February 25, 2000
Should the Budget Baseline Adjust Discretionary Spending Levels for Inflation?
An Examination of the Criticisms of this Approach
by James Horney and Robert Greenstein
One of the first debates over the budget for fiscal year 2001 is not about tax cuts or providing prescription drug benefits for Medicare beneficiaries it is about which baseline to use as the starting point in Congressional budget deliberations. In particular, it is about what levels of discretionary (i.e., non-entitlement) spending to assume for coming years in the budget baseline.
The decision about the level of discretionary spending to assume in the baseline is not a policy decision by Congress about the appropriate level of discretionary spending to provide. By definition, a baseline is simply a projection of what the level of spending will be if the President and Congress make no changes in current policies. It provides a standard to use in measuring the fiscal effects of various policy proposals; it is not a policy proposal itself.
The baseline in the Administration's budget assumes that under current policies, total discretionary appropriations over the next 10 years will equal the fiscal year 2000 appropriations level, adjusted for inflation. Some Congressional leaders are contending it is inappropriate to use this assumption in the budget resolution baseline because this assumption would:
- Produce a baseline with projected levels of discretionary outlays that are unrealistically high;
- Weaken fiscal responsibility; and
- Be tantamount to placing the entire discretionary part of the budget on autopilot.
This analysis examines the arguments raised against a baseline assumption that discretionary spending will stay even with inflation. The analysis finds that none of these arguments bear up well under scrutiny.
First, a majority of Congress and the President seem to have agreed that defense spending, which constitutes about half of total discretionary spending, should at least keep pace with inflation. Thus, for total discretionary spending to be held below the current discretionary spending level, adjusted for inflation, non-defense discretionary spending would have to be cut in real (i.e., inflation-adjusted) terms. Yet Republicans as well as Democrats want to increase non-defense discretionary spending for education, health research, and various other areas.
Moreover, over the past decade, non-defense discretionary spending has increased 20 percent in real terms. If spending for non-defense discretionary programs grew 20 percent during a period when Congress was struggling to eliminate deficits, it is highly implausible it will be cut in real terms during a period of budget surpluses. Budget experts such as Robert Reischauer, the former director of the Congressional Budget Office, have said that an assumption that overall discretionary spending will simply rise at the rate of inflation over the next 10 years is likely to prove unrealistically low, not unrealistically high.
Regarding fiscal responsibility, the Concord Coalition and the "Blue Dog" coalition of conservative House Democrats have both pointed out that using an assumption that discretionary spending will remain at current levels, adjusted for inflation, would promote rather than weaken fiscal responsibility, and that it is assumptions of lower levels of discretionary spending that present a threat to fiscal integrity. Assumptions that discretionary spending will be held below current levels, adjusted for inflation, are unrealistic and would lead to inflated surplus projections. Those inflated surplus projections, in turn, would open the door for larger tax cuts and/or entitlement expansions than would be fiscally prudent.
Finally, the attack line that assuming discretionary spending will stay even with inflation is tantamount to putting the government on autopilot makes for a catchy soundbite. But this charge is an empty phrase largely devoid of content. Any baseline assumption could be described as "putting the budget on autopilot" since baseline projections, by definition, are supposed to reflect the levels of spending, revenues, and surpluses that will occur if current policies remain unchanged.
Furthermore, the "autopilot" charge is tossed around in a fashion that implies that using a baseline that equals current discretionary funding levels, adjusted for inflation, reflects a policy decision to keep every single program and account in the budget at its current level in inflation-adjusted terms. Such a charge is nonsense. Such a baseline assumption is simply an assumption regarding the total level of discretionary appropriations. Within such an overall level, some programs could be increased while others were decreased or eliminated. The simplistic "autopilot" charge could just as easily (and just as wrongly) be used to describe a baseline that freezes discretionary funding at current levels, with no adjustment for inflation; it could be said this implies that every program will continue to get the exact same level of funding as last year even if some programs don't need the same amount of funds.
What is the Role of a Baseline?
A baseline is not intended to be a prediction of what budget outcomes will be; that depends in part on changes in policies Congress and the President choose to make. To the contrary, a baseline consists of projections of the levels of federal spending, revenues, and surpluses that would occur if budget policies remain unchanged. A baseline serves as "a benchmark for measuring the budgetary effects of proposed changes in federal revenues or spending."(1)
Since revenues and spending for entitlement programs are generally governed by permanent laws, projections of revenues and entitlement spending under current budget policies are generally based on the assumption that these laws will remain unchanged.(2) In the case of non-entitlement, or discretionary, spending which is determined each year in appropriation bills a strict assumption of no change in current laws would imply there would be no new appropriation laws and hence that all discretionary programs would run out of funds and cease to operate. That, of course, would not be a reasonable description of current policy.(3) Thus, baseline projections must make some assumption about the level of future appropriations.
Prior to the enactment of statutory caps on discretionary spending in 1990, the basic baseline projections produced by CBO and the Administration were based on the assumption that discretionary appropriations would be maintained at current-year levels, adjusted for inflation. That assumption was and still is specified in the baseline rules of the Balanced Budget and Emergency Deficit Control Act.(4)
After discretionary caps were enacted in 1990, the Congressional Budget Office and the Administration continued to produce projections of the levels of discretionary spending that would occur if appropriations were maintained at current levels, adjusted for inflation. For purposes of their basic projections of deficits or surpluses, however, they assumed that unspecified cuts below those baseline levels would reduce discretionary spending to the levels of the statutory caps in the years for which caps had been enacted and that discretionary spending would grow at the rate of inflation after the caps expired. From 1991 through 1998, actual appropriations did essentially comply with the caps, so assuming that appropriations would continue to do so was a reasonable description of current policy. But since discretionary spending enacted for the last two years greatly exceeded the caps, and complying with the caps in 2001 and 2002 would require draconian cuts in real expenditure levels for discretionary programs, virtually no observer believes that current policy is to hold discretionary spending to the level of the caps.(5)
Consequently, both the Administration and CBO have produced baselines that assume that current policy for discretionary spending means that appropriations will be provided each year from 2001 through 2010 at the level enacted for the current year, fiscal year 2000.(6) CBO produced two sets of projected discretionary spending levels based on the FY 2000 appropriations level one set in which appropriations are provided at the 2000 level for the next 10 years without any adjustment for inflation, and one set in which appropriations are maintained at the 2000 level adjusted for inflation. The Administration produced a single baseline, which assumes that discretionary funding will be maintained at the current level, adjusted for inflation.
The assumption that is made about discretionary spending levels has a large effect on the level of surpluses projected. CBO projects that non-Social Security surpluses will total $838 billion over the next 10 years if discretionary spending is maintained at current levels, adjusted for inflation. (The Administration projects non-Social Security surpluses totaling $746 billion over the same period, an amount close to CBO's considering the size of the budget.) But under the alternative assumption that discretionary spending will be frozen at the current levels for 10 years without any adjustment for inflation, CBO projects that non-Social Security surpluses will total $1.85 trillion over the next 10 years. Because such a large difference in projected surpluses could and likely would affect decisions about the affordability of large tax cuts or increases in entitlement spending, the choice of assumptions about discretionary spending has become a political issue.
Criticisms of and Facts About the Inflated Baseline
Some Members of Congress have argued that the baseline used in drafting this year's Congressional budget resolution should not assume that discretionary spending will be provided over the next 10 years through 2010 at the level enacted for 2000, adjusted for inflation. These Members favor alternative baseline assumptions that produce lower baseline levels of discretionary spending and thereby make it appear as though there is room for a larger tax cut. The three arguments that these Members most commonly make against using the assumption that discretionary spending in coming years will equal current levels, adjusted for inflation, are examined below.
Argument #1: The Inflated Baseline Assumes the Budget is Put on Autopilot.
Some Members have argued that the budget resolution baseline should not assume that discretionary spending will be maintained at current levels, adjusted for inflation, because that would imply the budget would be "put on autopilot" for 10 years and that spending would be continued for programs where such spending is no longer necessary.
The Members who make this argument have directed it against the baseline that assumes discretionary spending will grow at the rate of inflation. But the same argument could be made against a baseline that assumes discretionary appropriations will be frozen at current levels without any adjustment for inflation, or against any baseline that is based on current appropriations levels without changes in policies.
This argument seems to indicate confusion about the purpose of a baseline projection. A baseline is supposed to represent a benchmark for considering changes in policies. A baseline should indicate what the levels of spending will be if no changes in policies are made, not the levels of spending that will occur if the President and Congress alter policies. Having a baseline that reflects increases or decreases in spending that would occur if the President and Congress decide total appropriations should be increased or reduced is putting the cart before the horse.
In addition, this argument misconstrues the significance of a baseline projection that assumes that discretionary appropriations will be maintained at current levels, adjusted for inflation. Such a projection simply shows how much it will cost to continue providing the same aggregate level of goods and services as is currently being provided through discretionary appropriations. Using such a baseline projection for discretionary spending simply entails using this aggregate level as a benchmark. It does not imply that each and every account should be maintained at its current spending level, adjusted for inflation.
Argument #2: The Inflated Baseline Does Not Promote Fiscal Responsibility.
Some Members have argued that the budget committees should not use a baseline that assumes current levels of discretionary spending, adjusted for inflation, because that would weaken fiscal discipline. In fact, the opposite is very likely to be the case. The experience of the past two years demonstrates that using an artificially low baseline projection for discretionary spending erodes rather than strengthens fiscal discipline and compromises fiscal integrity.
Over the last two years, Congress and the President used a variety of stratagems designating regular appropriations as emergency spending, shifting military and civilian employee pay dates, and directing CBO to using unrealistically low estimates of outlays that CBO does not believe to be accurate to evade the discretionary spending caps. According to CBO, discretionary outlays will exceed the caps by $25 billion in 2000.(7) Budget discipline broke down because the caps were set at unrealistically low levels in the Balanced Budget Act of 1997. Once it became clear the caps would be exceeded, there was no real lid on discretionary spending.
Many budget analysts believe that setting realistic discretionary caps is the best method for ensuring discipline in the appropriations process. Relying on a baseline that assumes unrealistically low levels of discretionary spending is likely to encourage lawmakers once again to set caps on discretionary spending in future years that are unrealistically low and that will be disregarded when those years arrive. Instead of promoting restraint, baselines that assume unrealistically low levels of discretionary spending are likely to make it harder to control such spending in future years.
As we have seen with the caps established for 1999 and 2000, it is easy to assume low levels of aggregate discretionary spending in future years, but much harder to achieve those low levels when it comes time to enact actual appropriation bills. If discretionary spending is to be cut in future years below the current level, adjusted for inflation, it will take a concerted effort by policymakers who favor making such cuts to persuade other lawmakers that such cuts are justified during an area of budget surpluses. Yet those who say the baseline should assume that discretionary spending will be reduced below current levels, adjusted for inflation, generally pretend that policymakers will not need to change policies or reduce services to produce such reductions in real discretionary spending. (Anyone who doubts that cuts in spending below current, inflation-adjusted, levels entail real reductions in program activities should look at the reductions in military personnel and weapons that have occurred over the last 13 years; see box on page 8.) Those advocating low baseline assumptions for discretionary spending generally have been unwilling to stand up and explicitly call for reducing the services these programs provide. Their unwillingness to do so hardly inspires confidence that reductions in the services that discretionary programs provide actually will be achieved.
In addition to making it harder to control discretionary spending in the future, basing this year's budget debate on baseline projections that assume unrealistically low levels of discretionary spending over coming years may lead Congress to pass tax and entitlement changes that are fiscally irresponsible. Some Members of Congress already have pointed to CBO's projections that non-Social Security surpluses would total $1.85 trillion over the next 10 years if discretionary spending is frozen at the 2000 level, without any adjustment for inflation, as evidence that big tax cuts are affordable. Yet that baseline assumption assumes that by 2010, total discretionary spending will be cut by more than 20 percent in real terms (i.e., below the 2000 level, adjusted for inflation). Moreover, if defense spending is allowed to keep pace with inflation, holding total spending to the freeze level would require a cut of almost 40 percent in non-defense discretionary spending by 2010. It is inconceivable that discretionary spending will be cut that much. Enacting big tax cuts that consume most or all of an inflated surplus projection that is based on the assumption that discretionary spending will be frozen indefinitely is likely to result in non-Social Security deficits over the next 10 years, unless the economy performs even better than CBO and the Administration have forecast.
Concern that assuming unrealistically low levels of discretionary spending and consequently, assuming unrealistically large surpluses could lead to imprudent policy decisions has led both the Concord Coalition and the Blue Dog Coalition of conservative and moderate Democrats in the House of Representatives to argue that lawmakers should base their budget decisions on a baseline that assumes discretionary spending will be maintained at current levels, adjusted for inflation.(8)
Argument #3: Assuming That Discretionary Spending Will Equal Current Levels, Adjusted for Inflation, Overstates the Likely Path of Discretionary Spending.
Some Members of Congress object to using a baseline that assumes discretionary spending will be maintained at current levels, adjusted for inflation, arguing that actual appropriations levels will be lower than that. As noted above, a baseline is not intended to reflect future policy decisions, so this criticism would not be a reason for adopting another baseline even if the assertion were true.
But the assertion that actual appropriations in coming years will be below today's level, adjusted for inflation, almost certainly is not true. Total discretionary spending has kept pace with inflation for the past two years despite caps that mandated deep cuts below the inflation-adjusted levels. Moreover, non-defense discretionary spending has grown faster than inflation in every year but one since the caps were instituted in 1991. Total discretionary spending did grow more slowly than inflation from 1991 through 1998, but only because defense spending was cut significantly in the years following the end of the Cold War. With a consensus apparently emerging among a majority in Congress and the President that defense spending should at least keep pace with inflation in coming years, it is exceedingly unlikely that total discretionary spending will grow more slowly than the rate of inflation in the years ahead.
In fact, the assumption that discretionary appropriations in 2001 through 2010 will equal the currently-enacted fiscal year 2000 appropriations levels, adjusted for inflation, is itself likely to prove conservative for several reasons. First, funding for fiscal year 2000 is almost certain to be increased by additional, supplemental appropriations enacted this spring or summer to fund assistance the United States will provide in connection with Middle East peace efforts, peacekeeping activities in Kosovo and East Timor, additional disaster relief to respond to disasters that will occur in coming months, and other pressing needs.
Second, there is significant pressure to increase defense discretionary spending above current levels in real terms. If that occurs, keeping total discretionary spending at the same level as in 2000, adjusted for inflation, would entail making real cuts in non-defense discretionary programs. That seems unlikely, especially in light of the fact that spending for non-defense discretionary programs grew about 20 percent in real terms over the last 10 years. If this spending grew 20 percent while the nation was struggling to eliminate budget deficits, it is unlikely it will be cut in real terms in an era of budget surpluses.
In the longer run, simply keeping overall discretionary spending even with inflation probably is not realistic in an increasingly wealthy nation that has a growing population and rising demands for government resources to be devoted to activities such as health research, infrastructure improvements, education, and defense weapons and readiness. Just as families spend more in real terms when their real income grows, the government is likely to spend more in real terms to provide the increased level of government services and investments that citizens are likely to expect and demand.
Can Spending Be Frozen Without Changing Policies?:
The Case of the Pentagon
CBO estimates that defense outlays in 2000 will total $283 billion, exactly the same amount as defense outlays at the height of the Reagan defense buildup in 1987. But despite the same level of nominal spending now as then (i.e., the same level of spending without adjustment for inflation), the number of active-duty military personnel has been reduced from 2.2 million to 1.4 million since 1987 (a 36 percent reduction), the number of Navy ships has been reduced from 569 to 315 (a 45 percent reduction), the number of Air Force fighter wings (active and reserve) has been reduced from 38 to 20 ( a 46 percent reduction), and the U.S. military has closed more than 900 facilities around the world and 97 major bases in the United States.*
Why have these reductions which virtually all policymakers would agree represent a significant change in defense policy occurred when the dollar level of appropriations is the same now as in 1987? The answer is that these reductions have occurred because real defense spending (spending after adjustment for inflation) has fallen nearly one-third over the last 13 years. In discretionary programs, maintaining current policies generally entails staying even with inflation.
* Sources: House Armed Services Committee, "Defense Accomplishments of the 104th, 105th, and 106th Congresses," (November 18, 1999); Department of Defense, Annual Report to the President and the Congress.
As former CBO director Robert Reischauer recently wrote, Congress let non-defense discretionary spending grow by 20 percent in real terms "during its decade-long jihad to balance the budget," and it is hard to imagine Congress will be more closefisted when the budget is in surplus, the economy is robust, and voters are demanding more and better government services.(9)
Another individual who has argued that realistic projections of spending under current policies should adjust for more than just inflation is Republican Presidential contender George W. Bush. The Wall Street Journal reported a charge by rival Steve Forbes that under the administration of Governor George W. Bush, Texas state spending has grown 36 percent. The Journal reported that Governor Bush has called the Forbes charge false, stating that Texas spending has increased only 2.7 percent.(10)
The Forbes campaign calculated the 36 percent figure by comparing Texas state spending in the current year with spending in 1995, when Governor Bush took office. By contrast, the Journal reported, "Mr. Bush calculates just 2.7 percent growth largely because he adjusts for inflation and population growth."
As the Journal explains, Governor Bush takes the level of state spending when he took office and adjusts it for both inflation and state population growth since that year. The resulting figure, after these adjustments are made, forms the baseline against which Governor Bush measures the spending that occurred to determine the extent to which spending increased. In a campaign debate in New Hampshire on January 26, Bush once again presented numbers on budget growth in Texas after adjustment for inflation and population growth and explicitly stated he was making these adjustments. As the Journal article notes, the Bush point is "Simply put, government has to spend more money just to provide the same services."(11)
The arguments are exceedingly weak for using a budget baseline that assumes that discretionary spending levels in coming years will be below the current level, adjusted for inflation. A baseline should reflect what the level of spending will be if there are no changes in policies. The best description of no change in current policies for discretionary spending is that the goods and services provided through discretionary appropriations are maintained at current levels, which requires that total discretionary funding be maintained at the current level, adjusted for inflation. Such an assumption does not imply that each and every program will continue to be funded at the current real level.
Assuming that discretionary spending will be lower than the current level, adjusted for inflation, is more likely to promote fiscal profligacy than fiscal discipline. Assuming unrealistically low levels of discretionary spending would make the enactment of oversized tax cuts and entitlement spending increases more likely. In addition, if unrealistically low baseline assumptions for discretionary spending lead to discretionary spending caps that are unrealistic, those caps are likely to be breached in future years, leaving less budget discipline than realistic discretionary spending limits would provide.
1. Congressional Budget Office, Glossary of The Budget and Economic Outlook: Fiscal Years 2001 through 2010 (January 2000).
2. That means that changes in tax rules or entitlement programs that are scheduled to occur under current law are taken into account. In addition, section 257 (b) of the Balanced Budget and Emergency Deficit Control Act provides a limited exception to the assumption that laws are not changed. It requires that baseline projections should assume that entitlement programs with outlays greater than $50 million that are scheduled to expire under current law, as well as excise taxes that are dedicated to a trust fund and are scheduled to expire, will continue.
3. CBO estimates that discretionary outlays will total $603 billion in 2000, representing about one-third of all government spending. Almost half $283 billion is for defense. The remaining $320 billion is for a great variety of activities, including health and science research, veterans medical care, environmental programs, highway construction, Head Start, the Federal Bureau of Investigation, and the National Park Service.
4. See section 257(c) of the Balanced Budget and Emergency Deficit Control Act.
5. For instance, John Kasich, the Chairman of the House Budget Committee, has been quoted as saying that "the caps are not a relevant limit today on spending." Bureau of National Affairs, Daily Tax Report, January 28, 2000, page G-11.
6. CBO also produced a baseline that shows what projected spending and surpluses would be if discretionary spending were held to the level of the caps in 2001 and 2002.
7. Excluding $8 billion in additional payments to farmers, which were funded in an appropriation bill but are counted as entitlement spending.
8. See, The Concord Coalition, "Concord Coalition Warns Against Surplus 'Expectations' Game" (January 24, 2000), and The Blue Dog Coalition, "The Blue Dogs Respond to CBO's Revised Budget Projections" (January 26, 2000).
9. Robert D. Reischauer, "The Phantom Surplus," New York Times, January 28, 2000, page A27.
10. Jackie Calmes, "On Spending, Forbes Takes Cue from Bush," Wall Street Journal, (November 5, 1999).
11. Bush stated in the January 26 debate: "In terms of the budget, I've slowed the rate of growth down. And when you take out population growth and inflation, it is by far the slowest rate of growth ever in my state's history." Similarly, the Bush web site declares that "the three budgets adopted by Governor Bush cut the rate of growth in real per capita spending to 2.7 percent..." Real per capita spending is spending adjusted for inflation and changes in population.