Share |

Finally Some Real Evidence on the Stimulus: Less Effective than You Think

e21 | September 28, 2010
Getty Images

Have the various stimulus programs inaugurated since 2008 been successful or not? What aspects of stimulus programs are more or less effective? These are tricky questions to answer, yet crucial in determining the scale and nature of the government’s fiscal response to economic downturns. An excellent new analysis based on household surveys by Matthew Shapiro, Claudia Sahm, and Joel Slemrod sheds a great deal of light on the subject.

Unlike prior stimulus evaluations, which assume their conclusion by starting with the idea that tax cuts and government spending are inherently additive to the economy, this report is based on an actual empirical survey. As a result, the analysis reflects the realities on the ground rather than merely reinforcing the inputs of the researcher. There are a number of surprising results, but here is a review of some of the major findings

The chief surprise is simply how little of stimulus programs delivered through tax cuts boosted spending. The researchers find that in response to the one-time tax cuts in 2008, only 25% of households reported higher spending. This result is consistent with a large body of economic literature that suggests households smooth out their consumption over time, and so an unexpected windfall only slightly raises current levels of spending.

More surprising is the analysis of the tax cuts associated with the 2009 ARRA stimulus. These took effect through the form of reduced federal tax withholding on payrolls, and were widely expected to be more effective than one-time tax cuts in elevating aggregate spending. Yet the authors find that a mere 13% of households reported higher levels of spending due to the tax cuts. While the deteriorating economy may account for part of the decrease in the number of households responding to the stimulus, the authors are able to isolate the change in payment mechanism as the primary reason that the 2009 stimulus was less effective than the 2008 stimulus.

Among other things, this analysis highlights the limitations of using behavioral economics to inform policymaking. The choice of delivering the stimulus partially through reducing federal withholding was motivated by research in behavioral economics. Many from this school of thought suggested reductions in payroll withholding would not be particularly salient to taxpayers, and argued that households instead would be more likely to spend money as a result. It now looks like the opposite was the case. Taxpayers spent even less money when their payrolls were cut than when they received a one-time payment.

The report also challenges another key plank of the Administration’s policymaking: that tax cuts delivered to the poor offer more bang-for-the-buck than tax cuts delivered to the rich. This idea underlies a large number of policy choices, from unemployment insurance extensions to extending the Bush tax cuts for the middle-class but not the rich. Yet the authors find this was not the case for the stimulus program. When considering households that report they will respond to the tax cut by “mostly spending,” the researchers found:

Households who report being worse off financially are almost 7 percentage points less likely to report mostly spending the additional income from any given policy

Interestingly, the authors also find comparable effects for households that expect economic conditions to deteriorate in the future:

A household that expected their income to decline by more than 10 percent over the next year has a mostly-spend rate almost 8 percentage points lower than a household that expects their income to be unchanged

This is plausible if one thinks that poor or struggling households choose to save additional income rather than spend it, a seemingly wise decision. Yet that, in turn, implies that the “multiplier” of spending in response to, say, extensions of unemployment insurance may not be as high as many had projected. Expectations of the future loom large in predicting behavior today. Whether future disposable income is expected to be lower due to tax increases or a deteriorating economy, the natural reaction among households is to reduce spending today.

These results do not suggest that tax cuts, in general, are a bad way to deliver growth in prosperity. The study only examined the impact of temporary, one-time cuts in taxes, as opposed to permanent decreases in marginal tax rates. More permanent cuts in tax rates, especially if spending cuts are used to offset them, induce supply-side responses that raise long-term labor supply and output. Additionally, households facing a long-term decrease in taxes envision a long-run rise in wealth, which should induce a larger consumption response compared to a temporary boost in income.

Nor do these results necessarily suggest that the tax component of the stimulus was a bad idea. If the goal was simply to transfer income to poor or struggling households, the program may have worked as intended. Yet according to the stated goals of the stimulus – to boost spending, output, and employment – the tax component was a failure in the eyes of these researchers.

In fact, it appears that there is in fact a conflict between the goals of addressing household insecurity and boosting spending effectively. The idea that the poor are (especially) likely to spend money united the Administration’s progressive base. Extending aid for struggling families was good not only for its own reasons, but also because of its role in boosting spending and consumption. Yet, if spending/boosting policies and pro-poor policies are not identical, then the Administration must face tough tradeoffs.

For instance, if the goal were simply to maximize overall spending, a credible case could be made that a more effective stimulus measure would have been a large, one-time tax rebate to affluent households. Of course, for a variety of reasons, such a proposal would almost surely get blocked in Congress. (It’s also worth noting that just like with short-term tax policies aimed at low-income individuals, the permanent income hypothesis does suggest that even these new resources would be invested.)

Similarly, this research has implications for the withdrawal of the Bush tax cuts for individuals in the top income tax brackets. Supporters of raising these taxes have emphasized demand-side pressures. They generally argue that the rich are particularly likely to save money, rather than spend it, and money would be better allocated to lower-income households. Supporters of maintaining all the current tax rates – including those for households in the top brackets, like Alan Viard – have instead emphasized the long-run beneficial consequences of low tax rates.

This NBER paper suggests an additional, demand-side impetus for extending tax cuts for wealthy individuals. Higher income households may be no less likely to save incremental after-tax dollars than low-income households. Given that households in the top 5% of the income distribution earn over 37% of all income, it follows that increased taxes on this group are likely to disproportionately reduce consumption expenditures and cause the economy to contract in 2011.

Christina Romer, along with other Administration officials, has talked about how important it is to avoid the errors that prolonged the Great Depression. For this group, this means avoiding measures of fiscal austerity that they believe were responsible for the 1937 slowdown. If higher taxes on the rich would have particularly baleful effects on the economy, will the Administration take steps to avoid the tax increase, or are they more concerned about growing the tax base to finance future spending programs? Again, the identification of the high income households with “savers” prevented such difficulties from cropping up in the past; but if this research holds up, the Administration – and all other stimulus advocates – will face a new and challenging set of trade-offs.

Are we to draw definitive conclusions from just this one paper? That’s probably never a good idea, but these results are worth bearing in mind when hearing expansive claims of how many jobs would have been lost had no stimulus been enacted. We simply do not know the effect of the stimulus on the broader economy. We do now have evidence, however, that the tax cuts that made up both the Obama and Bush stimulus did a poor job at boosting overall spending, and that the nature of Obama’s tax cuts was particularly ineffective in terms of boosting output. As long as households remain burdened with debt, attempting to boost spending through temporary tax cuts will likely prove difficult.


e21 Projects & Partnerships