Way back in 2008 before the idea of fiscal stimulus became tightly linked with partisan politics, George W Bush signed a stimulus bill into law that passed congress largely on the backs of Democratic votes. The way it worked was that everyone got mailed a check (stimulus!) but the check was labeled a "tax rebate" (tax cuts!) so everyone was happy.
Unfortunately, the program did not lead to a massive surge in consumer spending which led many people to label it a failure and oftentimes to draw far-reaching conclusions about the nature of stimulus programs. As Martin Feldstein put it in The Wall Street Journal "The Tax Rebate Was a Flop. Obama's Stimulus Plan Won't Work Either." But there's new evidence from Christian Broda and Jonathan Parker that uses a more sophisticated design and shows the opposite — the Bush/Pelosi stimulus was effective and potentially offers a model for future politically tractable stimulus programs.
1. What's the evidence?
Broda and Parker are able to exploit the fact that the IRS lacked the institutional capacity to mail every single person a check simultaneously. That means we can compare the behavior of households that got a check early to the behavior of otherwise-similar households who hadn't yet gotten their check. Analysis of these spending patterns indicates that households spent 30 to 45 percent of their rebate checks during the quarter of disbursement and 20 to 30 percent more of it during the next quarter.
They also find that simply learning that the check was coming didn't influence spending behavior in a statistically significant way, and that the spending effect was much greater among low-income families than high-income ones.
The upshot is that standard economic models based on a forward looking "representative agent" are missing important relevant aspects of human behavior. In particular, the distribution of economic resources is important and putting extra cash into the hands of hard-pressed people results in more spending.
2. What doesn't the evidence tell us?
This is what's known as a partial equilibrium result, in other words a bit of an accounting exercise. We know that households spent the stimulus checks, and that put GDP up higher than it would have been had the checks not gone out but had everything else been exactly identical.
Yet had the checks not gone out, other things wouldn't have been identical. Perhaps the stimulus spending crowded out other private spending. Alternatively, perhaps the stimulus spending had a multiplier effect as the people who made money selling things to stimulus recipients went on to spend the money again.
3. Why is this important?
Despite the limits, this research tells us that one relatively simple class of stimulus program — send everyone money — does in fact produce the kind of behavioral response (more spending) that is desired.
It may strike you as funny that "mailing citizens money causes them to spend money" is an interesting result in empirical economics, but there is a strong theoretical tradition of denying that this is the case and many people seem to have felt that the 2008 experience vindicated that theory. Check-mailing stimulus, in turn, is interesting because even though it's nobody's favorite idea for boosting a depressed economy it's something that in the past was able to win bipartisan support. Conservatives fear that more spending on government programs will permanently increase the size of government, while liberals fear that tax cuts will permanently starve the government of revenue. Sending out checks, by contrast, manages to largely sidestep endemic ideological conflict.