Stimulating the Economy Through Tax Policy: Principles and Applications

Congressional and Administration leaders are now examining the appropriate fiscal policy response to the recent terrorist attacks. Since the attack, the Federal Reserve has provided liquidity to financial markets, and has reduced its key lending rate twice, by a total of 100 basis points. Congress and the President agreed on a $40 billion spending package for defense, rescue and rebuilding efforts, and an airline bailout of $5 billion, plus additional loan guarantees.

The key short-term fiscal issues are whether additional stimulus proposals would help and what form assistance should take. Spending initiatives aimed at quickly stimulating the economy-including the rapid clean-up and rebuilding of New York City and efforts to attenuate the costs of economic slowdown-will likely form part of any stimulus package. Tax cut proposals, however, have now taken center stage in the debate.

This paper focuses on the stimulus potential of alternative tax cuts. We do not address whether a stimulus should be provided. Rather, we examine the principles that will lead to the most effective tax stimulus, should policy makers choose this path, and then examine some recent proposals in light of those principles.

The most effective stimulus proposal would maximize its "bang for the buck." That is, it would direct as much of the tax cuts as possible to stimulating new consumer spending and/or new business investment in the short run; and it would do the least possible damage to long-term fiscal prospects and hence minimize any upward pressure on interest rates. Temporary rebates to individuals and temporary business subsidies for new investment fit these principles. A package with substantial long-term revenue costs, on the other hand, could do as much harm as good: It would raise interest rates, which would offset much of the direct benefits of any stimulus by reducing business and housing investment, as well as consumption that is affected by interest rates or stock market values. Many of the stimulus proposals currently being considered do little or nothing to address the need to stimulate the economy in the short-run, and would exacerbate long-term fiscal problems. Proposals to cut tax rates on capital gains or on corporate income are particularly problematic along these dimensions. Such proposals may be worth discussing in other contexts, but they clearly represent the wrong policy response at the current time.

I. Principles

The appropriateness and effectiveness of a tax stimulus depends on the underlying economic outlook-to indicate what is needed-and the budget outlook-to indicate the financial constraints. The terrorist attack disrupted the workings of an already weakening economy, and many economists believe it pushed the economy into a recession. But the economy's long-term prospects remain strong. The 10-year budget outlook, which was relatively auspicious at the beginning of the year, has deteriorated rapidly due to the tax cut enacted this spring, the weakening of the economy before the terrorist attack, and the further weakening of the economy after the attack. The economic outlook thus suggests the need for policies that stimulate the economy in the short run. The budget outlook suggests that the long-run revenue impact of stimulus policies should be limited, so as to avoid exacerbating the nation's long-term fiscal challenges, which would raise interest rates and undermine the effectiveness of the stimulus. This reasoning suggests five principles for designing the most effective tax stimulus package. The first two relate to keeping the package focused on stimulating new behavior. The last three relate to maintaining budget discipline.

  • Structure any business tax incentives to encourage new investment, not to provide a windfall for previous investment. This includes temporary provisions to provide accelerated depreciation, expensing, or tax credits for new investments in the near future.

  • Design any household tax reductions to maximize effect on demand. Tax cuts should therefore be focused on low- and middle-income households who tend to have a higher propensity to spend out of their income than do high-income households. Timing any new tax cut for households to coincide with the holiday season may be an effective way to encourage it to be spent.

  • Allow only temporary items. This will limit the long-term cost of the package and reduce the temptation for policy-makers to try to push through long-standing proposals that may or may not have merit for other reasons, but do little or nothing to stimulate the economy in the short run.

  • Set an overall stimulus budget first. This will further reduce the likelihood of a run-away, "Christmas tree" spending and tax package. Fed Chairman Alan Greenspan and former Treasury Secretary Robert Rubin recently suggested a "budget" for the stimulus of $100 billion, including those items that have already been enacted since the attacks, and therefore new stimulus of roughly $50 billion. Subsequently, President Bush advocated new stimulus of $60 to $75 billion. (There appears to be some confusion as to the meaning of the overall stimulus ceiling. In our opinion, the overall stimulus budget should include all stimulus legislation proposed and enacted since September 11. It should cover all stimulus costs, as estimated by the Congressional Budget Office and Joint Committee on Taxation, over the next ten years.)

  • Minimize the long-term costs. Combining short-term stimulus with long-term fiscal discipline provides more stimulative impetus to the economy than a stimulus package alone, since it restrains any increase in interest rates that could undermine the effectiveness of the stimulus.

II. Tax cuts for households

Another rebate for consumers

Declines in consumer confidence, spending, and employment have led to calls for a second round of consumer rebates to directly bolster consumer demand. Economic research suggests that households tend to immediately spend between 20 percent and 70 percent of any temporary income tax cuts they receive. Although it is too soon to fully evaluate the impact of the rebate checks distributed this summer, the checks appear to be generating only a small increase in consumption. In August, personal after-tax income rose by 1.9 percent, in significant part due to the rebate, but personal spending only rose by 0.2 percent. A survey undertaken by the University of Michigan in August and September found that only 19 percent of respondents said that they were going to spend their rebates.

To make another round of rebates more effective as a stimulus, policy-makers could target the rebates to lower and middle income households. Economic research suggests that these households have tendencies to spend a greater proportion of any new income than high-income households do—and the more the rebate is spent, the more effective it is as a stimulus. Furthermore, roughly 30 million lower and moderate income households did not receive a rebate earlier this year, even though they pay federal payroll and excise taxes. To better target such households, the rebates should be based on employee Social Security and Medicare payroll taxes. (Although the rebate would be based on payroll taxes, it should be financed through general revenues.) To maximize the stimulative impact and minimize administrative hassles, the rebates could be set as one amount for anyone with earnings in calendar year 2000 over a minimum level.

Another way to stimulate the spending potential of any rebate may be to time the rebate checks for the holiday shopping season. Holiday purchases of presents for friends and family are highly seasonal. Intuitively, putting more resources in the hands of lower- and moderate-income families during the holiday season seems likely to bolster the volume of shopping during the holiday season. Vendors—who have already experienced one rebate round—may now be particularly astute about marketing sales and deals based on the rebates.

Accelerate the previously legislated tax cuts

Another proposal for household tax cuts is to accelerate to January 1, 2002 the income tax rate cuts—or some portion thereof—currently scheduled to occur on January 1, 2004 and January 1, 2006. This proposal is problematic for several reasons.

First, and most importantly for its stimulative impact, the acceleration is not well-targeted to generating additional spending in the short run. The vast majority of the costs of accelerating the rate reductions would occur after 2002: The acceleration would also reduce revenue in 2003, 2004, 2005, and 2006. Since most of the cost would not go to providing stimulus in 2002, when it is needed most, the proposal does not maximize its bang-for-the-buck. A temporary one-year tax cut would generate the same stimulus in 2002 at much less cost. (There is also a good chance that, when any accelerated tax cuts become operative in years between 2003 and 2006, the economy will already be growing rapidly. Thus, the proposal could very likely fall prey to the familiar pattern of the federal government's stimulating the economy at times when it does not need to be stimulated.)

Furthermore, the accelerated rate cuts would only apply to high-income households—those who are in the top 25 percent of the income distribution. Most taxpayers are in the 15 percent bracket. The tax rate cut for such taxpayers has already been implemented through the addition of a new 10 percent bracket at income of up to $12,000 for married couples, and no further changes are scheduled before 2008. The majority of taxpayers would thus receive no benefit from the accelerated rate reduction. As noted above, the marginal propensity to consume income among the higher-income group who would benefit from the acceleration is below that of lower- and moderate-income groups. Since most of the cost would occur after 2002, and since the amounts that would flow to individuals in 2002 would be concentrated among those with relatively low propensities to consume, accelerating the tax cuts would have a low bang-for-the-buck.

The proposal has two other shortcomings. First, the equity of giving tax cuts to those in the highest-income groups, while ignoring those in lower-income groups (who are most likely to be the ones losing their jobs in a downturn), is questionable. Second, preliminary estimates suggest that accelerating the rate cuts would "bust the stimulus budget" figures noted by the President, Chairman Greenspan, and former Secretary Rubin.

In terms of longer-run considerations, accelerating the rate cuts would have the political effect of helping to lock in further tax rate cuts for the highest-income taxpayers. Yet in the near future, Americans will need to reconsider such further reductions in tax rates for high-income taxpayers, as the economic and budget outlook after the terrorist attack and the current slowdown become clearer. Regardless of one's views about whether the tax cut passed last spring was sound fiscal policy at that time, it is clear that the nation's finances are now much worse off than they were then. Careful consideration will have to be given to the policy adjustments necessary to maintain a sound long-term fiscal position (particularly in light of ongoing anti-terrorism costs).

III. Business tax cuts

Incentives for new business investment

An alternative set of proposals would aim to stimulate business investment. This could include an investment tax credit or accelerated depreciation schedules (or expensing, a form of accelerated depreciation) for new investments. Of the business-oriented tax proposals currently under discussion, such temporary investment incentives are the most consistent with the principles delineated above. These incentives would provide a relatively strong bang-for-the-buck because they are targeted on new investments and because they are temporary, thus minimizing the impact on interest rates. As discussed below, these features place them in sharp contrast to corporate income tax rate cuts.

It is important that any investment incentives be temporary for four reasons. First, making the incentive temporary would encourage firms to shift investments into 2002—and therefore maximize the stimulus effect in 2002. Second, a temporary incentive involves significantly lower budgetary cost—and therefore less harmful pressure on interest rates—than a permanent incentive. Third, the costs and benefits of permanent tax incentives for investment are complicated, and debate over whether such permanent incentives would be advisable would divert policy-makers from the immediate task at hand. Finally, allowing permanent investment incentives would open the door to other permanent components of the stimulus package, which would undermine its effectiveness.

The precise form of a temporary investment incentive should reflect administrative and other issues. For example, some practitioners believe that accelerated depreciation or partial expensing may be slightly easier to implement than an investment tax credit. Moreover, whatever their form, it is also important to note that such incentives are not fool-proof. Their impact on investment may be limited, especially when firms already have significant cash-on-hand, there is excess capacity, and aggregate demand is falling. The various concerns should serve to reduce expectations about the impact of temporary incentives for business investment. None of them suggest, however, that either a permanent incentive for investment or a corporate tax rate cut would be more effective as a stimulus.

Corporate tax rate cuts

Some commentators and business leaders are advocating reductions in the tax rate on corporate income as the best method for stimulating the economy. Advocates claim the tax cut would raise stock market values (thereby stimulating consumer spending ), reduce costs of goods to households, and increase investment. A corporate tax rate cut, however, is ill-suited to address the nation's short-run economic challenges.

A corporate tax rate cut is a poor stimulus for three reasons. First, it is an inefficient way to stimulate new investment because it provides a windfall to the income earned on investments made in previous years. Firms would benefit if they are profitable, even if they are making no current investments or are reducing their current investments. Second, the corporate tax rate cut provides little, if any, immediate assistance to firms that are currently facing losses and therefore are not currently paying corporate income taxes. These are, however, the firms that are disproportionately in need of assistance during the downturn. Third, permanent reductions in corporate tax rates are expensive—a 10 percentage point decline in the corporate tax rate reduces projected surpluses by about $900 billion over the next decade, including about $700 billion in tax revenue losses and $200 billion in additional interest on the public debt. This decline in the surplus would put upward pressure on long-term interest rates, which in turn would erode most or direct positive effects of a corporate tax cut on the stock market and on new investment.

Some have claimed that cutting the corporate rate would raise stock values and thus stimulate consumer spending. This chain of reasoning is problematic for several reasons. First, even if it were valid, the corporate tax rate cut would imply a very small short-term stimulus relative to its long-term cost. Our estimates suggest that consumption would rise by $36 billion to $60 billion in the next year, or between 4 and 7 percent of the 10-year cost if interest rates did not adjust. Other tax reductions noted above could achieve a much larger short-term stimulus per dollar spent.

As noted above, however, the decline in the surplus would raise interest rates. Separate estimates of the relationship between the budget and interest rates by President Clinton's Council of Economic Advisers (CEA) and Harvard Professor Martin Feldstein (chairman of the CEA under Ronald Reagan) imply that the cut in corporate tax rates would cause interest rates to rise sufficiently to wipe out almost all of the stock market effect claimed by advocates, and hence also eliminate almost all the effect on consumer spending.

A second claim is that the tax cuts would be immediately passed along to consumers in the form of lower prices for goods and services. But most evidence suggests that the corporate tax is borne by owners of capital in general or owners of corporate capital, not by consumers. In addition, recall that firms that are currently losing money do not pay corporate income taxes. Thus, if savings were passed to consumers, survival would become even more difficult for those firms already hit hardest by recent events.

A third claim is that corporate tax rate cuts would raise investment by reducing the cost of capital investment. This argument is correct as far as it goes, but it ignores several critical factors. For example, for a tiny fraction of the cost of a corporate tax rate cut, a temporary investment tax credit (ITC) could generate the same or larger incentive to undertake new capital investment during the next year. The ITC is cost-effective because it only subsidizes new investment, while the corporate tax cut reduces burdens on both new investment and current income (which is the return to old investments). Also, because the ITC would be temporary, revenue losses would not extend into the future and the impact on interest rates would be much smaller. Accounting for all of these factors, a permanent corporate tax rate cut costs between 10 and 30 times the amount of a temporary ITC that generates the same reduction in the cost of new capital investment over the next year. (Other forms of temporary tax incentives for new investment would similarly be much more cost-efficient than a corporate tax rate cut.)

Some claim that a corporate tax cut would generate new investment by raising corporate cash flow. But from the end of 1999 to the middle of 2001, non-financial, non-farm corporations raised their liquid financial assets by $100 billion and their total financial assets by $700 billion. This increase did not stop a significant decline in investment earlier this year, but it does show that the decline was not due to any shortage of cash on hand.

A final claim is that corporate tax rate cuts would help reduce layoffs. Such an effect seems unlikely. After all, corporate income tax cuts do not help firms that are already losing money, and these are the firms most likely to lay off workers.

Capital gains tax cuts

Another proposal would reduce the maximum tax rate on long-term capital gains from 20 percent to 15 percent. Whatever its merits in other contexts, a capital gains tax cut has several crucial drawbacks under current circumstances.

A permanent capital gains tax cut is poorly designed to address the short-term economic problems at hand. Lower tax rates on capital gains are typically supported as a way to raise saving, but what the economy needs right now is more spending, not more saving. Lower capital gains rates are also claimed to raise economic growth, but the effect is probably tiny and certainly many years in the making. In contrast, what the economy needs right now is a short-term boost. In addition, a capital gains tax reduction is not an efficient way to stimulate new investment, because the tax cut would apply to capital gains on existing assets, and those gains are a return to prior investment. Finally, capital gains tax cuts would reduce long-term revenue and exert upward pressure on long-term interest rates.

A temporary reduction in capital gains tax rates is perhaps even more problematic, since it would encourage people to sell their stocks now. If people took the revenue from their asset sales and reinvested in the market, there would be little effect on any economic aggregate. If instead they spent their realized gains on consumption goods, the result could reduce stock market values, which could hurt consumer confidence and business investment.

IV. Conclusion

The current economic and budget outlook suggests the need to focus on policies that stimulate the economy in the short run and do not damage the long-term fiscal outlook. Policies that are not consistent with these principles will generate weaker economic stimuli than those that do. There are limits, however, to the effects of any type of tax cut in stimulating economic activity, particularly in times of uncertainty when people and firms are delaying major economic decisions.

This article will be supplemented periodically as new tax proposals are discussed in the policy arena. For further detail on these issues, see William Gale, Peter Orszag, and Gene Sperling, "Tax Stimulus Options in the Aftermath of the Terrorist Attack," Tax Notes, October 8, 2001.