Assessing Yet Another Round of the Structural Unemployment Arguments
No matter how much elites insist that our unemployment problem is structural, they don't have the data on their side.
David Brooks has the 2012 version of the structural unemployment argument in his editorial today, "The Structural Revolution." Here's rooting for this one, as the previous arguments haven't held up all that well.
The 2010 version of the argument had to do with an increase in JOLTS "job opening" data, data that turned out to be incorrectly estimated by the BLS (as we learned in 2011). The 2011 version focused either on the idea that the unemployed had bifuricated into a normal unemployment market and a long-term, zero-marginal productivity market (it hadn't) or that the "regulatory uncertainty" of the Obama administration was holding back the economy (which, as Larry Mishel found, wasn't backed by the data).
There's been a ton of situations where these structural unemployment arguments came charging down the runway only to hit a cement wall of data. One "oops" moment was Raghuram Rajan citing Erik Hurst in claiming that unemployment would be three points lower if it wasn't for "structural" reasons, and Hurst having to publicly point out his preliminary research said nothing of the sort. Another was Rajan arguing, in June of 2011, against monetary policy. Why? Because "one view is that corporate investment is held back by labor-market rigidities (wages are stubbornly too high)....There is, however, scant evidence that the real problem holding back investment is excessively high wages (many corporations reduced overtime and benefit contributions, and even cut wages during the recession)." Empirically that means that there shouldn't be any bunching of wage changes at the zero mark. Here's what the San Francisco Fed found early this year:
Apparently none of that changed anything for anyone. So what do we have now? I want to address three specific points in Brooks's essay which I think are wrong in a very useful way. First, Brooks argues that "Running up huge deficits without fixing the underlying structure will not restore growth." The argument here is that a larger deficit will not help with short-term growth. I'll outsource this to Josh Bivens, addressing a similar argument from Adam Davidson:
This is the reverse of the truth – there is wide agreement that debt-financed fiscal support in a depressed economy will lower unemployment. Now, it’s true that there are holdouts from this position. And others who think the benefits of lower unemployment are swamped by the downsides of higher public debt (they’re wrong, by the way). But, the agreement is much more widespread – ask literally any economic forecaster, in the public or private sector, that a casual reader of the Financial Times has heard of if, say, the Recovery Act boosted economic growth. They will all tell you “yes.”
You won’t find anywhere near such a consensus on long-run tax or education or health care policy. In fact, public finance economists can’t get unanimous agreement on if, in the long run, income accruing to holders of wealth should be taxed at all (it should, by the way). In short, anybody waiting for the current unpleasantness to pass and for economists to unite in harmony in future policy debates shouldn’t hold their breath...
Lastly, Davidson notes that there is a rump of economists (he calls them, reasonably enough, the Chicago School) that argue that debt-financed fiscal support cannot help economies recover from recessions. But, it’s important to note that there is pretty simple evidence that can be brought to bear on this Keynesian versus Chicago debate. Nobody denies, for example, that the government could borrow money and just hire lots of people – hence creating jobs. What the Chicago school argues is that this borrowing will raise interest rates (new demand for loans will increase their “price,” or interest rates) and this increase in interest rates will dampen private-sector demand. But interest rates have not risen at all since the Recovery Act was passed and private investment has risen, a lot.
Second, Brooks argues that "there are the structural issues surrounding the decline in human capital. The United States, once the world’s educational leader, is falling back in the pack." If this is the case -- that our problems are a lack of education and investment in human capital -- then recent college graduates would have significantly lower unemployment rates than most, or they would be the same, or if they were higher then they'd come down even faster. Also from EPI, Heidi Shierholz, Natalie Sabadish, and Hilary Wething, "The Class of 2012":
Young people with recent college degrees have high unemployment rates. That's not good, either for Brooks's argument or for the huge number of young people being devastated by the weak economy and the weak response of elites.
Third, we have Brooks arguing that there are issues "surrounding globalization and technological change. Hyperefficient globalized companies need fewer workers. As a result, unemployment rises, superstar salaries surge while lower-skilled wages stagnate, the middle gets hollowed out and inequality grows." Some occupations require high skills and have sufficient demand, but some occupations require mid-skills and are disappearing. (Low-skill jobs should be fine on unemployment, but low on wage growth, in most versions of this "job polarization" theory.)
Let's take BLS CPS unemployment data by occupation, March 2007 and March 2012, and see if you can tell me which occupations require these high-end skills from their low 2012 unemployment rates:
I'm having trouble seeing them in the data.
So here's the important thing about the demand-side recessions: If I wanted to come up with a "supply" theory for Brooks, I'd say, looking, at the data above, that we have too many college graduates and too many business and professional workers. I'd also say we have too many non-college graduates and too many service workers. I'd also say we have too many of all ages, all educations, and all occupations. Something is weak at a fundamental level in the economy, which is impacting everything, even before we get to the pressing issues related to job polarization or education. That weakness is demand, and that is where the policy response should be. Don't tackle it, and the longer-term problems are even harder to manage.
As David Beckworth noted, "[t]his evidence in conjunction with that of downward wage rigidity, excess money demand, and the Fed handling the housing recession just fine for two years should remove any doubt about there an aggregate demand problem. The real debate is how best to respond to this problem." The evidence he referred to was the SF data noted above along with the tracking he found between sales being reported as the "single most important problem" by small businesses and the unemployment rate:
Mike Konczal is a Fellow at the Roosevelt Institute.
Follow or contact the Rortybomb blog: