Thursday, November 12, 2009

Impact of ARRA on US unemployment

The Obama administration's official announcement that the spending measures of the $787 bn fiscal stimulus (or ARRA) has so far created or saved just over 640,000 jobs has triggered off an intense debate. Critics have questioned the veracity of the figures, especially those relating to jobs saved, and point to the rising unemployment figures to rubbish these claims.



Supporters though argue that since the figures do no reflect measures such as tax cuts, boosted unemployment benefits or jobs created indirectly by stimulus spending, the actual numbers may be closer to one million. They also argue that the stimulus saved the country from slipping into a Depression and contributed to the second quarterly growth of 3.5% for this fiscal year. As Jeff Frankel writes, the biggest problem is with estimating the jobs saved since "one cannot estimate accurately, let alone prove, what would have happened in the absence of the stimulus package". Paul Krugman has this evidence that the ARRA is working.

The extent of the recession has been so deep that initial estimates and forecasts of the stimulus support required have been wide off the mark. The most striking indicator of this is the steep divergence between the estimates of unemployment with and without stimulus spending and the actual unemployment figures due to the mistaken assumptions behind the baseline figures. Greg Mankiw has been continuously updating on the unemployment forecasts made by Christina Romer and Jared Bernstein based on the estimated impact of the $787 bn stimulus measures with the actual unemployment figures.



The depth of the recession, already the longest since the war, and the devastation caused to the credit markets has meant that a long drawn out, at best U-shaped or at worst an L-shaped, recovery is inevitable. In fact, in anticipation of continuing weakness, there have been mounting calls for a second round of stimulus spending. In fact the NYT has gone on to assert that "without another round of effective stimulus, the worst recession in modern memory will likely become — at best — the weakest recovery in modern memory".

Already many of the stimulus spending measures which have run out, like the extension for unemployment insurance etc, have been extended for some more time. Given the persisting weakness in aggregate demand due to the damage suffered by households and businesses, it has been estimated that the output gap is is substantial and is likely to persist well into the next few years. Unemployment rates could remain high for many years to come.



Fundamentally, unemployment has always been a lagging indicator and therefore in previous recessions too the unemployment figures have continued to rise well into the economic recovery period. And it may be no different (or even worse, if Paul Krugman is to be believed) this time round too.



Mark Thoma has an excellent post in his new blog, which feels that the US unemployment rate could have a higher rate of normal unemployment (from its present normal of around 4%), in view of the possibility of high structural unemployment, even after normalcy is restored. This would be due to the structural changes that have taken as a result of globalization and the recent crisis - reduction in resources employed in housing, finance, and automobile production, higher savings rate and correspondingly lower consumption expected from households in the future etc.

Nouriel Roubini feels that it is most likely that the unemployment rate will peak close to 11% and will remain at a very high level for two years or more. He claims that the weakness in labor markets and the sharp fall in labor income will ensure a weak recovery of private consumption and an anemic recovery of the economy, and increases the risk of a double dip recession.

Update 1
Paul Krugman points to the not-so frightening US public debt as a share of GDP, Taylor Rule and yields on TIPS to argue that inflation may not be an immediate danger. Larry Summers too feels not much cause for alarm with US public debt.

Update 2
Mark Thoma explains why unemployment lags output in recoveries - businesses waiting to see whether the recovery is for real; firms bring back their on-the roll, and not laid off, slack employees back to work at the first signs of recovery (labor hoarding); having discovered how to reorganize to increase productivity, the demand for labor on the upside will be smaller than the amount lost on the downside thereby causing sluggishness in the recovery of labor etc

Update 3
CBO's estimates of the impact of ARRA on employment and economic output as on September 2009 is available here (pdf here).

Update 4
See Paul Krugman on the impact of the stimulus.

Update 5
Menzie Chinn on the importance of structural unemployment in the current recession.

Update 6
NYT examines the debate on the impact of ARRA on output and employment, and feels that ARRA contributed towards mitigating the effects of the recession and if anything may have been designed on assumptions (with the stimulus, the jobless rate was estimated by the Obama administraiton to peak at 8.1 percent) which were overly optimistic.



John Taylor though expresses doubts about these claims.

Update 7
Economic research firms IHS Global Insight, Macroeconomic Advisers and Moody’s Economy.com all estimate that the ARRA has added 1.6 million to 1.8 million jobs so far and that its ultimate impact will be roughly 2.5 million jobs, an estimate the CBO considers conservative.

David Leonhardt writes that aid to states and cities may be the single most effective form of stimulus and calls for another round of spending focussed on that. Unlike road- or bridge-building, it can happen in a matter of weeks and unlike tax cuts, state and local aid never languishes in a household’s savings account. He also proposes extended jobless benefits, which also tend to be spent, and tax credits carefully drafted to get businesses to hire and households to spend, like the cash-for-clunkers program.



Update 8
The CBO estimates that in the last quarter of 2009, the stimulus "added between 1.0 million and 2.1 million to the number of workers employed in the United States, and it increased the number of full-time-equivalent jobs by between 1.4 million and 3.0 million" (people who had been working part time but became full time, or otherwise increased their hours). It also estimates that the inflation-adjusted gross domestic product was most likely 1.5 to 3.5 percent higher in the fourth quarter than it would have been without the stimulus.

Update 9 (10/3/2010)
Mark Thoma has this post which argues that the Fed could not have done much more to stimulate employment. He feels that that "further easing by the Fed may not have much additional effect on long-term real interest rates, that even if rates could be brought down, consumers and businesses would be unlikely to respond by increasing investment and the consumption of durables - firms already have considerable idle capacity, so why build more, and consumers are pessimistic about their futures, so why buy on credit - and that there is an inflation risk from further easing".

Update 10 (16/3/2010)
Econbrowser points to a WSJ survey of forecasters whose results indicate that instead of the 0.15% growth rate recorded in 09Q4 y/y growth, the growth rate would have been -0.93%. For 2010Q4 Q4/Q4 growth, they forecast 3% growth, and in the absence of the ARRA, they would have predicted 2.2% growth.

Update 11 (25/3/2010)
Ray Fair from the Cowles Foundation has this estimate of the macroeconomic impact of the US stimulus bill passed in February 2009. It finds that both the real output and employment increased due to the stimulus bill.

Update 12 (29/5/2010)
The US CBO has released its latest assessment of America's 2009 fiscal stimulus, for the first quarter of 2010. It estimates that the in the first quarter of calendar year 2010, ARRA’s policies raised the level of real (inflation-adjusted) GDP by between 1.7-4.2%; lowered the unemployment rate by between 0.7-1.5 percentage points; increased the number of people employed by between 1.2-2.8 million; and increased the number of full-time-equivalent jobs by 1.8-4.1 million compared with what those amounts would have been otherwise. It also estimates that the effects of ARRA on output and employment are expected to increase further during calendar year 2010 but then diminish in 2011 and fade away by the end of 2012.

Update 13 (12/6/2010)
See David Leonhardt and Edward Glaeser on the impact of AARA.

Update 14 (1/7/2010)

CBO assessment of the jobs created by ARRA



See also this on the broad market impact of the stimulus.

Update 15 (27/7/2010)

David Leonhardt points to the fact that cash-strapped local and state governments in the US did not have to lay-off teachers, policemen, and other local government staff in large numbers across the country as proof of the effect of the stimulus (transfers to these governments were a major component of ARRA). Most major analysis of ARRA estimate that the stimulus is responsible for something like 2.5 million jobs that would not otherwise exist today.

No comments: