In this vein, Matt Mitchell at the Mercatus Center argues that "the data are making it increasingly more difficult to argue that redistribution through unemployment benefits is both humanitarian and stimulative." The data, Mitchell believes, are in a study of the 2008 Bush stimulus package. This so-called Stimulus I provided tax rebates of between $300 and $600 to single households making up to $75,000 and jointly filing households making up to $150,000, with a $300 bonus for each eligible child.

The 2009 study shows that, in the words of Mitchell, "the poor" were less likely to spend their tax rebates than "the wealthy." (About one-fifth of households reported mostly increasing spending, while more than half reported mostly paying off debt.) He then argues that this flies in the face of the theory - laid out in an Alan Blinder Wall Street Journal op-ed - that unemployment insurance stimulates spending:

The key to this reasoning is [Blinder's] assertion that the poor are more likely to spend a marginal dollar than the wealthy.

But Mitchell has wholly conflated "the poor" with "the unemployed." Unemployment insurance targets assistance to a group of people whose consumption, research shows, falls by more than 20 percent in the absence of the insurance. The benefits prevent a sudden income shock from forcing a household to cut back spending dramatically, reducing aggregate demand in the economy. Further, the benefits help unemployed homeowners afford mortgage payments, keeping them out of foreclosure and aiding the ailing housing market.

"The poor" might need assistance, as well, but their spending decisions are quite different from the unemployed. In fact, criticism of the type of tax cut/tax rebate assistance provided by Stimulus I centered on the very concern that households would save the money and use it to pay down debts instead of spending it. This concern is even built into the calculation of the fiscal "bang for the buck" that results from government stimulus: in October of 2009 Mark Zandi estimated that a dollar in refundable tax rebates produces $1.22 in economic activity, but that extending unemployment insurance benefits produces $1.61 in economic activity, the most of any stimulus measure.

Indeed, there was discussion of the potential benefit of targeting assistance to lower-income households. But this came primarily in reaction to tax cuts that would have been a sop to the wealthy. While the stimulative benefits of the Bush stimulus might have been weak (Zandi, for his part, thinks it moderately effective), it still allowed many lower-income households to pay down debt, freeing up cash for the future, and ended up phasing out pointless rebates for the wealthiest taxpayers.

In short, research on Bush's 2008 stimulus provides important information about how households react to tax rebates during an economic downturn. In particular, it suggests that policymakers should be wary of tax cuts as a means of stimulating aggregate demand.

What it says nothing about, though, is the humanitarian or stimulative impact of extending unemployment benefits.