Observations on the U.S. Labor Market

One of these is not like the others – local government employment stabilizes, but state does not – structural unemployment decreases – Obamacare does not cause the rise in part time employment

First, a reminder for those with short memories, part of the reason why the current recovery is so lackluster in terms of employment growth, even relative to the preceding jobless recoveries.

Figure 1: Log government employment relative to 2009M06 (blue), relative to 2001M11 (red), relative to 1991M03 (green), seasonally adjusted. Source: BLS via FRED, author’s calculations.In contrast to previous recoveries, government employment in this recovery is declining.

Second, the breakdown in government employment losses is interesting. It has been concentrated mostly at the local level.

Figure 2: Government employment relative to 2009M06 at state (blue), local (red), and Federal (green) levels, in 000’s, seasonally adjusted, 2009M06-2013M08. Excludes temporary Census workers. Source: BLS via FRED, and author’s calculations.Notably, of the decline of 512 thousand in local government, 318.3 thousand has been accounted for by the education sector. I suppose that’s okay ‘cause our students got ‘nuff learnin’ (we don’t need no stinkin’ simultaneous equations!).

Third, consider the question: How much of unemployment is structural? Prakash Loungani hasupdated his earlier estimates based upon information from the sectoral dispersion in stock market returns.

Figure from Prakash Loungani (Sept. 2013).He writes:

The estimate of structural unemployment has declined over the past year in line with the decline in the actual unemployment rate. In the figure below, the solid (black) line shows the U.S. unemployment rate declining from nearly 10 percent in 2009:Q4 to about 7 ½ percent today. The dotted (red) line is the estimate of structural unemployment; it too has declined over that time and the latest estimate of structural unemployment is 6.2 percent. There is a lot of concern about U.S. long-term unemployment. In this case too, there has been a decline in the estimate of the structural component of long-term unemployment, but it is more gradual than in the case of total unemployment.

If there was an argument for more heavily weighting supply side measures more heavily than demand side measures (on which count I’m dubious), those arguments have dissipated.

Fourth, Obamacare and the rise in part time employment. From Furman and Stevenson writing in CEABlog

Over the 41 months since the Affordable Care Act was signed into law, 91 percent of the increase in employment is due to full-time work. This pattern has roughly held over the past twelve months, with 86 percent of the increase in employment due to full-time work.

The authors also quote a FRB SF study: “We have shown that part-time work is not unusually high relative to levels observed in the past, most notably in the aftermath of the early 1980s recession.”

They conclude:

There is no systematic evidence that employers are shifting employees to just below 30 hours per week. CEA has analyzed the Current Population Survey to estimate worker hours in more narrow ranges than those published by the BLS. This analysis indicates that there has been no noticeable shift in employment just above or below the 30 hours per week threshold, a conclusion that has also been corroborated by outside analysts.

This figure illustrates their assertion:

Figure from CEA (Sept. 2013).Hence, while the Obamacare-causes-part-time-employment-increases meme dominates in certain circles (I usually see it in listings of anecdotes, and polls), I don’t think the data is determinative on this count, and if there is an effect, it so far is not detectable. Over the longer term, it may go the way of the “QE and hyperinflation” meme. [1]

This piece is cross-posted from Econbrowser with permission.