Likely Effects of the Tax Rebate Checks

Following my recent musings about the tax rebate checks, several people asked about the likely economic consequences of this sort of policy.

My friend and colleague, Nick Souleles, is one of the leading experts on these matters, so I asked him for a short primer on what we learned from when rebate checks were sent out in 2001.

Here is a list of three of the best recent contributions, in chronological order. Taken as a group, they both provide some useful guidance, and highlight remaining uncertainties — and the thing I really like is how these papers highlight some of the ways in which economists are willing to get their hands dirty with some particularly innovative research strategies.

1. “Consumer Response to Tax Rebates,” by Matthew Shapiro and Joel Slemrod, American Economic Review, 93(1), March 2003.

The authors use a fairly direct way of figuring out the effects of the 2001 rebate checks: They ask people.

The answer: Around one-quarter planned on spending it, one-third planned to save, and the rest were planning on using it to pay off debt. Interestingly, the rich were more likely to spend the rebate, while those on lower incomes planned on paying off debt.

2. “Household Expenditure and the Income Tax Rebates of 2001,” by David Johnson, Jonathan Parker and Nicholas Souleles, American Economic Review, 96(5), December 2006. A nice summary from the N.B.E.R. Reporter is available here.

This paper shows some real ingenuity: the authors note that the 2001 tax rebate checks were sent out over a ten-week period, with the date depending on the second-to-last digit of your social security number, which is effectively a random number. Sensing a research opportunity, the authors persuaded the B.L.S. to add some questions to the Consumer Expenditure Survey, allowing them to compare the timing of the rebate payment and the timing of expenditures.

All told, they estimate that on average around two-thirds of the rebate was spent during the ten-week disbursement period and the subsequent three months. Interestingly, around a third of the total response came from a rise in spending on clothes.

3. “The Reaction of Consumer Spending and Debt to Tax Rebates — Evidence from Consumer Credit Data,” by Sumit Agarwal, Chunlin Liu and Nicholas Souleles, Journal of Political Economy, 115(6), December 2007 (un-gated version here).

This paper begins with one particularly compelling observation: credit card companies know our social security numbers (and hence who got their rebates when), and they also know a lot about our spending and saving patterns.

And so once the authors were able to get a large credit card company to share with them (anonymized) data, their research project was made.

Recall that paper #1 had found that nearly half of all respondents expected to use the rebate to pay down their debt. It turns out that this was the initial response of many, but then over the ensuing nine months, spending rose by enough to account for around two-fifths of the average rebate. And for those who were liquidity constrained, spending rose even further.