Chairman Nussle, Ranking Member Spratt, and Members of the Committee:
Thank you for inviting me to testify today. It is always an honor to appear before this committee. I applaud the Committee’s efforts to focus on the important role of uncertainty in shaping budget outcomes and policies. I’d like to address four questions.
The first question is how good are the growth assumptions and baseline projections made by the Congressional Budget Office? I think it is fair to say that the CBO does a very good job of forecasting future economic activity and its relation to revenues and spending. The CBO’s projections are often wrong, sometimes by sizable amounts. It is particularly difficult to predict and understand turning points. But my impression is that CBO does as well as, or better than, anyone else who tries to divine the economic future.
The second question is what causes the baseline projections to be so uncertain? The primary source of uncertainty in the baseline projection is uncertainty regarding the overall size of the economy in the future. Not only is future GDP difficult to predict precisely, but small differences in assumed growth rates can generate large changes in budget outcomes.
For example, in January, CBO predicted a real growth rate averaging 3.0 percent per year over the next decade. This is a reasonable estimate, but everyone would agree there is substantial uncertainty surrounding the exact figure. If the annual rate of economic growth turns out to be just three-tenths of a percentage point larger or smaller than CBO projects, the difference in the deficit will be about $700 billion over the next 10 years (CBO, Economic and Budget Outlook for Fiscal Years 2005-14, January 2004, Appendix B).
Compared to the size of the economy, uncertainty about other economic factors is usually significantly less important in generating budget projections. Such factors include the rate of inflation, the level of interest rates, and the level of the stock market. While it is true that changes in capital gains have significantly influenced revenues over the past several years, it has required enormous, atypical changes in capital gains to have that effect, and the effects have been far smaller than those due to changes in the size of the economy. Typical changes in inflation, interest rates, and the stock market generate modest changes in budget projections.
Given the size of the economy, and given the other economic factors—like capital gains, inflation, and interest rates—the actual structure of tax policy plays a very small role in generating uncertainty in the baseline projections. To be more specific, the current tax system taxes different people and different forms of income at different rates. Even if we taxed everyone at the same rate on all activities, the great majority of uncertainty would still be present in budget projections.
A third issue is that the baseline projection is becoming an increasingly misleading measure of realistic budget outcomes. This increases everyone’s uncertainty regarding what the actual budget outcome will be, given the baseline. This trend is due to several factors, most notably: (i) the enormous increase in the use of expiring tax provisions and sunsets in the tax code; (ii) the presence of specific unsustainable policies, like the built-in growth of alternative minimum tax, which no one believes will be allowed to persist; and (iii) the presence of an overall unsustainable fiscal policy stance, such as we now face. When policy-makers enact laws that are fiscally unsustainable or that are designed in a blatant manner to skirt budget rules, the effect is to increase uncertainty about the course of future federal policy. This likely has negative effects on the economy.
The fourth issue is what to do about uncertainty in the budget projections. There are two policy questions: how to reduce uncertainty and how to respond to whatever uncertainty is left in the system.
If the main source of uncertainty is the economy itself, then policy-makers will always face a significant amount of uncertainty. But policy-makers can still reduce uncertainty in several ways. The best way to reduce uncertainty is avoid creating fiscally unsustainable policies. A second way is to avoid creating artificial sunsets.
These issues will come to a head in dealing with whether to make the 2001 and 2003 tax cuts permanent. At first glance, one might think that making the tax cuts permanent would reduce uncertainty since it indicated a long-term Congressional commitment to lower tax rates. Making the tax cuts permanent, however, will substantially increase the severity of the fiscal gap between revenues and spending and hence it will serve to increase uncertainty about how and when the overall fiscal gap will 3 be closed. In fact, making the tax cuts permanent in light of the current fiscal situation is sufficiently irresponsible that it may lead to investors wondering not just how and when the fiscal gap will be closed, but whether it will ever be closed. That is, several knowledgeable observers, most notably Treasury Secretary Robert Rubin, have raised the concerns that ever-increasing federal debt projected for the medium- and long-term could seriously disrupt financial markets and the broader economy.
In contrast to extending the tax cuts, letting them expire would reduce uncertainty about the tax system as well as reduce uncertainty about the overall fiscal gap. Likewise, establishing a budget-conscious solution to the AMT problem would go a long way toward reducing uncertainty about future tax rates and reduce uncertainty regarding whether, how, and when the fiscal gap will be closed.
Policy-makers should also shape policies explicitly in light of the fact that uncertainty is present. First, I think that realistic budget projections should be given more weight in the policy process than they currently are. The baseline is useful for understanding the effects of legislation, but is extremely misleading as a guide to what likely budget outcome are going to be. The reason why is that the baseline is constructed according to certain fixed rules for projecting spending and revenues that do not reflect likely or plausible outcomes. If policy-makers paid more attention to alternative, realistic forecasts of likely budget outcomes rather than the baseline, they would not be so surprised by many of the changes in the budget projections.
Second, policy makers should distinguish between projections that are more certain and those that are less certain. Population growth, for example, is relatively certain in the medium term. If you know how many 50 year olds are alive today, chances are you can make a pretty good guess at how many 70 year olds will be alive in 20 years. On the other hand, if you know the size of the economy today, that may provide very little guidance for estimating the change in the size of the economy over the next 20 years. Along similar lines, almost every study finds that the nation faces a substantial long-term fiscal shortfall, even though there is substantial uncertainty about many economic factors over the long-term. Simply put, population aging and health care technologies will create budgeting problems under almost any scenario.
Third, Congress should consider ways to adopt budget rules that not only limit both tax cuts and spending increases but also do so in a way that does not let every available dollar be allocated as soon as it appears. Families face substantial economic uncertainty about future wages, health, asset returns, and so on. They often manage that uncertainty by trying to keep a financial reserve. A literal financial reserve might be hard to establish at the federal level, but budget rules that limited the extent to which policymakers could create future tax cuts and spending increases would be a step in the right direction.
My last point is that perhaps the biggest danger in acknowledging the importance of uncertainty is that it will allow policy makers to shrug off any potential fiscal problems with the insight that the problems might go away on their own. Current, plausible budget 4 projections suggest significant medium-term deficits and very sizable long-term deficits. But the presence of uncertainty suggests that these problems could conceivably go away on their own. This observation might tempt politicians to delay seeking solutions, hoping to avoid the daunting economic and political risks associated with large-scale tax increases and spending cuts. Indeed, in the last few years, the majority of legislators have chosen not only to ignore the long-term fiscal problems but to make them substantially worse by enacting and advocating significant tax cuts and spending increases. But it is extremely unlikely that the current fiscal problem will resolve itself without policy actions, and delays in addressing the issue will only make the eventual solutions more extreme and painful. It would be a travesty if the explicit recognition of uncertainty in the policy process led policy-makers to choose to avoid dealing with the serious fiscal problems the country faces.