What If Stimulus Advocates Were (Half) Right?

X
Story Stream
recent articles

In "The Job Impact of the American Recovery and Reinvestment Plan" dated January 9, 2009, Christina Romer and Jared Bernstein outlined the rationale for, and the expected results from, President Obama's "stimulus" program. Their report included the now-famous graph that warned that without action, unemployment would rise from 7.4% in December 2008 to 9.0% in mid-2010, after which it would begin to decline.

The graph also predicted that the application of $775 billion in stimulus would limit the jobless rate to a peak of 8.0%, which would be reached in the third quarter of 2009. With stimulus, by mid-2010 the unemployment rate would be at 7.2% and falling.

Congress increased the size of the stimulus program to $787 billion and passed it. We now know the results.

Rather than the predicted 7.2%, the unemployment rate in June 2010 was 9.5%. Worse, labor force participation fell to 64.7%, from the 65.8% level that existed at the time that Romer and Bernstein prepared their report. As a result, total employment in June 2010 (BLS Household Survey) was only 139.1 million, rather than the expected 145.1 million. This represents a shortfall of six million jobs from what the stimulus program was supposed to deliver.

It is obvious now that Romer and Bernstein were spectacularly wrong about what stimulus would produce. But what if they were right about what would have happened without the stimulus? After all, Romer and Bernstein are eminent economists, and it was the stimulus (especially at the scale proposed) that was the new and untested - and therefore unknown - element. Clearly, Romer and Bernstein had a much better foundation upon which to base their forecast of the "base case" than they did for the stimulus case.

If Romer and Bernstein were right about the base case, then, rather than creating jobs, stimulus actually destroyed them. By June 2010, stimulus was supposed to have "created or saved" 2.8 million jobs, on its way to a total of 3.7 million by the end of 2010. However, total employment in June 2010 was actually 3.2 million less than what Romer and Bernstein projected it would have been without stimulus.

It is possible to assess the efficacy of stimulus by looking at the experience of other nations during the current recession, which has been worldwide in scope. Examining twelve nations (Australia, Canada, France, Germany, Greece, Ireland, Italy, Japan, Portugal, Spain, the U.K., and the U.S.), it is possible to compare the percentage change in total employment of each country from the end of 2007 to the end of 2009 with the amount of fiscal stimulus applied. For the purposes of this analysis, the "net stimulus" of each nation is considered to be the sum of its 2008 and 2009 budget deficits minus twice its 2007 deficit, all expressed as a percent of GDP. This approach is designed to show the effectiveness of increasing a country's fiscal deficit as a tool for countering rising unemployment.

The twelve-country comparison suggests that stimulus is actually counterproductive. Ranking the nations in ascending order of net stimulus, the six lowest-stimulus countries applied an average of 5.4 GDP percentage points of net stimulus and saw their total employment decline by an average of 0.6% over the two-year period. In contrast, the six highest-stimulus nations applied an average of 13.7 GDP percentage points of net stimulus while seeing their total jobs go down by an average of 5.21%. The three countries with the highest amount of net stimulus (the U.S., Spain, and Ireland) also had the highest percentage employment losses.

Despite all of the academic work on "multipliers", it is clear that stimulus can never produce an increase in total demand. This is because the Treasury must sell bonds and take money in before the same money can be sent out in the form of government spending or tax rebates. However, it is somewhat surprising how destructive to employment stimulus appears to be. Based upon the evidence, it is reasonable to conclude that Romer and Bernstein were right about the base case in early 2009, and that Obama's stimulus program has reduced total employment by 3.2 million jobs from where it would be right now if we had not passed the stimulus bill.

It makes sense that stimulus (government borrowing and spending) would destroy jobs. Government bonds are bought with investment dollars. If there is no change in incentives to call forth more output, then if the government sells more bonds, other forms of capital must be liquidated in order to buy them. Without tools (capital), workers can produce nothing and their jobs disappear.

According to Recovery.gov, as of June 30, 2010, about 53% of the stimulus funds had been disbursed. This required the sale of an incremental $420 billion in government bonds. Right now, U.S. private non-residential produced assets per private-sector worker are about $160,000. Accordingly, $420 billion in stimulus bond sales could be expected to force the liquidation (or prevent the creation) of 2.6 million average private-sector jobs. However, given the fact that this recession has impacted lower-paid jobs more severely than higher-paid positions, it is plausible that taking $420 billion of capital out of the private sector could account for the entire employment shortfall (3.2 million jobs) from the Romer-Bernstein base case. This would be the case if the average job destroyed by stimulus were supported by $130,000 of private non-residential produced assets.

If this relationship continues to hold, the bond sales required to fund the remaining $367 billion of stimulus could be expected to destroy (or prevent the creation of) an additional 2.8 million jobs between now and the end of 2011. Added to the 3.2 million jobs lost to stimulus already, this implies that total employment at the end of 2011 would be 6.0 million lower than the Romer-Bernstein base case. If this were to occur, and the labor force participation rate were to remain at the depressed level of June 2010, the reported unemployment rate in December, 2011 would be 9.5%, the same as that of June, 2010.

It will be interesting to see how the employment situation plays out if the current economic policies (including stimulus) are allowed to continue for another 18 months.

While the administration, most economists, and the mainstream media still seem to believe in stimulus, opinion polls show that the vast majority of Americans do not believe that stimulus creates jobs. They appear to be right.

 

Louis Woodhill (louis@woodhill.com), an engineer and software entrepreneur, and a RealClearMarkets contributor.  

 

Comment
Show commentsHide Comments

Related Articles