For other versions of this document, see http://wikileaks.org/wiki/CRS-RL30948 ------------------------------------------------------------------------------ Order Code RL30948 CRS Report for Congress Received through the CRS Web Fashioning A Tax Cut Trigger: Economic Issues Updated January 29, 2002 Marc Labonte Economist Government and Finance Divison Congressional Research Service ~ The Library of Congress Fashioning A Tax Cut Trigger: Economic Issues Summary Congress is currently considering economic stimulus bills in response to the recession. Some members are concerned by the re-emergence of a federal budget deficit, and the possibility that these bills would make the deficit larger after the recession has ended. In addition, a large tax cut was passed in June 2001, the Economic Growth and Tax Relief Reconciliation Act (P.L.107-16), and the reductions in tax rates in this law will be phased in over several years. Several Members of Congress are interested in linking a "trigger" policy to these tax cuts, as recommended on several occasions by Alan Greenspan, Chairman of the Federal Reserve. Trigger policies would make the enactment of each year's phase-in contingent upon achieving yearly budget surplus or debt reduction targets. In other words, the phase-in would occur only if the surplus were large enough to make it affordable in the eyes of trigger proponents. They also note ways in which actual budget outcomes are likely to deviate from the baseline assumptions, due to both policy changes and unrealistic baseline assumptions. A trigger policy has several interesting economic implications: ! A simple trigger would be pro-cyclical. That is, it could prevent tax cuts from occurring during recessions, thus undesirably reducing aggregate spending further. More complexly crafted triggers may be able to take this problem into account. ! Most of the major provisions of P.L.107-16 are phased in by 2006. But 84% of the on-budget surplus occurs from 2007-2011. Thus, a trigger tied to P.L.107-16 may be unable to prevent the erosion of the bulk of the surpluses. ! Baseline forecasts are uncertain, and they become more error-prone when forecasted further into the future. While a trigger may be meant to compensate for this uncertainty, the mechanics of tying a trigger to an uncertain forecast can lead to unexpected outcomes. Notably, the choice of what target to link the trigger to leads to different budgetary outcomes if surpluses turn out to be smaller than forecasted. ! Some proponents of a tax cut who hope that it will increase long run growth by increasing saving and work incentives fear that a trigger will create uncertainty in decision making that will negate the tax cut's saving and labor effects. The report also offers an overview of the political issue concerning whether triggers can be circumvented through parliamentary maneuvers in the event that a target is missed. Critics claim that this has occurred when mechanisms similar to triggers have been used in the past. They fear that if this is the case, the trigger will be rendered ineffective. This report will be updated as events warrant. Contents Would a Trigger Be Pro-Cyclical? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 How Long Can the Trigger Effectively Last? . . . . . . . . . . . . . . . . . . . . . . . 3 Might a Trigger Policy Fail for Political Reasons? . . . . . . . . . . . . . . . . . . . 4 Uncertainty in Budget Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Should a Trigger Target the Past, Present, or Future? . . . . . . . . . . . . . . . . 5 Would a Trigger Create Uncertainty in Decision Making? . . . . . . . . . . . . . 8 What Number Should Be Targeted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 List of Tables Table 1: Comparing the Baseline Projection with the Administration Tax Cut . . 4 Fashioning A Tax Cut Trigger: Economic Issues Congress is currently considering economic stimulus bills in response to the recession. Some members are concerned about the re-emergence of a federal budget deficit, and the possibility that these bills would make the deficit larger after the recession has ended. In addition, a large tax cut was passed in June 2001, the Economic Growth and Tax Relief Reconciliation Act (P.L.107-16), and the reductions in tax rates in EGTRRA will be phased in over several years. Several Members of Congress are interested in linking a "trigger" policy to these tax cuts, as recommended on several occasions by Alan Greenspan, Chairman of the Federal Reserve. Trigger policies would make the enactment of each year's phase-in contingent upon achieving yearly budget surplus or debt reduction targets.1 In other words, the phase-in would occur only if the surplus were large enough to make it affordable in the eyes of trigger proponents. Since most of the provisions of the economic stimulus bills are temporary, it is more likely that a proposed trigger would be applied to EGTRRA. A primary purpose of a trigger is to allow some portion of the surplus to be maintained if economic projections turn out to be overly optimistic or if future policy decisions make the tax cuts unaffordable. They also note ways in which actual budget outcomes are likely to deviate from the baseline assumptions, due to both policy changes and unrealistic baseline assumptions.2 This report considers several interesting economic implications of formulating a trigger policy. These include the pro-cyclicality of a trigger policy, how the phase-in provisions of the tax cut correspond with the forecasted surpluses, political considerations concerning a trigger's effectiveness, the uncertainty inherent in budget projections, the policy implications of choosing different budgetary figures to target, and how triggers may affect individuals' decision-making. All of these considerations point out the complexity of designing an effective trigger. Would a Trigger Be Pro-Cyclical? Although policy decisions ultimately determine the budget balance, the budget surplus is also extremely sensitive to economic factors like growth, inflation, and 1 Different trigger policies target different variables. Some target the size of the surplus, some target the level of the publicly held national debt, and some target interest payments on the national debt. In practice, some variables may prove to be a more direct way to accomplish one's goals than others since each is sensitive to somewhat different factors. For the sake of simplicity, this report will assume that the size of the surplus is being targeted, unless otherwise noted. 2 See CRS Report RL30901, Projecting the Surplus: A Discussion of Issues, by Marc Labonte and Philip Winters. CRS-2 unemployment. A trigger policy that does not make a distinction between whether changes in the surplus are being driven by automatic budget stabilizers or policy changes is open to the criticism that it would be pro-cyclical. How large are these automatic stabilizers? The Congressional Budget Office (CBO) estimates that the recession will reduce the surplus by $148 billion in 2002 by lowering tax revenues and automatically increasing certain categories of outlays. The $148 billion reduction does not include the cost of any policy changes or technical re-estimates.3 A pro-cyclical policy is one that exaggerates the effects of the business cycle. For example, pro-cyclical monetary policy would raise interest rates in recessions, dampening aggregate demand, and lower interest rates in expansions, adding to aggregate demand. Tax cuts increase aggregate spending in the economy in the short run. This may help return the economy to full potential if the economy is in recession, but may increase inflationary pressures if the economy is already at full potential. Instituting tax cuts during a downturn would be a counter-cyclical policy.4 Some trigger policies could be considered pro-cyclical because they would prevent a tax reduction phase-in from occurring because poor economic growth caused the surplus to shrink below the target, even if spending and tax policy had been very restrained. Under these circumstances, it can be argued that the trigger would exacerbate the downturn by restraining aggregate demand. Alternatively, if the economy grew rapidly, the surplus could be above the target, regardless of spending and tax behavior. If the growth of the economy were rapid enough to be inflationary, allowing the phase-in would add to inflationary pressures. Can a trigger policy be designed that avoids this problem? One option would be to delegate to some body, for instance Congress or an executive department, possibly the Commerce Department, the responsibility of evaluating each year whether deteriorating economic conditions would make a trigger detrimental. If conditions were found to be deteriorating, the phased-in reduction would be implemented regardless of the whether the target had been met. This type of proposal could be criticized on the grounds that it is based on a subjective decision, and thus could be prey to the sort of political pressures that critics fear would undermine a trigger (see below). Is there a way to base the decision on a more objective standard? In principle, it should be fairly easy. CBO calculates a "structural surplus," or "full employment surplus," as well as the actual surplus. The structural surplus is based on the notion that business cycle effects can be stripped out of tax revenues and outlays to reveal how large the surplus would be if the economy, regardless of its current state, were performing at its full potential. From this, CBO could extrapolate what other figures, like the publicly held national debt and net interest payments, would be if the surplus equaled the "structural surplus." The trigger might be applied to the structural surplus just in situations when the economy is below its full potential, so that phased- in tax cuts are not forgone because of recession. Or it might also be applied to 3 CBO, Budget and Economic Outlook, January 2002. 4 For more information, see CRS Report RL30839, Income Tax Cuts, the Business Cycle, and Economic Growth: A Macroeconomic Analysis, by Marc Labonte and Gail Makinen,. CRS-3 situations where the economy is above full employment, in which case the actual surplus may be above the target, but taxes would still not be cut because the surplus would have been larger if other tax and spending policies had been as restrained as intended.5 Of course, the structural surplus is based on CBO's assumptions concerning the sustainable rates of economic growth and unemployment, which is ultimately a subjective decision. Many economists believe that monetary policy is a better tool than fiscal policy for managing aggregate demand in the short run. If this is the case, they may argue, implementing or preventing a tax cut on short run grounds is misguided, because monetary policy can lead the economy to full employment regardless of the fiscal stance. This argument could be applied in favor of a simpler trigger as well: even if the trigger were pro-cyclical, it can be argued that monetary policy will adjust to neutralize the trigger's effect on aggregate demand anyway, making the argument a moot point. How Long Can the Trigger Effectively Last? There is an asymmetry in timing between the phase-ins of EGTTRA and the projected surpluses. 84% of the surpluses are forecast to occur from 2007-2011, and all of the on-budget surpluses occur from 2010-2011, as seen in Table 1. (These forecasts do not contain the revenue costs of any economic stimulus proposal, or any other proposal that was not signed into law by January 2002.) By contrast, most of the principal elements of EGTRRA, in terms of revenue effects, will be completely phased in by 2006. This includes the reductions in marginal tax rates and the reduction in the marriage penalty. Only the reduction in estate tax rates occurs primarily after 2006, and the revenue effects of this provision will be largely outside the 10 year budget window. Thus, it can be argued that the major uncertainty in forecasts occurs after the trigger could effectively safeguard against uncertainty. In other words, if the 2007-2011 surpluses fail to materialize, the trigger will do little to offset the revenue loss. 5 Under the law, as was determined in the 1986 Supreme Court case Bowsher v. Synar, automatic "triggered" changes in law must be based on estimates by the executive branch, in this case through the Office for Management and Budget (OMB). This case declared the 1985 Balanced Budget and Emergency Deficit Control Act, or the original Gramm-Rudman- Hollings Act, in part unconstitutional. While OMB does not calculate a structural surplus at present, it could do so in the future by using the same methods that CBO uses. CRS-4 Table 1: Comparing the Baseline Projection with the Administration Tax Cut (Billions of Dollars) 2002- 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2011 CBO Baseline -21 -14 54 103 128 166 202 250 294 439 1,602 Surplus CBO Baseline On- -181 -193 -141 -108 -99 -76 -56 -23 4 131 3,122 Budget Surplus EGTRRA -38 -91 -108 -107 -135 -152 -160 -168 -187 -130 -1,275 Source: CBO, Budget and Economic Outlook, January 2002. Note: Revenue estimate of EGTRRA is based on the law expiring on December 31, 2010. Revenue estimate of EGTRRA does not include larger debt service costs that result from the tax cut. Might a Trigger Policy Fail for Political Reasons? There are at least two arguments against triggers based on political, rather than economic grounds. Some opponents of a trigger policy ­ and proponents of tax cuts ­ argue that a trigger policy would have a perverse effect on public policy. They argue that opponents of tax cuts might intentionally boost outlays during the appropriations process enough that the surplus would miss its target and the phase-in tax cuts would be prevented. In other words, they believe that a trigger may breed more fiscal profligacy, rather than less.6 Some trigger proposals include targets on the growth of discretionary spending as well to counter this possibility. Other opponents argue that trigger policies would give a false sense of security, but would ultimately be undermined by those who would use parliamentary tactics to skirt the trigger. They point to the failure of the Gramm-Rudman-Hollings Act to control the deficit as intended through triggered cuts in outlays in the late 1980s and early 1990s. They argue that tough trigger targets would validate "cheating" to contravene the trigger and lenient trigger targets would offer no meaningful fiscal discipline. While Congress cannot simply ignore a trigger once it is law, there are many ways it could be circumvented, depending on how the trigger is written. For example, it might be circumvented by re-passing the tax cut or passing supplemental appropriations after the trigger is applied. And future Congresses always retain the right to pass a bill that would revoke the trigger outright. Proponents of triggers argue that even if the Gramm-Rudman-Hollings Act did not work as well as intended, it did create some fiscal restraint and the budget deficits would have been even larger in its absence. As evidence, they point to the fact that discretionary spending fell from 10% of gross domestic product (GDP) in 1986 to 8.7% of GDP in 1990. Some opponents of a trigger policy have argued that the surplus is better safeguarded if the parts of the tax cut that have not yet been phase-ins are revoked now. They argue that under current forecasts, any trigger targets are highly unlikely to be met anyway, particularly if the target is on-budget balance. Thus, if future provisions of the tax cut are already predicted to be unaffordable, lawmakers should make certain they do not become law today, rather than take the chance that a trigger 6 For example, see Stephen Entin, "The Trouble With Triggers," Congressional Advisory #110, Institute for Research on the Economics of Taxation, February 13, 2001. CRS-5 will be subverted in the future. If the surplus proved larger than expected, then the revoked provisions could be reinstated. Uncertainty in Budget Projections Since the trigger is likely to target today's 10-year budget projection, one should consider the uncertainty of such projections.7 For example, CBO projected in 1997 that the FY2000 budget would be in deficit by over $100 billion, even after adjusting for policy changes since 1997. In fact, the FY2000 budget was in surplus by $236 billion. Even for the following fiscal year, budget projections are fairly inaccurate; as the projection is extrapolated further into the future, it has historically been less and less accurate. From 1981-1999, the average absolute value of the error in CBO's estimated surplus for the following budget year was 1.1% of GDP, after adjusting for policy changes. This percentage of GDP equaled $109 billion in 2000. The average error for 5 years in the future was 3.1% of GDP. Most of the inaccuracy came from economic assumptions that turned out to be incorrect. Nor does the accuracy of forecasts seem to be improving: policy changes explain only 43% of the $4 trillion reduction in CBO's 10-year baseline surplus since January 2001, the other $1.6 trillion reduction in the baseline surplus is due to forecast re-estimation. Thus, forecasts tend to become more inaccurate as one extrapolates further into the future, which is troublesome considering that 84% of the on-budget surpluses are forecast to occur from 2007-2011. This suggests that it would be desirable for a trigger to leave a considerable amount of room for forecast error in outlier years. But as the next section indicates, this uncertainty cannot easily be built into a trigger policy. A trigger could gain considerable accuracy if targets for outlier years were re- estimated as they are approached, but some would argue that this would defeat the purpose of a trigger ­ to provide objective fiscal discipline in advance. If the trigger were re-estimated as forecasts are updated, some believe that future manipulation of the updated targets would allow unaffordable phase-ins to occur (or, conversely, prevent affordable phase-ins from occurring). Should a Trigger Target the Past, Present, or Future? A trigger must target some budget outcome. But does it make more sense to target past, present, or future outcomes? In other words, should the trigger be applied to the previous fiscal year, after it has been observed? Should it apply to the current fiscal year's budget, after it has just been passed? Or should it apply to the current fiscal year's budget blueprint, at the beginning of the budget process? Each choice has advantages and disadvantages. Applying the trigger to the future would involve targeting the surplus generated by the baseline projection for the coming year at the beginning of the budget process. For example, in January 2004, the target for FY2005 phase-ins would be applied to the FY2005 baseline. Notice that this is the only trigger target that is influenced solely by changes in the economic and technical conditions underlying the baseline, 7 For more information, see CRS Report RL30854, Uncertainty in Budget Projections, by Phillip Winters. CRS-6 whereas the other two targets are applied to a surplus influenced by both economic changes and that year's policy changes, and cannot easily distinguish between the two. It can be argued that this type of target would place the trigger in the overall budget context. This would make the tradeoffs between spending, tax cuts, and debt reduction over the next 10 years most overt for the policymaker designing a trigger today. The drawback, critics would argue, is that it would allow the target to be met at the beginning of the budget process, allowing the phase-ins to be enacted, before compromises and amendments had been made in the budget that might contravene the target. It would also judge the target on budget forecasts that have historically proven to be inaccurate. Furthermore, the temptation might exist to present an overly optimistic baseline simply to meet the target.8 Targeting the "present" would involve applying the target to the budget balance after that year's budget had been passed (but before it had been implemented). For example, the target for FY2005 phase-ins would be applied directly after the FY2005 budget was passed in September 2004. One advantage of this type of target is the fact that the budget forecasts have historically become much more accurate by the end of the budget resolution. But applying a trigger to the final budget resolution could produce avoidance measures of a different type. A budget could be passed that met the target, and then supplemental appropriations and emergency spending could be undertaken after the phase-in was approved. It appears that many methods were employed to avoid triggering the spending cuts mandated by the Gramm-Rudman- Hollings Act when the deficits were increasing in the late 1980s and early 1990s. Such activities, it can be argued, would be equally hard to prevent with a tax cut trigger. Targeting the "past" would involve targeting the past year's actual result to determine whether the next year's phase-ins would occur. For example, in October 2004, the actual surplus for FY2004 would determine whether or not the FY2005 phase-in would occur. Applying the trigger to the past fiscal year's results may be a less intellectually coherent option, but the more practical way to avoid some of the pitfalls outlined above. Some may argue that it does not seem particularly logical to punish last year's mistakes or excesses in this year's budget, especially if this year's budget does not repeat those mistakes.9 On the other hand, proponents of this approach might argue that "hindsight is 20-20," and the only way to ensure that a target is not contravened by purposely rosy forecasts or parliamentary maneuvers is to give policymakers the chance to see how events played out in fact before applying the trigger. It would also benefit from being applied to an actual result which is not, therefore, subject to forecast uncertainty. 8 Tying a trigger to a structural surplus exacerbates this problem. The size of the full employment surplus depends on what rate of economic growth is deemed consistent with full employment. Thus, a target could be met by assuming whatever rate of economic growth generates a full employment surplus equal to the target. However, there is little consensus regarding what the "full employment" rate of growth is. 9 If the primary purpose of the trigger is to pay down the debt, then a trigger applied to the past fiscal year may be seen as an effective response: if the debt is not paid down this year, it accompanies us to the next year. CRS-7 Targeting the "past" is the only option whose rules do not prevent a deficit from purposely occurring (assuming cheating did not occur in the other two options). This is because the trigger is being applied to a different fiscal year than when the phase-ins occur. For example, imagine that the 2005 phase-ins are being determined in September 2004 by the 2004 actual on-budget surplus. Thus, nothing prevents an on- budget deficit from occurring in 2004; the only result would be that the 2005 phase-in would not occur. Alternatively, the target could be met in 2004, allowing the 2005 phase-in to occur. But this target would say nothing about the size of the 2005 surplus; thus, it would not prevent a 2005 phase-in from pushing the 2005 budget into deficit. Game theory analysis leads to the conclusion that the decision to have the trigger apply to the past, present, or future becomes extremely important when considering the uncertainty of budget projections.10 If the surpluses exceed the baseline projection, the target is non-binding and irrelevant. Likewise, if the surpluses disappear or turn out to be so small that the tax cut is deemed unaffordable, then any type of target will prevent the phase-ins from proceeding. But imagine that the surplus projections turn out to be smaller than expected but still large enough for a tax cut. For example, the tax cut is estimated to cost about $2 trillion when extra interest costs are included, and the on-budget surplus is forecast to equal about $3 trillion over 10 years. Assume that policymakers would like to spend $2 trillion on a tax cut and $1 trillion on other policy priorities which have not yet been enacted. If the surplus turns out to total only $2 trillion, either the tax cut or the other policy initiatives could be enacted, but not both, if the budget is to remain in surplus. Any trigger that is based on actual results would not build in these types of distinctions to account for forecast uncertainty. Hence, how the trigger is designed would determine whether the tax cut or the other policy priorities would be given priority in this case. By its nature, the trigger that targets the "past" result will automatically ensure that the tax cut would take priority over other policy initiatives in the coming fiscal year. By targeting the last year's results, the phase-in, in effect, becomes the first policy decision of the new budget. This type of trigger, based on an outcome that occurs after policy changes, cannot easily differentiate between changes in the surplus due to changes in economic conditions and changes due to policy. Thus, the trigger cannot be designed to accommodate those policymakers who would choose to give less priority to the tax cut than other policy initiatives if the surplus shrank for economic reasons. Perversely, with this type of trigger, once Congress discovers that next year's baseline surplus has been projected to fall for economic reasons, next year's phase-in could be prevented from occurring only by purposely increasing spending enough this year to miss the target. A trigger applied to the present would have a very different effect. This type of trigger, occurring at the end of the budget process, would also be unable to distinguish between changes in the surplus being driven by economic changes rather than policy changes. But unlike the trigger applied to the past, this trigger would be the last policy decision of the budget. Thus, the designer of a trigger policy cannot design the target in a way that gives priority in advance to tax cuts rather than other 10 The following analysis assumes that targets would not be updated as forecasts are updated. CRS-8 policy initiatives, or vice versa, in the light of forecast uncertainty. Instead, the ability to choose priorities would fall to the Congress passing the budget in question. Once Congress learns that the baseline surplus has shrunk for economic reasons, it would have to decide whether to set aside the remaining surplus for tax cuts, by meeting the target, or other policy initiatives, by missing the target. In many respects, a trigger that targets the present is equivalent to passing each year's phase-ins as a separate law each year. A trigger applied to the "future" is the only type of trigger that can be designed to distinguish between economic changes and policy changes that affect the surplus. That is because it is applied to the baseline surplus, before policy changes occur. Thus, it follows that only a trigger applied to the future can be designed to favor either tax cuts or other policy initiatives in the event that surplus falls for economic reasons. A policymaker who wished to give priority to the tax cut would target a baseline surplus equal in size to only the cost of the tax cut. A policymaker who wished to give priority to other policy initiatives would target a baseline surplus equal to the entire baseline surplus. It also follows from this analysis that once another policy initiative becomes law, it generally would have priority over the tax cut phase-ins regardless of what type of trigger were used ­ at that point, it, unlike the phase-ins, would not be automatically revoked if the surplus forecast falls. Would a Trigger Create Uncertainty in Decision Making? Those who believe that people's incentives are highly influenced by marginal tax rates may oppose a trigger mechanism on the grounds that it would create uncertainty in individuals' planning and decision making regarding work, saving, and consumption. In their opinion, this would defeat one of the primary goals of tax reduction ­ increasing economic growth through greater incentives to work and save. For example, individuals might be hampered in their investment planning because they would be unable to predict whether or not a tax cut phase in will occur. This argument seems more persuasive in the case of corporate taxes since businesses do formally model their capital outlay plans several years in advance. But opponents find this argument unpersuasive when applied to individuals because they believe that most individuals' future plans are vague and subject to change. What Number Should Be Targeted? When deciding how to determine what number to target for a trigger, there are a few points to keep in mind: ! Budget projections are nominal figures, based on an assumed inflation rate. For most numbers, it may be more consistent to adjust the target for inflation. Otherwise, the target may be missed in nominal terms, for better or for worse, because the actual inflation rate was different than forecast, even though the target was met when measured by real purchasing power. Revenues tend to CRS-9 grow more rapidly than outlays when inflation rises.11 An exception to this would be a target for the national debt if the trigger's goal were to effectively retire the debt. Ignoring the effect of inflation on the surpluses, the national debt will be the same in nominal terms regardless of how high or low inflation proves to be in the future. ! The budget surpluses, especially the on-budget portion, are projected to grow considerably each year of the 10-year forecast. If a trigger were meant to prevent "fiscal profligacy," a target that stays at a constant surplus (or interest payment) level could allow for more "profligacy" each year. In this case, targeting a percentage of the forecast might be considered. ! If the target were based on the baseline estimate, the target would need to take into consideration the fact that some baseline assumptions are legalistic, rather than realistic.12 The target would more likely be met if baseline assumptions were modified when appropriate. ! If the trigger were applied to the surplus on a certain date each year, the targeted amount would need to account for the fact that tax revenues are highly seasonal because many taxes are paid quarterly or yearly. This makes the publicly held debt highly seasonal as well. Likewise, day-to-day funding requirements of the Treasury lead to an imprecise relationship between the size of the surplus and the reduction in the publicly held debt. Since the seasonality is predictable in both cases, estimating the debt at any given time of the year is fairly simple, but important to acknowledge. It would be inappropriate for a trigger based on, say, April's data to be applied to May's data. Triggers that were based on a comparison to the same point in the previous fiscal year would avoid this problem. ! Presumably, any phase-in would occur (or not occur) in January, at the beginning of the new tax year. The mismatch between the budget fiscal year, which begins in October, and the tax calendar year, which begins in January, may be corrected by an adjustment in the target used for the trigger. Since a phased-in rate reduction would occur in January, the revenues of each fiscal year where a phase in had occurred would be generated from one quarter of revenues at the last tax year's higher rate and three quarters of revenue at the next tax year's lower rate. This gives an inflated impression of what revenues are being generated by current tax policy. This problem could be avoided by targeting the surplus generated during only the last three quarters of the fiscal year. ! A debt target, rather than a surplus target, may not prevent the on-budget portion of the budget from returning to deficit. In accounting terms, this is because the debt is a cumulative "stock," determined by the sum of past 11 CBO, Budget and Economic Outlook, January 2001, page 115. 12 CBO elaborates on this concept on p. 7 of The Budget and Economic Outlook, January 2001. For a discussion of alternative baseline assumptions, see CRS Report RL30901, Projecting the Surplus: A Discussion of Issues, by Marc Labonte and Philip Winters. CRS-10 deficits, while the surplus is a yearly "flow," determined by each year's budget balance. For example, imagine that the surplus were larger than expected in 2002 so that debt reduction greatly exceeded the debt target. If the reduction that occurred in 2002 were large enough, it would imply that the debt target could be met in 2003 even with an on-budget deficit. ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL30948