Getting the numbers right on unemployment

In an earlier post on growth in Medicaid as a percent of GDP one reader suggested that the trend line in unemployment is down.   And indeed since mid-2009 the line has been generally down (the graph here is from the Bureau of Labor Statistics).   My reader's point was really not focused on the post about Medicaid but it is important to respond to it none-the-less.

Employment numbers are complicated.  The unemployment rate connotes a figure derived from dividing the number of all unemployed individuals by the number of people in the workforce.   Seems simple but that number does not include those people who either voluntarily or involuntarily have taken themselves out of the workforce.  It is also a crude measure and thus does not differentiate between people who are fully employed and those that have, for reasons of the current macro situation, chosen to take a job below their skill level.  

There are many ways to compute the true unemployment rate in the country.  One (presented from a Financial Times graph) is to look at the of the US population that is employed.  And for this recession, even after the NBER announced that the recession had ended, the recovery has been described by all stripes of economists as a "jobless" recovery.  Some economists like Paul Krugman suggest that has happened because there was not enough of an intervention in the economic from stimulus packages.   Others argue (and I would agree) that part of the problem came from not too little government intervention but too much.  Regardless, the numbers are what they are - this recovery has had a puny recovery in the number of Americans employed.  That is a problem of significance.