Why the Stimulus Failed

New research on what actually happened to a trillion dollars. WSJ, 09-08-11

Even zero jobs growth in August doesn't seem to have disrupted President Obama's faith in the economic policies of his first three years, so one theme we'll be listening for in tonight's speech is how he explains the current moment. Why did his first jobs plan—the $825 billion stimulus—so quickly result in the need for another jobs plan?

For readers who want to know, an important account is offered in a pair of new Mercatus Center working papers by the George Mason economists Garett Jones and Daniel Rothschild, who did field research on what they call the supply side of the stimulus.

The Keynesian theory was that a burst of new government spending would take up some of the slack in aggregate consumer demand. This was justified in 2008, again in 2009, and is still defended now based not on real-world observation but on abstract macroeconomic models that depend on the assumptions of the authors. The Congressional Budget Office's quarterly studies—often cited to claim the stimulus created tens of thousands of new jobs—are based on such a model. By informative contrast, Messrs. Jones and Rothschild interviewed actual people who received stimulus dollars and asked how they spent the money.

In the first paper, the authors survey 85 different businesses, nonprofits and local governments across the country and conclude that "As is often the case when economic models are transferred from the blackboard to actual public policy, there was a gap between theory and practice."

One of the major patterns Messrs. Jones and Rothschild uncovered was that the top-down stimulus was poorly targeted. In one redolent example, a federal contractor said he was told to use smaller, nonstandard tiles that are harder and more expensive to install in order to increase the cost of the project. That way, the government could claim the money was moving out the door faster. The famous Milton Friedman line about government ordering people to dig with spoons to employ more people comes to mind.

In another case study, a budget shortfall forced a mid-size city to lay off 185 public workers—but the city received a $4 million stimulus grant to improve municipal energy efficiency. The manager of a construction company received funds for "the last thing on our list; and truthfully, the least useful thing." It happened to be a crane and a forklift.

A carpenter forman surveys a construction site funded by federal stimulus in Lakewood, Colorado, in 2010.

The authors are careful to note that such anecdotes do not mean that all of the stimulus was a waste, and they did find some success stories. The problem is that all but the most reductionist Keynesians of the Paul Krugman school believe it matters what the government spends money on. A dollar that eventually will be taken out of the private economy through borrowing or higher taxes to fund pointlessly expensive projects—a la the tiny tiles—is not the way to nurture a recovery.

The second paper suggests that the stimulus did not "create or save" nearly as many jobs as the models indicate. On the basis of 1,300 interviews, Messrs. Jones and Rothschild estimate that merely 42.1% of the firms that received grants hired people who were unemployed. Instead, they poached workers from their competitors.

"This suggests just how hard it is for Keynesian job creation to work in a modern, expertise-based economy," they write. The stimulus "was implemented at a time when the Keynesian model had every chance of succeeding on its own terms. The high level of unemployment and the rapid deadline for spending created both the supply of workers and the demand for workers. If the job market results are so lackluster in this setting, economists should expect even weaker stimulative results during more modest recessions."

The lesson of such on-the-ground knowledge is that the stimulus was a lost opportunity. In practice it became a shotgun marriage between an economic theory justified by computer models and 40 years of liberal social priorities (clean energy, Medicaid expansions and the rest). This produced the 9.1% unemployment we now have.

The economy would have benefitted far more if the government had instead improved the incentives for people and businesses to invest, produce and grow. The President probably won't mention any of this, but it does explain why he has to give his latest speech.