The Problem with Income Statistics

There's been some recent data on which both the left and the right seem to be concentrating.   It seems that this economic recovery has not been as robust as anyone thought it would be.   In some stats wages have actually declined.   The numbers have been very harsh on workers without a high school diploma but even not very friendly to college graduates.   In January, for example, total income (according to the Feds who keep those numbers) income declined month to month by 3.6%.  Real wages for many income groups have declined.

Some part of the left (like former Enron advisor Paul Krugman for example or Robert Reich) believe that incomes have declined because there is not enough government.   They continue to subscribe to the "public squalor" argument first advanced by John Kenneth Galbraith (if we only had more government we would be better off).   Some of the right (for example, John Taylor and Paul Ryan) sees the slow income growth as a function of too much government.   If we could just reduce the government's hand in the economy and reduce the level of debt, incomes and economic growth would be kick started.

The chart at the left is one from Catherine Mulbrandon's wonderful resource on income statistics.  It shows job growth by sector by income between 2000 and 2011.  So, for example,  during the period the number of government jobs in the economy grew and their average income was a bit over $60,000.  Health care and social assistance grew significantly but their median income was a bit lower.

Mulbrandon's chart, which was published widely this week in places like Wonkblog got me to think about the issue.   And while I generally subscribe to the arguments from people like Taylor and Ryan - there is a bit more going on.     As Mitra Toosi (of the Bureau of Labor Statistics) has suggested the workforce is composed of three groups (at least as it applies to this question) entrants (those starting to work), stayers (those continuing) and leavers (those retiring or leaving the workforce for some other reason).    The mix of those could have a profound effect on aggregate income statistics.   The chart above is from a paper by Toosi which projects workforce participation to 2050 by age.   Note that even then there will be some remnants of the boomers still in the workforce.  But the dynamics of the three groups may influence the aggregates.

Let's try an example to illustrate this.   I retired from my CEO position at the end of 2011.    My successor was hired with an income that was about a third lower than the final level of compensation I received.  Over time her compensation will grow to match or exceed what I achieved (assuming she does well - and I make that assumption).  At the same time, while I retired comfortably, my income has declined from when I was working full time.   If we had a two person economy, aggregate income would have declined.   But neither of us is worse off.

All this is not to denigrate the arguments for reducing deficits and for putting the economy on a sounder path to economic growth.   But it is to say that when you hear some politician or pundit making a point about income statistics, take the words with a lot of caution.