28 May 2016

Secular Stagnation

Proponents of secular stagnation love to show the following graph
as evidence that the US labor market has hardly recovered from the last recession, making the last eight years a so-called 'jobless recovery.' Explanations for this vary, but two seem to be most popular:
  1. There is a savings glut in the United States, meaning that the nominal interest rate that is required for savings/investment equilibrium is significantly negative, resulting in a persistent aggregate demand shortfall.
  2. Recessions have negative effects on potential output/employment, making complete recover to full employment unlikely without significant government intervention.
The narrative seems to make sense: to the extent that the employment to population ratio is indicative of the state of the American labor market, the recovery has, to a large extent, been 'jobless,' and real GDP remains far below it's 1990 to 2007 trend:
 Furthermore, the real federal funds rate remains below normal levels, a prediction in conjunction with number 1 above:
I'm not entirely convinced of this case, though. For one thing, the employment to population ratio is hardly a good indicator of the labor market -- employment simply should not revert to the same proportion of the population irrespective of demographic concerns. Part of the reason economic growth was relatively rapid in the 1980's is that there was a secular increase in the employment to population ratio driven by women entering the labor force. Similarly, part of the reason post-2007 growth has been relatively slow is because of a secular decline in the employment to population ratio driven by an aging population:
Since the NBER recession trough in June 2009, the employment to working age population ratio has recovered a lot more than the employment to population ratio, and this is with data that only goes until March of 2015. Also, broader measures of unemployment that include those 'marginally attached' to the labor force -- people who have expressed a desire for work in the last 12 months -- as well as people who are currently part time but would like to be full time show a much more robust recovery in employment than the employment to population ratio suggests.
 The idea that employment remains significantly depressed in the United States is dubious. Yes, there has been a significant decline in employment levels that has not recovered, but this decline is most likely secular and, while there is still room for improvement in the labor market, it is clear that the employment to population ratio vastly exaggerates how much room there actually is.

What about real GDP? Hasn't the decline there been significant and persistent? Yes, but most of that can be explained by the largely demographically driven slow growth of employment since 2008, leaving the rest to be explained by low productivity growth. Does it makes sense to suggest that recessions caused by drops in aggregate demand have negative effects on productivity?

Unfortunately the theory in this area is rather lacking; while there are endogenous growth models, their real ability to explain TFP growth remains extremely dubious in my opinion -- when the causal link between, e.g., R&D employment and TFP is just assumed, I don't see there being much of a possibility for success. That being said, it certainly seems reasonable that long spells of detachment from employment result in the eroding of skills, for example, so the argument remains undecided in my opinion.

But I digress, the most important part of secular stagnation is the notion that the labor market has significantly more slack than it actually does, and thus it seems that the likelihood of a persistent decline in employment is small.

No comments:

Post a Comment