Don’t blame the robots: It’s not productivity growth that’s holding job growth back

The Wall Street Journal ran an article a couple of days ago implicitly arguing that accelerating productivity growth is a prime reason why labor market recovery from the Great Recession has been so sluggish. Another reporter asked me about it yesterday, so I figured I’d write up a couple of thoughts on it.

First, we should be clear that the pace of labor market recovery since the Great Recession has not been uniquely bad; since the trough of the recession, private sector employment growth has actually been exactly in line with the (admittedly too-slow) recoveries from the recessions of the early 1990s and 2000s. Overall employment growth has actually outperformed the recovery from the early 2000s recession. Figure A below shows the trends for private sector employment. Note that the jobs lost during the latest recession dwarf those lost during other recessions – but since the official recovery began, job growth has been on-par with recent recoveries. Note that policymakers should not be graded on this generous curve – it’s a disaster that we haven’t had a better recovery from that perspective. But one doesn’t need to generate new theories to explain this allegedly atypically bad recovery – it just hasn’t been atypically bad.

Second, and in line with Dean Baker’s response to the article, productivity growth has not been particularly fast since the Great Recession. Figure B below shows the behavior of productivity averaged over all recessions between 1947 and 1981, the average of the early 1990s and early 2000s recoveries, and growth since the Great Recession. So, again, one cannot argue that fast productivity growth presents unique challenges in the current recovery since its performance just hasn’t been all that unique.

Lastly, and maybe wonkiest, fast productivity growth doesn’t change the validity of Keynesian diagnoses of what the economy needs at all. In fact, it would just strengthen them. The root of the Keynesian diagnosis is that there is a large gap between aggregate demand and potential supply in the economy – or, a large “output gap.” Figure C below shows the problem – the large output gap between actual and potential GDP is the reason why we have such high unemployment today. Productivity growth just pulls the potential GDP curve upwards, which means, all else equal, that the output gap will rise (on the chart I illustrated this with the “actual, if productivity growth accelerated” line).

But, the obvious solution to this problem is simply to push up demand to make actual GDP equal potential GDP again. Basically, accelerating productivity growth would just make measures to boost demand more necessary, and would insure that no adverse supply-side response (say accelerating inflation or rising interest rates) would kick-in.

The root cause of today’s underperforming economy remains insufficient spending by households, businesses and governments to fully employ all those who want a job. And the cure for this is simply policy measures to boost spending. Yes, I’m sure this has gotten boring for many economy watchers who want newer and more exciting diagnoses and cures, but sometimes what’s true is pretty boring.

One quick thought on why explanations based on productivity growth can sound convincing: At any point over the past century you could have walked into a factory and been told about the big technological improvements that had been made over the past four years. If you’re a business writer who walks into a factory today looking for a root cause of the labor market’s doldrums, guess what? You’ll be told about the big technological improvements made over the past four years, and then you might think, “hey, that’s why the jobs aren’t here!” But, if you had walked into a factory in 2000 – when the unemployment rate reached 3.8 percent – You also would’ve been told about an amazing four-year run of technological advance. In the end, high rates of unemployment are about demand falling short of supply, period.