Tuesday, March 28, 2017

Variations on the Phillips curve: labor force participation and wage growth

 - by New Deal democrat

Over the last month or so, in a few posts I have looked at the relationship between labor force participation, wages, and unemployment.  Last week I looked at several variations on the Phillips Curve -- the proposed relationship between inflation and the unemployment rate. While over a single business expansion the relationship seems to work, i.e., lower unemployment rates correlate with higher inflation, that hasn't been true over a longer term secular basis, and it specifically reverses during and after severe recessions, where higher (but declining) unemployment rates have been correlated with higher (and declining) inflation. 

But a much tighter relationship appears to exist between the labor force participation rate (the total employed plus unemployed as a share of population) and wage growth:

We have two secular regimes where higher participation is correlated with higher wage growth separated by a brief transition where higher participation was correlated with a sharp decline in wage growth.

Let's break it down.  First, here are the inflationary 1960s and 1970s, where there was also a great deal of labor bargaining power due to strong unions:

Here is the low inflation late 1980s to the present, where there has been very little labor bargaining power:

In both of these cases, for a total of almost 45 of the last 50+ years, higher wage growth has been correlated with higher labor force participation.

Here is the brief transition period during the 1980s Reagan Administration, where both inflation and labor bargaining power sharply declined:

The bottom line is that, once we take into account labor bargaining power, there appears to be a very good and durable relationship between changes in prime age labor force participation and 
growth in wages. 

But of course, correlation is not causation, and I have suggested in prior posts that if anything, wage growth may lag labor force participation, with some complex mutual causation. I will wrap this thought process up in one final post later this week.