Data for May and June suggest that the rebate payments provided by the Economic Stimulus Act of 2008 may not yet have provided much of a boost to consumption. The increase in personal saving in May and June approximately matched the size of the rebate checks in those months.For those keeping score: that appears roughly consistent with the permanent income hypothesis, or perhaps Ricardian equivalence, but not the Keynesian consumption function. One caveat: as Shapiro and Slemrod note, we can't assume savings rates would have remained constant. Or, well, economists can assume whatever we want of course, but its not necessarily correct - negative shocks to household wealth (housing market) and expected future income (sluggish economy), if unanticipated, would have led people to save more. So the stimulus "worked" to the extent consumption spending was higher than it otherwise would have been, and to the extent this had a multiplier effect. Therefore, we need a structural model to estimate the counter-factual (i.e., the Lucas critique applies). In any case, as Shapiro and Slemrod note (and some argued at the time), one way to make sure the money is spent is for the government to spend it directly.
Here are the official numbers from the Bureau of Economic Analysis. Personal saving was $45.6 billion in May, compared to $48.1 billion in stimulus payments in May. Personal saving in June was $23.0 billion, compared to $27.9 stimulus payments in June. In stark contrast, personal saving averaged only $2.9 billion per month during the first four months of the year.
The personal saving rate tells the same story. After running under 0.5% during the first four months of 2008, it jumped to 4.9% in May and 2.5% in June. The change in the personal saving rate corresponds closely to the size of the rebate as a percentage of disposable income. The figure shows how most of the rebate payments appear to have gone straight into saving.
The fact that personal saving jumped by nearly as much as the increase in stimulus payments suggests that most of the rebate checks were saved, at least temporarily, but does not establish it definitively because we cannot be sure what the level of personal consumption expenditures would have been without the rebate checks. If personal consumption expenditures would have collapsed absent the stimulus payments while personal disposable income would have held steady, then the personal saving rate would have spiked up in May and June in any event, and its sharp rise when the rebates arrived does not then indicate that most of the rebate payments were saved.That most of the rebate checks were saved is, though, consistent with the results we find using the University of Michigan Survey of Consumers. When consumers were asked whether their stimulus check would lead them to “mostly spend, mostly save, or mostly pay down debt,” only 18% answered that it would lead them to mostly spend more.
Andrew Samwick notes:
So, in the cold light of day, what happened with the rebates was that the government issued more debt than it would otherwise have issued, and consumers now hold less debt than they otherwise would have. That looks like a shifting of liabilities from individual balance sheets to the government balance sheet. In other words, it was a bailout of consumer debt.Of course, our government's liabilities are our liabilities, but they carry a lower rate of interest than household debt, so we are saving on interest and since we are net borrowers from the rest of the world, that does represent a net national saving, rather than just a redistribution (its sort of like one of those "loan consolidations" where you combine your bills into "one low monthly payment"). There is a redistributional impact, though, to the extent that the future tax code is progressive, a larger share of debt will be paid by higher-income households.
In the Journal a few days ago, Martin Feldstein opined that the policy (which he had supported) was a flop. That generated a number of responses, nicely rounded up and added to by Mark Thoma.