The Corner

The one and only.

A Stimulus Retrospective


The election season has prompted Americans to review President Obama’s presidency, especially his signature “stimulus,” the American Recovery and Reinvestment Act (ARRA). Any honest assessment cannot conclude that its impact on the economy was zero. You simply can’t throw $800 billion at the economy and have nothing to show for it. The real question is of course, was its impact worth it? That is, the question is not could the stimulus affect the economy; it is should it have been used to try?

Let us admit at the outset that the actual effect the stimulus had on the economy is ultimately unknowable. Knowing for sure would require comparing the actual economic history with the one that would have prevailed in its absence — a counterfactual. We don’t know that and will never know for sure.

That leaves us with an exercise in educated guesswork, which is exactly what the Congressional Budget Office (CBO) does periodically when it comes up with its required estimates of the stimulus’s economic effect. News coverage of these reports typically focuses on the top-line estimate for increased employment owing to the stimulus. What gets lost in translation is the broad range of CBO estimates, that is, the considerable uncertainty in the underlying assumptions.

The CBO estimates hinge on the so-called multiplier effects of the policies included in the stimulus law. These vary from policy to policy, and across estimated multiplier effects. No two analysts would likely agree on the same multipliers. CBO plays it straight, providing both the low-bound estimates for these multipliers, as well as the high-bound estimates. All have some foundation in the economics literature, where there is anything but consensus. However, they do allow us to consider the policy choices in terms of trade-offs.

There is not a single major provision of the stimulus law for which the low-bound multiplier estimate was greater than 1. That is, CBO openly acknowledges that a dollar of stimulus expenditure could generate less than a dollar of output. Accordingly, it is entirely possible that the stimulus did impact the macroeconomy, but was not the right policy call — it may well on the whole not have been worth the expenditure in taxpayer dollars. And there is credible research to support the low-bound scenario. As Kevin Hassett noted in a February issue of National Review, a recent study by Valerie Ramey of the University of California, San Diego, suggests overall government spending is closer to having a multiplier of about 0.5.

At the other extreme, most of the major provisions under the high-bound scenario are associated with multipliers greater than 1, providing the possibility of a good bang for the buck.

We think it’s safe to assume the multiplier effect was about 1. It’s worth noting that this is below the roughly 1.4 that the administration used in their estimate that predicted the unemployment rate would be about 6 percent right now thanks to the stimulus.

A dollar for dollar return on the stimulus expenditure might suggest to some that the stimulus was the right call. Wrong. In the long run, even assuming this multiplier effect, we’re worse off from the stimulus. CBO makes just this point: “In contrast to its positive near-term macroeconomic effects, ARRA will reduce output slightly in the long run.” The reason for this is because the stimulus was debt-financed, and ultimately the additional debt will crowd out private capital, reduce the capital stock, and depress wages. So, we’re worse off in the long run than if ARRA had not been enacted. 

Remember as well the roughly $7.8 trillion in intervention by the Federal Reserve, an order of magnitude larger than the stimulus, and the Troubled Asset Relief Program (TARP). Supporters of the stimulus would argue that our economy was “going over a cliff” and the stimulus saved the day. But compared to the other measures taken, its absence would not have ended the world as we know it. This is even more valid when one considers its longer run implications.

The stimulus has been advertised as classic Keynesian fiscal policy, designed to smooth out troughs in the business cycle. But supporters of this policy approach often ignore the other aspect of this school of thought — to smooth out peaks as well. When the economy is doing poorly, a good Keynesian would support deficit-financed expenditures to mitigate a flagging economy. But in the boom years, a good Keynesian would forgo the fiscal gas pedal and apply some breaks. This is the moral equivalent of eating ice cream today, while committing to broccoli later. But that’s not what we’re seeing. Instead, despite running four years of $1-trillion-plus deficits, the stimulus’s chief supporter, President Obama, has proposed an average ten-year deficit of $668 billion. The president wouldn’t even pass Keynesian economics. Instead we get a ten-year ice cream party.

Unfortunately the consequences of the stimulus won’t stop in the short run. The CBO did not address the adverse long-run effect of the risk of a sovereign-debt crisis. With a national debt having surpassed our GDP, the U.S. is now in the range where other nations have experienced inhibited growth from large levels of indebtedness. This begins a vicious cycle — with slower growth comes flagging revenues, increased spending via transfer programs, and growing debt. Interest payments begin to rise and place policy-makers in an ever-tightening fiscal strait jacket. And then, with frightening rapidity, investors lose confidence. The implications of this scenario are plainly evident from the recent experience in Europe — but the economic fallout from a debt crisis in the world’s largest economy is difficult to fathom.

The debate over the stimulus bill will continue. We believe that the evidence argues against either extreme: Keynesian success or no impact. It also, however, suggests it was not a good idea.