November 27, 2001
TAX CUTS ARE NOT AUTOMATICALLY THE BEST STIMULUS:
A Response to Glenn Hubbard
by Peter Orszag and Joseph Stiglitz(1)
PDF of report
If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.
On November 16, Glenn Hubbard, the Chairman of President Bush's Council of Economic Advisers, published an op-ed article in the Washington Post entitled, "Tax Cuts are the Best Stimulus."(2) In that op-ed, Hubbard argued that "it is a major fallacy to praise new spending plans as 'stimulus,'" and that tax cuts are far more effective than expenditure increases in stimulating the economy. Hubbard placed particular emphasis on business tax cuts, writing that "business tax relief is essential because it encourages firms to expand investment and employment."
The Hubbard op-ed recapitulates many of the arguments that various Bush Administration officials have been making in recent weeks. Given the current state of the debate over a stimulus package, an assessment of these arguments is in order. This purpose of this brief analysis is to examine these arguments, as reflected in Mr. Hubbard's article.
As explained below, we find these arguments to be inconsistent with sound, mainstream economic thinking. Basic economic analysis indicates that increased government expenditures can indeed be stimulative, and, in fact, are often more effective as stimulus measures than tax cuts.
Indeed, Mr. Hubbard's own figures reveal the weakness of his reasoning. In the op-ed, he claims the President's stimulus plan would create 300,000 more jobs. The Joint Tax Committee has estimated that the tax-cut components of the Administration's plan would cost more than $90 billion in fiscal year 2002. The cost for each job created or saved under the Administration's stimulus package thus would exceed $300,000. The reason that the cost per job created or saved would be so high is that the Administration's package is poorly designed to provide short-term stimulus to the economy. Policymakers can design a stimulus package with a much larger bang for the buck if they can move beyond the tax-cut ideology that seems to underlie both the Administration's proposals and Mr. Hubbard's article.
The Basic Flaw in the Hubbard Article
The principal reason that Mr. Hubbard's arguments are misguided and the main reason the Administration's package would be relatively ineffective as a stimulus measure is that they largely ignore the central feature of a recession: lack of demand. In a recession, the primary problem is that the nation's firms face a reduction in demand for their products not that they lack available workers, equipment, or anything else needed to produce goods and services. Indiscriminately injecting cash into such firms through tax breaks, without linking the tax breaks to new business activity, would do little if anything to address the underlying difficulty.
Firms that are faced with reduced demand for their products lay off workers, regardless of how much cash they have. The managers of firms have a fiduciary responsibility to maximize their profits, and in the face of reduced demand for their product, firms therefore typically reduce costs by cutting back on production, which triggers layoffs. As the number of unemployed workers increases, a downward economic spiral can occur. Households with unemployed workers, facing a sharp decline in their incomes, cut back on spending and further reduce the demand for products. That, in turn, leads to additional layoffs. This harmful cycle, by which an economic slowdown can build into a more serious recession, can be arrested or broken by boosting demand for the goods and services that American companies produce. Only when a company faces renewed demand for its products will it end the process of shedding workers and begin to create new jobs. As a result, the primary objective of a stimulus package should be to spur spending on these products.
An effective stimulus package consequently should expand the aggregate demand for goods and services in a timely way. Mr. Hubbard notwithstanding, there is little question that increases in government expenditures can be quite effective in boosting aggregate demand and thereby stimulating the economy in the short run.
Temporary expansions in unemployment insurance, for example, would spur increased consumer spending. Households in which a worker is laid off experience a significant decline in income. They thus are likely to spend a high percentage of any additional income they receive while out of work. The extra spending on unemployment benefits that a temporary expansion of unemployment benefits provides thus has a direct economic benefit it keeps more workers employed at firms that produce the products the unemployed workers purchase with their additional cash. Temporary expansions in unemployment insurance consequently are a "win-win" proposition: They are quite effective in helping more people keep their jobs during an economic downturn, and they also assist those who are unfortunate enough to have lost their jobs.
Analyzing Mr. Hubbard's Arguments
Mr. Hubbard argues against increasing expenditures on government programs as part of a stimulus package for two reasons. First, he contends, "a dollar spent by the government is one fewer that can be spent by private businesses." Second, he argues that, "new spending programs almost never go away..." Both of these assertions make catchy sound-bites. In the context of an economic downturn, however, neither is correct.
The notion that a dollar spent by the government "crowds out" one dollar of spending by private businesses is correct only when the economy's resources are fully utilized; in that case, additional demand on those resources by the government necessarily reduces the demands that can be placed on them by the private sector. But when the economy's resources our workers and plants and equipment are not fully utilized, government spending does not displace private-sector resources on a dollar-for-dollar basis. Indeed, during an economic downturn, government spending can "crowd in" additional private-sector activity by spurring overall demand and thereby making it more likely that firms will be willing to make new investments. Mr. Hubbard's argument about government spending fully crowding out business spending thus is puzzling; it would be valid only if the economy were fully utilizing its resources, which is clearly not the case now.(3)
The latest data, for example, show that the capacity utilization rate the proportion of plant and equipment capacity being used in production fell to 74.8 percent in October, its lowest level since 1983. Furthermore, any implication that the economy is fully utilizing its resources would be inconsistent with another statement Mr. Hubbard makes namely, that "the economy needs help now."
Mr. Hubbard's second argument against including expenditure increases in a stimulus package that "new spending programs almost never go away " also is unfounded. This statement implies that federal spending tends to rise inexorably over time and that program expansions initiated in response to recessions inevitably continue after the economy recovers. Congressional Budget Office data show, however, that total federal spending has fallen from 22.3 percent of the economy (i.e., of the Gross Domestic Product) in 1991 to 18.2 percent in 2000, and that federal spending is significantly lower now, as a share of the economy, than it was during the 1980s.(4)
Furthermore, under almost any plausible budget scenario for coming years, federal spending is projected to decline further as a share of the economy after the downturn ends and by 2010, to reach its lowest level since 1965.(5)
In addition, Congressional leaders who have proposed including increases in program expenditures in a stimulus package have called for making such measures temporary; Administration rhetoric to the contrary, no permanent entitlement increases are being proposed. History suggests that commitments to make these program expansions temporary would be honored; in every recession of recent decades, expansions in unemployment insurance were provided that were considerably more generous than the limited expansions the Administration is now proposing, and in every instance, these unemployment insurance expansions were withdrawn when the economy recovered and unemployment declined. In short, the marked decline in recent years in federal spending as a share of the economy, the temporary nature of the program expansions being proposed, and the historical record regarding temporary program expansions in previous recessions all contradict Mr. Hubbard's argument.
Mr. Hubbard's Key Themes
As we have just seen, a key theme of the Hubbard op-ed that temporary spending increases would be ineffective as stimulus but pose longer-term dangers to the economy is without foundation. So is another basic theme of his article: that corporate tax breaks would automatically constitute effective stimulus.
A number of the corporate tax breaks included in the House stimulus bill or the Administration's stimulus package would be ineffective as stimulus measures because they would give large sums of cash to corporations without providing much, if any, incentive for the corporations to use this cash to make new investments or retain more workers. The proposed repeal of the corporate alternative minimum tax and the House provisions allowing a longer carry-back of business losses are examples of such provisions. Both measures would provide tax breaks to firms regardless of the firm's current investment or employment activities.
Indiscriminately injecting more cash into firms may, at first blush, sound like a good idea because it would give firms more money that they could spend. But the primary problem currently facing most firms is that their customers are reducing their purchases; for the average firm, the problem at hand is not a lack of available cash. Indeed, from the end of 1999 to the middle of 2001, holdings of liquid financial assets by nonfarm, nonfinancial corporations rose by more than $100 billion (or more than 17 percent), and holdings of total financial assets by such firms rose by more than $700 billion (or almost 9 percent).(6)
Furthermore, if the problem is that some firms are prevented from undertaking new activities because they are cash-constrained, the tax incentives should be focused on such firms and conditioned on undertaking new activities. Yet the corporate tax provisions in the House stimulus plan and the Administration stimulus package do not generally target firms that are cash-constrained or that would expand their operations in response to the tax incentive. For example, according to data from Citizens for Tax Justice, General Motors would receive $833 million from full repeal of the corporate alternative minimum tax.(7)
Yet General Motors has roughly $8 billion in cash and, according to a company spokesman, it has no plans to increase the amount it invests each year in plants and new products.(8)
Such tax breaks would instead primarily feed through into additional cash for corporate shareholders, who tend to be disproportionately higher-income. Since higher-income households tend to spend relatively little of any additional income they receive, little additional demand for goods and services would be created.(9)
Mr. Hubbard's Imaginary Tax Increase
In his op-ed, Glenn Hubbard argues that the Senate Finance Committee legislation would raise taxes on employers. This argument is misleading with respect to the version of the legislation that the Finance Committee approved, and is simply incorrect with respect to the version of the legislation that was considered on the Senate floor.
The version of the legislation the Finance Committee approved would have financed its temporary expansions in unemployment insurance benefits out of the Federal Unemployment Insurance Trust Fund. Under a statute known as the Reed Act, when reserves in that trust fund surpass certain levels, the additional reserves must be transferred to state unemployment accounts. States can effectively use the transferred funds for several purposes, one of which is to reduce unemployment insurance tax rates. If a portion of the reserves in the Federal Unemployment Trust Fund were used to finance the provision of additional unemployment benefits during the current recession, the "excess" amounts transferred to states in future years while still substantial would be smaller than they would have been if none of the federal trust fund reserves had been used for additional benefits now. It thus could be argued that using Federal Unemployment Trust Fund reserves for additional federal unemployment benefits during the current recession would result in smaller state reductions in unemployment insurance taxes in the future than would otherwise be instituted.
To the extent that the Finance Committee provisions would cause the ensuing reductions in state unemployment insurance taxes to be smaller than would otherwise be the case, however, the change would occur in future years, not today. Mr. Hubbard misleadingly implies in his op-ed that the Finance Committee bill would result in counterproductive tax increases that would occur now while the economy is weak. Furthermore, under the bill the Finance Committee approved, states would still be expected (according to Congressional Budget Office estimates) to receive approximately $20 billion in these "excess" federal unemployment funds over the next 10 years, which they could use to cut unemployment insurance taxes. (If no federal trust fund reserves are used for unemployment benefits during the current downturn, states would receive an estimated $40 billion over the next ten years.)
Because of fears that the Finance Committee's action would be criticized as ultimately resulting in a tax increase, the Senate Democratic leadership changed this provision of the legislation when the Finance Committee bill came to the Senate floor shortly before Thanksgiving. In the version of the legislation placed before the full Senate, the temporary unemployment insurance expansions would be financed from general revenues rather than from the Federal Unemployment Insurance Trust Fund. As a result, the transfers of "excess" trust fund revenues to the states would be unaffected; states would still receive the full $40 billion in "excess" funds from the Federal Unemployment Insurance Trust Fund over the coming decade. As a result, the misleading argument that Mr. Hubbard employed in his article is not applicable to the legislation the full Senate considered.
In short, there is no basis for concluding that the Senate Finance Committee legislation, as placed before the Senate, would raise taxes on employers. Mr. Hubbard's assertion on this matter is incorrect.
To be sure, some corporate tax breaks would boost spending. In particular, temporary corporate tax breaks that are tied to new investments by firms can help to spur more short-term spending by the firms. A firm that decides to build a new plant or to purchase new computers now rather than in the future in order to take advantage of a temporary investment tax incentive will have to purchase materials, services, or equipment from another firm, such as a construction company or computer manufacturer. The resulting short-run effect on the economy is similar to that which occurs when substantial numbers of unemployed workers boost their expenditures because their unemployment benefits have been increased. Unfortunately, most of the corporate tax breaks included in the House stimulus legislation and the Bush Administration's stimulus proposal do not provide such incentives. Generic support for corporate tax breaks as economic stimulus regardless of whether the tax breaks are tied to new investment on the simplistic grounds that these tax breaks would provide firms with more cash is unwarranted.
Two final points relating to Mr. Hubbard's article are worth noting. First, many of the Administration's stimulus measures appear to be focused more on longer-term considerations than short-term stimulus. Any contention that we need more tax cuts to strengthen the economy's long-term prospects, however, would be inconsistent with Mr. Hubbard's statement that the recent terrorist attacks "did not undermine long-term productivity growth, so economic fundamentals remain strong." If the long-term fundamentals remain strong and the problem is that the economy is weak in the short run a sound assessment of our circumstances why isn't the Administration focusing on spurring the economy now when it needs help? Why is its stimulus package dominated by permanent and multi-year tax cuts that would produce little or no short-term stimulus? Moreover, a stimulus package is not free: It leaves us with more government debt than we otherwise would have, which can adversely affect longer-term economic performance. A package with such a low bang for its buck produces little short-term benefit to justify its budgetary cost.
Second, although some recent data suggest the economic downturn may be relatively short-lived, the outlook remains highly uncertain and many economists remain concerned that the downturn could be significantly worse than currently expected. Given such uncertainty, a well-designed stimulus package should adjust automatically to changing circumstances. In other words, the best-designed stimulus measures would expand automatically if the downturn turns out to be more severe than currently expected but contract automatically if the economy recovers more rapidly than anticipated. Expansions in unemployment insurance benefits work in precisely this manner: As unemployment mounts, the expansions in unemployment benefits automatically enlarge to cover more workers and in so doing provide additional stimulus. Conversely, the scope and cost of expansions in unemployment benefits automatically contract to cover fewer workers when unemployment declines. By contrast, the Administration's proposed tax cuts do not respond to changes in economic activity in this manner. In fact, many of the Administration's tax proposals such as the proposal to accelerate reductions in marginal tax rates for higher-income taxpayers would grow more costly if the economy experiences a strong recovery than if it does not.
The bottom line is that many proposals to increase program expenditures including temporary expansions in unemployment insurance would stimulate the economy, minimize job loss, and adjust automatically to changing economic conditions. In comparison, many of the tax breaks the Administration is championing would have a low "bang for the buck" in bolstering demand, would do little to help avoid layoffs, and would not provide an automatic insurance policy of injecting additional stimulus into the economy if the recession turns out to be more severe than expected. Although Mr. Hubbard strains to come to opposite conclusions, his article is unconvincing.
1. Peter Orszag is the Joseph A. Pechman Senior Fellow in Tax and Fiscal Policy at the Brookings Institution, and previously served as a senior economist on the Council of Economic Advisers. Joseph Stiglitz is Professor of Economics at Columbia University, one of the recipients of the 2001 Nobel Prize in Economics, and a former Chairman of the Council of Economic Advisers.
2. Glenn Hubbard, "Tax Cuts are the Best Stimulus," Washington Post, November 16, 2001.
3. On a related front, Mr. Hubbard's assertion that proposed expansions in unemployment insurance would increase unemployment is based on the argument that expanded unemployment benefits would reduce the incentive to seek work. That argument ignores the fact, however, that in an economic downturn, expanded unemployment benefits would spur consumer spending and thereby help keep more workers employed. Fundamentally, Hubbard's view assumes that even in a recession, unemployment is predominately caused by a lack of effort in finding a new job, rather than a lack of available jobs. We disagree: in a recession, inadequate aggregate demand and therefore a lack of available jobs plays an important role in causing unemployment. In the context of a recession, any negative incentive effects from expanded unemployment benefits are likely to be outweighed by the positive impact on aggregate demand. As a result, the net effect of expanding benefits in a recession would likely be to reduce, not increase, unemployment.
4. The figures for appropriated spending are equally striking. Such spending amounted to 6.3 percent of GDP in 2000, the lowest level on record, going back to 1962 (the first year for which these data are available). Appropriations constituted 12.7 percent of GDP in 1962 and did not fall below 7.0 percent of GDP until 1996. In part, the steep decline in appropriated spending reflects the end of the Cold War and the resulting decision to reduce the size of the armed forces. Even so, domestic appropriations also are very close to their lowest level on record as a share of the economy, going back to 1962.
5. In a document released on a bi-partisan basis in early October, the Chairmen and ranking members of the House and Senate Budget Committees listed an array of potential future expenditures, including a prescription drug benefit, farm legislation, increased education spending, measures to reduce the ranks of the uninsured, increased defense and security expenditures, and costs of responding to national disasters, among others. If all of the potential expenditures the Budget Committee listed are incurred, federal expenditures would still decline to 17.7 percent of GDP by 2010, which would be the lowest level since 1965. See Richard Kogan, "Where Has All The Surplus Gone?" Center on Budget and Policy Priorities, November 1, 2001.
6. Federal Reserve Board, "Balance Sheet of Nonfinancial Corporate Business," Table B.102, September 18, 2001.
7. Citizens for Tax Justice, "House GOP Stimulus Bill Offers 16 Large, Low-Tax Corporations $7.4 Billion in Instant Tax Rebates," October 26, 2001.
8. Glenn Kessler, "Minimum Tax Spurs Maximum Pressure; Bush, Business Push For Repeal, Refunds," Washington Post, November 23, 2001.
9. Mr. Hubbard raises doubts that lower-income households tend to spend a higher percentage of income than higher-income households. The weight of the most recent evidence, however, suggests that lower- and middle-income households do have higher propensities to consume out of available resources. See, for example, Karen E. Dynan, Jonathan Skinner, and Stephen P. Zeldes, "Do the Rich Save More?" NBER Working Paper 7906, National Bureau of Economic Research, September 2000; Jonathan Parker, "The Consumption Function Re-estimated," mimeo, August 1999; and Jonathan McCarthy, "Imperfect Insurance and Differing Propensities to Consume Across Households," Journal of Monetary Economics, November 1995, pages 301-27.