and the succeeding four fiscal years) and the actual outcomes for each fiscal year from 1986 through the most recently completed fiscal year (1999). Next, CBO excluded the differences due to the effects of legislation enacted after the projections were made. Finally, it calculated the absolute average error (the average of the differences without regard to whether the projected deficit was higher or lower than the actual outcome) of the deficit projections for each of the six different time periods covered each year. That is, CBO calculated the average absolute error of projections for fiscal years that began the October 1 before the projections were made (the current year), the average absolute error of projections for fiscal years starting October 1 of the same year the projections were made (the budget year), the absolute average error of projections for fiscal years starting October 1 of the next year (the first outyear), and so on for projections for each fiscal year through the fourth outyear.

CBO found that the absolute average error for the projections of the deficit for the budget year (the year starting about eight or nine months after the projections were made) is equal to 1.1 percent of gross domestic product (GDP) for that fiscal year. Based on CBO's current economic forecast, an average error in the surplus projected for fiscal year 2001 (the budget year right now) would be $112 billion. CBO also found that the absolute average error for the deficit projected for the following fiscal year (the first outyear) was even larger — 1.6 percent of GDP. Thus, for fiscal year 2002, which will begin more than 20 months from now, an average error in the projected surplus would produce a swing of $170 billion — that is, if CBO is as accurate as it has been on average over the last 14 years, we can expect that the total budget deficit will be $170 billion higher or lower than the $212 billion CBO is currently projecting for 2002 (not including the effects of legislation that may be enacted).

This is not intended as a criticism of CBO — particularly since I was at least partially responsible for some of the projections that have been so far off. It is simply intended to indicate that projecting the course of the federal budget is highly uncertain; the best possible estimates are still likely to be wrong by many billions of dollars.

Congress cannot do anything about the uncertainty of budget projections. You can, however, choose how to deal with that uncertainty when deciding how the budget process should work. And I think that it is reasonable to ask whether locking in a budget plan for two years is the appropriate response to such a high degree of uncertainty. I have often heard Members of Congress argue that the government should be run more like a business. It can be argued that businesses today are facing more uncertainty than ever in trying to plan for their future — for instance, even just a very few years ago who imagined the opportunities and the challenges that the internet presents for American business today. How are successful businesses responding to this uncertainty? By increasing their flexibility and responding more rapidly to changes in their environment. They have not responded by reducing how often they thoroughly reconsider their business plans, although they might extend the time horizon of those plans. It is hard to imagine the CEO of any corporation telling his stockholders that in this dynamic economy the company will be well served by locking in place for two years a business plan based on today's market outlook, spending a year reviewing whether that plan is working, and only then making significant changes in that plan.

There may be good reasons for the Congress not to revise the federal budget as often as some companies change their business plans, but it is hard for me to believe that it is best for the Congress to respond to the changing budget situation — not to mention the changing needs of American citizens — only every two years.

Of course, most biennial budget plans do not preclude changes in the budget in the non-budget year. Budget resolutions can be revised, reconciliation bills pursuant to instructions in the revised resolution can be considered, and supplemental appropriations can be enacted. But if such revisions and new budget legislation are routine, then the promised benefits of biennial budgeting will certainly not be achieved. Those benefits can only occur if Congress does not spend any appreciable time on the budget in the non-budget year, but spending little or no time on the budget every other year is not the best response to the constantly changing economic and budget outlook.

Some might argue that biennial budgeting is worth trying because the worst that will happen is that the Administration and the Congress will end up spending just as much time on the budget in the off year as they spend each year now. Therefore, no harm would be done. But biennial budgeting could do harm to the process. For instance, the President and Congress could fail to make needed changes in the budget in the second year, with potentially serious results. But, even if they do end up making extensive changes in the budget each year to respond to the changing outlook and needs, the resulting decisions might be less desirable than they would be under the current process. Because the time spent on the budget in the off year will not be part of a thorough, structured review of the budget, decisions about changes in the budget in the second year will be made on a more ad hoc basis than occurs under the current system. Thus, not only might the potential benefits of biennial budgeting — more time for other activities — fail to materialize, but budget decisions may be less well-considered and less responsive to the needs of the American public.