Tax Rebates as Stimulus

It's very difficult to get truly objective, nonpartisan analysis of stimulus issues. Most people look to the CBO for authoritative, unbiased analysis. While I respect the CBO and think it's reputation or honesty is well deserved, I also know that bills are often drafted with the CBO's scoring methodology specifically in mind. The majority party also gets to pick the head of the CBO, and there's no denying that it selects someone with similar ideological sympathies.

In truth, the best place to look for nonpartisan expert analysis is the Congressional Research Service. The CRS is Congress' private think tank; it's staffed with experts on every conceivable area of policy, both substantive and procedural. CRS analysts are true experts in their field, and they do a great job sifting through the academic research and highlighting the most reliable results. Congress is fiercely protective of the CRS—the CRS's memoranda to Congressmen and Senators are confidential, and, amazingly, CRS reports aren't free to the public. (OpenCRS.com usually has slightly older versions of CRS reports, which are updated frequently.)

So to cut through all the partisanship surrounding the stimulus debate, I went to the CRS reports (which my firm bizzarely pays for). The one thing that surprised me in reading over all the relevant CRS reports is that the recent evidence on tax rebates shows that they're actually quite effective as stimulus. For example, this is what the CRS report titled, Tax Cuts for Short-Run Economic Stimulus: Recent Experiences, says about the 2001 rebate:

The rebate met some important standards for an effective tax cut stimulus. Unlike many stimulus proposals in the past, particularly in the 1960s and 1970s, where the stimulus occurred during the recovery rather than the recession phase (potentially adding to inflationary pressures), its impact occurred during the recession. In addition, tax cuts are most effective as a stimulus if they are spent, and the tax reductions affected lower and moderate income taxpayers who have a high propensity to spend.

At the same time, there was some concern that lump sum payments might be spent in the same fashions as a continued increase in income through tax reductions. There was some evidence that temporary rebates in the past were not spent. It appears, however, that most of the rebate was spent fairly quickly: at least 20% to 40% in the quarter received and two-thirds by the end of the second quarter after receipt.

Two-thirds of the rebate was spent within 6 months? This surprised me, since, according to Conventional Wisdom, the 2001 rebates failed as stimulus because most of it was saved rather than spent. Indeed, the most common criticism of last year's stimulus bill was "we tried tax rebates in 2001 and it didn't work."

The report cites a 2006 study in the American Economic Review by David Johnson, Jonathan Parker, and Nicholas Souleles, titled Household Expenditure and the Income Tax Rebates of 2001 (ungated draft here). The paper is quite compelling. It relies on a tremendous natural experiment, nicely summarized in the abstract:

Under the Economic Growth and Tax Relief Reconciliation Act of 2001, most U.S. taxpayers received a tax rebate between July and September, 2001. The week in which the rebate was mailed was based on the second-to-last digit of the taxpayer's Social Security number, a digit that is effectively randomly assigned. Using special questions about the rebates added to the Consumer Expenditure Survey, we exploit this historically unique experiment to measure the change in consumption expenditures caused by receipt of the rebate and to test the Permanent Income Hypothesis and related models. We find that households spent about 20-40 percent of their rebates on non-durable goods during the three-month period in which their rebates were received, and roughly another third of their rebates during the subsequent three-month period. The implied effects on aggregate consumption demand are significant. The estimated responses are largest for households with relatively low liquid wealth and low income, consistent with liquidity constraints.

The authors estimate that the rebate directly increased nondurable consumption expenditures by 2.9% in the third quarter of 2001 (the quarter in which the rebates were received), and by 2.1% in the fourth quarter. And that's without including multiplier effects.

The CRS report also says that the preliminary evidence on the 2008 rebate suggests that it worked surprisingly well:

Some preliminary evidence has been reported on the 2008 rebate. Although aggregate data do not show a rise in consumption to correspond with the rise in disposable income, these simple aggregate observations may not be very informative. In a preliminary study of households, Broda and Parker found that 20% of the rebate was spent in the first month. They predict that the 2008 rebate will have a significant effect on spending in subsequent months. Their study also found higher levels of spending for lower income households and those with fewer liquid assets.

The study by Broda and Parker, titled The Impact of the 2008 Tax Rebates on Consumer Spending: Preliminary Evidence, was published at the end of July. It relies on the same natural experiment as the Parker, Johnson, and Souleles study, making it extremely difficult to quibble with the methodology. Significantly, Broda and Parker found that the 2008 rebate was even more effective than the 2001 rebate (again, this is before including any multiplier effects):

Despite much recent concern that households would save the money, we find that households are doing a significant amount of extra spending because of the stimulus payments. The typical family increased their spending on food, mass-merchandise and drug products by 3.5 percent when their rebate arrived, relative to a family yet to receive its rebate. Based on these estimates and on results from previous analysis, we estimate that demand for overall nondurable consumption in the second quarter of 2008 has been boosted by 2.4 percent as a direct result of the stimulus payments, and will be held up by around 4.1 percent in the third quarter of 2008.

Unfortunately, the vast majority of commentators continue to ignore the evidence when discussing the stimulus bill, relying instead on Conventional Wisdom. James Surowiecki channeled this conventional wisdom in his New Yorker column last week:

Past tax rebates, as many economists have argued in recent weeks, haven’t seemed to boost consumption as much as was hoped. Some estimates suggest that when a rebate was handed out in 2001 less than half of it was spent. And while the results of last year’s rebate seem to have been somewhat more encouraging, much of it still went unspent.

This is, as we've just seen, demonstrably false. The evidence shows that people spent most of their rebates in the first 6 months (which, when we're talking about multi-year infrastructure projects, is quite fast).

Maybe this, rather than a misguided hope for bipartisan support, is why the Obama administration included a $145 billion individual tax credit in the stimulus bill. The business tax cuts were clearly included to woo Republicans, as all the evidence shows that business tax cuts are completely ineffective as stimulus. Since the stimulus bill passed the House with exactly zero Republican votes, maybe the Obama administration should give up on bipartisanship and strip the business tax cuts out of the bill.

After looking at the available evidence, however, I'm forced to support an increase in the individual tax credit. If the tax credit was increased to $200 billion, for example, the evidence suggests that roughly $133 billion of that would be spent within the first 6 months. That would tide us over nicely as we wait for the slower infrastructure projects to come online.