UPDATE: The report on job polarization in Oregon has been released, click the link for more information including a set of slides.
Lately there has been an amazing amount of pixels spent on describing the quality and pay of the job growth. Most recently both the Wall Street Journal and the Federal Reserve Bank of Atlanta compare industry sectors by average pay and lament the fact that most of the jobs are in lower-paid industries. I want to push back on these conclusions a little bit and also expand upon them. In short, yes, the U.S. economy is adding a large number of low-paying jobs, however we are also seeing relatively strong growth at the top end of the employment scale as well. We are missing the growth in jobs at the middle of the income distribution. If this all sounds familiar, it is. This is the continuation of the job polarization process analyzed, described and researched extensively by MIT’s David Autor. I am borrowing from his work (along with Acemoglu), the great work the Federal Reserve Bank of New York has done and also Jaimovich and Siu. I have a forthcoming research paper on job polarization in Oregon and will presenting on the subject at Tim Duy’s Oregon Economic Forum in October. What follows is a very brief overview of some of these trends at the national level.
First, while I do not completely disagree with the articles being written recently, I do want to point out that the fundamental changes occurring in the labor market are best seen at the occupational level not the industry level. When discussing cognitive or routine or manual type functions and how technological change is impacting them, these generally cut across industries by occupation. When I look at employment growth from 2010 to 2012, using occupational groups, I get the following (trying to match the great graphics that both the WSJ and Atlanta Fed use).
What we see here is strong job growth at both the top and bottom ends of the wage spectrum. Yes, food preparation and personal care account for a disproportionately large share of jobs gained in recent years, but so too have business and financial services, healthcare practitioners, computer and mathematical occupations and management. Where we have seen slower growth is in the middle. The light blue bars, which I term lower middle-wage jobs account for about 40% of all occupations in 2012 yet account for just 26% of the growth. The dark blue bars, which I term upper middle-wage jobs, account for another 19% of all occupations and 0% of the growth. This, by definition, is job polarization. These patterns have been ongoing in the U.S. and in Oregon for decades and as Jaimovich and Siu point out, the process is exacerbated during recessions. These middle-wage occupations fell much further during the Great Recession and are growing much slower coming out, if at all. Expectations are for these occupations to continue to grow in absolute numbers as the expansion continues, however the relative growth rates will likely continue to lag.
This has been just an excerpt, but is important to add to the ongoing national discussion. Look for my Oregon specific research report in the next month or so. Finally, while the above looks at broader occupational groups, one can dive into individual occupations. Although BLS warns against such work as you do lose some coverage given definitional changes and tweaks, it can still be informative. Below I take all individual occupational data from 2010 and 2012 and break employment down into approximate deciles and show the changes. Here, in particular, one can see the lack of job growth in the middle-wage range.