Friday, May 14, 2010

Male Employment-Population Ratios Recovering

This morning, I checked in on the most recent BLS data release.  I'm particularly interested in the long term pressure on male employment in the US from the triple pressures of automation, globalization, and immigration.  The overall situation is summarized in this graph of the male employment-population ratio for men ages 25-54 from 1948 to the present:

As you can see, it goes down with each recession, and then doesn't fully recover after.  It declined particularly badly with the great recession recently.

In recent months, it is starting to recover, however, as this next graph shows.  This is the same data, but just for the most recent cycle (since 2003).


I wanted to look at it broken out by education level, but to do that, we have three problems: a) the data can only be had for ages 25 and up (ie introducing some confound from retired men), b) the data only start in 1992, and c) the data are not seasonally adjusted.  For example, here is the data for high school graduates:
Problems a) and b) there's nothing to be done about, but c) I can fix, using the same seasonal adjustment method as in this post.  Therefore, here are the seasonally adjusted data for employment population ratio, men ages 25+, broken out by highest educational level:


As you can see once we have seasonal adjustment, the employment population ratio has now started to recover for all educational groups.  The recovery is only just beginning however, and if history is any guide, it won't recover to the level prior to the recession (although the high-school dropouts are an exception for some reason that I don't understand).

No comments: