tax policy center
Tax Policy Center
   Entry 6 of 9  

The Bush Tax Cuts: Are tax cuts an effective way to reduce government spending?

The "starve the beast" theory relies on the intuition that if government has less revenue, it will be pressured to reduce spending. We offer an alternative theory that we call "coordinated fiscal discipline," which holds that policymaking institutions go through alternating periods of fiscal largesse and fiscal restraint. Periods of largesse see both tax cuts and spending increases, whereas periods of restraint involve both tax increases and spending cuts. One mechanism by which this might happen is that granting large tax cuts to some groups makes it less politically feasible to rein in the desires of other constituencies for increases in their preferred spending programs. Several suggestive pieces of evidence favor the "coordinated fiscal discipline" view over the "starve the beast" view.

  • Budget rules and legislative agreements have proved effective at reducing spending and balancing the budget in the past only when restrictions were placed on both tax cuts and spending increases at the same time. For example, a series of budget rules were imposed first in 1990 and then extended in 1993 and 1997 in an attempt to reduce budget deficits. These rules imposed restraints on both sides of the budget. Tax cuts and increases in mandatory spending (that is, spending obligated by virtue of previously enacted laws, often on entitlement programs such as Medicare or Social Security, rather than enacted by laws in the current year) had to be paid for with tax increases or cuts in other mandatory spending. Discretionary spending (that is, spending set by annual appropriation) was made subject to caps. Likewise, the budget deals that were enacted to reduce deficits in 1990 and 1993 involved both spending cuts and revenue increases. This series of policies helped contribute to the budget surpluses that emerged starting in 1998.
  • Many in Congress who voted for the Bush tax cuts have also voted for substantial increases in government spending. One study examined the voting records of members who had signed a "no new taxes" pledge. The signers voted overwhelmingly in favor of the Bush administration’s tax cuts and have committed themselves not to vote for future tax increases. Yet 86 percent of signers also voted for the extremely expensive Medicare Modernization Act of 2003, which extended Medicare to cover prescription drugs, despite estimates that it would increase federal spending by $395 billion over 2004 to 2013 and substantially more thereafter. Almost three-quarters of those who had taken the no-new-taxes pledge supported the pork-laden 2004 highway bill, which would have spent about $300 billion over six years had it passed. Thus many of the same people who voted for tax cuts also voted for large permanent spending increases, at a time when budget deficits were large and growing-hardly behavior consistent with a strategy of "starving the beast."
  • The "starve the beast" theory suggests that lower revenue should be accompanied by lower spending, and higher revenue by higher spending. Conversely, the coordinated fiscal discipline view implies that lower revenue will accompany higher spending, and higher revenue will accompany lower spending. The data over the past twenty-five years generally display a pattern consistent with the latter view.
  • Testing these alternative hypotheses empirically requires that the spending and revenue data must be "standardized" to remove the effects of the business cycle on the federal budget. The reduced economic activity in a recession tends to lower tax revenue automatically, without any policy intervention, while spending on welfare, unemployment insurance, and related programs rises automatically as well. An economic upswing has the opposite effects. Thus the movements of the business cycle will make taxes and spending move in ways that appear to favor the coordinated fiscal discipline hypothesis. Taking out these spurious effects is thus necessary for a fair test of whether taxes and spending tend to rise and fall together or to move in opposite directions.
  • The results of such an analysis show that even after controlling for the business cycle, changes in spending and changes in taxes moved opposite to each other, rather than with each other, over three recent major periods. In the first, between 1981 and 1992, revenue fell and total outlays rose. In the second, between 1992 and 2000, revenue rose and spending fell. Finally, between 2000 and 2004, revenue fell relative to GDP but spending rose. (This period included the September 11, 2001, terrorist attacks and the beginnings of the wars in Afghanistan and Iraq, but only about half of the increase in noninterest spending was due to increased defense and homeland security appropriations.) All of these patterns are inconsistent with the "starve the beast" view.
  • It is true that between 1981 and 1992, as revenue fell, standardized noninterest spending fell, too. But this pattern can hardly be taken as strong evidence of effective fiscal discipline. Standardized noninterest spending fell by only 0.4 percent of GDP, while the ratio of public debt to GDP almost doubled, from 26 percent in 1981 to 48 percent in 1992, the largest peacetime growth in the debt-to-GDP ratio in U.S. history aside from the Great Depression. Thus lower revenue has proved to be neither necessary (witness the 1990s) nor sufficient (witness the 1980s and the period since 2000) to reduce federal spending.
  • The past several years underscore the questions about the starve-the-beast view. It is hard to believe that spending would have increased by much more than it did between 2000 and 2004 if the tax cuts had not been enacted. Discretionary spending rose from 6.3 percent of GDP in 2000 to 7.8 percent in 2004 (defense spending accounts for 60 percent of that increase). Mandatory spending meanwhile rose from 9.8 percent of GDP in 2000 to 10.7 percent of GDP, and future growth in such spending was ensured by the passage of a large new entitlement program, the Medicare prescription drug benefit.
  • The formal econometric evidence on whether tax reductions are followed by spending reductions is mixed: the available studies were mostly done back in the 1980s, and there are results supporting each side. The more recent evidence does suggest that larger budget deficits constrain both spending increases and tax reductions. This evidence, however, does not distinguish between the "starve the beast" and the "coordinated fiscal discipline" theories. The pattern of larger budget deficits leading to higher taxes and lower spending is based on historical experience in which both spending reductions and tax increases were considered jointly as part of fiscal restraint packages. This pattern does not prove (nor does it disprove) that revenue reductions will induce spending reductions. In short, there is simply no compelling evidence that tax cuts constrain spending.
   Entry 6 of 9