New evidence that the Obama stimulus was a bust

The debate in the popular press and on blogs about the efficacy of the 2009 stimulus bill generally has been superficial on both sides. Those who tout the bill point to a study by liberal economists Alan Blinder and Mark Zandi purporting to show that the fiscal stimulus enacted under Presidents Bush and Obama lowered the unemployment rate by approximately 1.5 percentage points. However, as Lawrence Lindsey points out in the Weekly Standard, this study assumed that the recent stimulus has been roughly as effective, dollar for dollar, as past programs. It then multiplied past measures of job creation by the number of dollars spent this time around.
This simplistic methodology assumes, rather than measures, the efficacy of the most recent incarnation of stimulus. It thus begs the question it purports to answer.
Moreover, there are reasons why it should not be assumed that the Obama stimulus has been as effective as prior ones. Larry Summers has said that stimulus should be timely, targeted, and temporary. The Obama stimulus was not timely because, as Lindsey observes, most of the funds were distributed through the cumbersome government contracting process and then, in many cases, were channeled through state and local governments. It was not targeted, at least not in the proper sense, because the money was allocated according to the preferences, not of economists, but of congressional Democrats who based their decisions on political considerations including the desire to reward favored interests.
The most popular argument that the stimulus flopped is also specious. It relies on a paper by Christina Romer and Jared Bernstein that predicted that if the Obama stimulus package passed, the unemployment rate would peak in the middle of 2009 at 8 percent and would be at 7.5 percent by now. Because this has not happened, and because the unemployment rate also exceeds that which Romer and Bernstein said would obtain without stimulus, some argue that the package therefore failed.
The main problem with this argument is that Romer and Bernstein underestimated the severity of the recession. In fact, unemployment was already higher in the first quarter of 2009, when the stimulus was passed, than the 8 percent peak they assumed. Thus, their estimated numbers plainly cannot serve as a basis for judging the impact of the stimulus.
Lindsey proposes to correct for this problem by raising the starting point used by Romer and Bernstein to correspond to the actual unemployment figures in the first quarter of 2009 – 8.2 percent. From there, he uses the path they projected. Alternatively, to give defenders of the stimulus the benefit of any doubt, he uses the average unemployment figures from the second quarter of 2009 – 9.3 percent. The first approach fully compensates for the standard objection to using the Romer-Bernstein paper to judge the stimulus. The second approach overcompensates for that concern.
Lindsey finds that, under both alternatives, the unemployment rate with the stimulus significantly exceeds the corrected Romer-Bernstein estimates. With the correction (alternative 1), the unemployment rate for the second quarter of 2010 with the stimulus should be around 8 percent; with the over-correction (alternative 2), it should be at 9 percent. In reality, though, the unemployment rate was nearly 10 percent.
Romer and Bernstein also estimated what the unemployment rate would be without the stimulus. Lindsey found that, with the correction, their model predicts an unemployment rate without the stimulus that is essentially identical to the actual unemployment rate in both the first and second quarters of 2010. In the last two quarters of 2009, the actual unemployment rate exceeded the rate of unemployment that the corrected model predicted would occur with no stimulus, and was equal to the rate predicted by the overcorrected model.
Lindsey’s conclusion – which seems unassailable – is that after correcting for the unduly low starting point used by Romer and Bernstein, the actual performance of the economy with respect to jobs is almost exactly what they said would it would be if we had done nothing, instead of passing the $800 billion stimulus package. This strongly suggests that the stimulus package enacted by the Democrats was a bust.

Responses