Counting the Jobs Produced by the Stimulus

When the stimulus package was enacted last winter, the administration said its goal was to create or save 3½ million jobs by the end of next year. How closely has the administration come to achieving that goal? A couple of weeks ago the White House issued an interim report on jobs directly created or saved as a result of one part of the stimulus package, the grants or contracts directly made by the federal government or indirectly provided through federal aid to state and local governments. The report has been subject to minor carping and major criticism. The headline number in the report was based on reports by companies and government entities receiving funds under the stimulus program. A number of reporters have found fault with these reports. Many of the specific criticisms are no doubt valid. On the whole, however, the headline number in the report – 640,000 jobs created or saved – understates the net effect of the stimulus program on overall employment. The reason is the indirect effects of the stimulus on consumer spending and employment, which are not counted in the headline number.

“Data” is the plural of “anecdote.” Most of us are much more comfortable with anecdotes than we are with careful data analysis. It is therefore perfectly understandable when newspapers and television feature anecdotes rather than statistical analysis to describe the state or trend of the economy. 

Unfortunately, almost every anecdote is open to competing interpretations. This was obvious when the White House released its summary of 150,000 reports submitted by recipients of federal grants or contracts under the 2009 stimulus program. The reports were collected to help policymakers and taxpayers keep track of the employment effects of stimulus spending.  According to the Wall Street Journal, a Kentucky shoe store owner gave credit to the stimulus program for creating or saving 9 jobs as a result of an $890 contract to supply work boots to the Corps of Engineers. While the Journal treated this claim with appropriate skepticism, one can imagine business owners who are persuaded to remain in operation as a result of a modest order that spells the difference between insolvency and a tiny profit. 

In essence, the reports distilled by the White House provided evidence from 150,000 anecdotes.  According to the Administration’s summary, the reports offered evidence that 640,000 jobs have been directly created or saved as a result of spending or promised spending under grants or contracts funded by the stimulus package. Jared Bernstein, the Vice President’s chief economist, emphasized that the 640,000 count represents an incomplete tally of the total jobs added or saved as a result of the stimulus package. It ignores, for example, the jobs created or saved as a result of personal tax cuts or hikes in unemployment compensation checks. We cannot collect anecdotes from Walmart, Safeway, or Disney World telling us how many jobs have been produced by higher consumer spending induced by the stimulus package. When a customer buys groceries or takes a vacation in Florida, how can we tell what percentage of the spending was induced by a tax cut or a bigger unemployment check? We must rely on elaborate, less transparent data analysis to uncover the indirect effects of the stimulus package. When the indirect effects are included, White House economists estimate that over a million jobs have so far been added or saved as a result of the stimulus.

The Wall Street Journal suggests that the White House estimate of 640,000 jobs directly saved or created may overstate direct job creation by 20,000 positions. Even if the Journal’s estimate is correct, the difference represents less than 2% of the total number of jobs directly or indirectly saved and created by the stimulus. Economists do not agree on how to estimate all the direct and indirect employment effects, of course. Most of us would agree, however, that the total job effects are likely to be greater than the 640,000 jobs identified in Recovery.gov. The current employment situation is certainly grim, and it has gotten much worse since the President took office. The job market is significantly healthier, however, than it would have been if the stimulus package had not passed.

Unless the labor market deteriorates much further, I am pessimistic about the political prospects for another major stimulus package. The Administration’s opponents have been successful in sowing doubts about the wisdom of the last stimulus. According to a recent CBS/New York Times poll, a clear majority of Americans now believes the stimulus package has either made the economy worse or had no beneficial impact. Only about a third of respondents thinks the stimulus has produced economic gains so far. 

In this political environment it is unlikely Congress will pass a major new stimulus package anytime soon. What is more likely – indeed, what is essential – is the continuation of stimulus programs that are currently scheduled to expire.  Last week the House and Senate extended unemployment protection for workers who have lost jobs in the current recession. These protections ought to be extended until the job market improves significantly and it becomes easier for laid off workers to find jobs. If unemployment is likely to remain over 9% for an extended time, there is a compelling case for additional public infrastructure investment. Given high unemployment in the construction and capital goods industries and federal borrowing costs that remain near a post-war low, it makes sense to invest in public capital projects over the next few years. If the federal government does not have adequate plans for such investments, it should start making them soon.