Last week, Paul Krugman posted a chart prepared by Lawrence Mishel of the Economic Policy Institute that aimed to explain the gap between growth in compensation for the median worker and productivity growth from 1973 to 2011. Here is Krugman’s brief discussion:
Larry Mishel has a systematic breakdown of the reasons for worker income stagnation since 1973. He starts with the familiar divergence: productivity up 80 percent, the compensation (including benefits) of the median worker up only 11 percent. Where did the productivity go?
The answer is, it’s two-thirds the inequality, stupid. One third of the difference is due to a technical issue involving price indexes. The rest, however, reflects a shift of income from labor to capital and, within that, a shift of labor income to the top and away from the middle.
What this says is that widening inequality makes a huge difference. Income stagnation does not reflect overall economic stagnation; the incomes of typical workers would be 30 or 40 percent higher than they are if inequality hadn’t soared.
A few immediate questions came to mind: was the economy-wide productivity increase identical across all sectors and across all categories of workers? Mishel, to his credit, provides more detail regarding his sources. The comparison Mishel is making is between hourly compensation of production/nonsupervisory workers in the private sector and productivity for the total economy. That is, we’re talking about production workers in manufacturing and nonsupervisory workers in services. This category is not coextensive with all workers, however.
Back in 2005, Alan Reynolds of Cato observed the following:
All such comments allude to “average earnings” of production workers in manufacturing and non-supervisory workers in services. But that data series does not purport to measure hourly pay at all, much less a typical worker’s wage. The figures cover only 62 percent of all jobs, not 80 percent, if government workers and the self-employed were included. And that just begins to explain the confusion.
In fact, this data series is so misleading it is finally being phased out by the Bureau of Labor Statistics (BLS), to be replaced over the next four years by one that covers all private employees. For one thing, as the BLS explains: “the production and non-supervisory worker hours and payroll data have become increasingly difficult to collect, because these categorizations are not meaningful to survey respondents. Many survey respondents report that it is not possible to tabulate their payroll records based on the production/non-supervisory definitions.”
Not everyone will embrace Reynolds’ analysis, and I would be eager to see other estimates regarding the share of total employment accounted for by the category of production/nonsupervisory workers.
What we do know is that we’re dealing with moving targets over time. Consider, for example, the changing civilian employment-population ratio, which you’ll find at the Federal Reserve Bank of St. Louis. There was a steep increase in the ratio for most of the period covered by Mishel, presumably due to rising female labor force participation, yet there has been a pronounced decrease in recent years, due to a number factors ranging from early retirements to the continuing impact of the deterioration of the labor market prospects of less-skilled men and the associated increase in the disability rolls, etc. This doesn’t undermine the story Mishel is trying to tell, but it seems worth of note.
Moreover, the sector composition of the U.S. economy has changed over this time, as Spence and Hlatshwayo have emphasized in their work on the evolving structure of the U.S. labor market, which we’ve discussed on a number of occasions. Employment growth in the U.S. has tended to come from the nontradable sector while increases in value-added have tended to come from the tradable sector.
Virtually all (97.7 percent) of the incremental employment [from 1990 to 2008] stems from the nontradable sector. This occurred despite dramatic labor-saving technology in information processing that ran across all sectors of the economy.
The leading employment sectors are government and health care, in that order, both on the nontradable side. Together these two sectors generated more than 10 million addition- al jobs over the period, accounting for almost 40 percent of the increment. Health care added 6.3 million jobs on a base of 10 million. Government added 4.1 million on a base of 18.4 million.
There are a few ways to think about this shift in sector composition and how it might impact productivity. It could be that Mishel was actually understating the extent of the productivity wedge for private sector workers by using economy-wide productivity rather than private sector productivity. Measuring the productivity of public sector workers is notoriously difficult, as is measuring the productivity of workers in the rapidly expanding health sector.
But what if much of the private sector productivity gains were being driven by a minority of private sector workers? Spence and Hlatshwayo observe the following:
The tradable sector experienced job growth in high-end services including management and consulting services, computer systems design, finance, and insurance. These increases were roughly matched by declines in employment in most areas of manufacturing.
How many of these workers are best described as nonsupervisory? And might declines in employment in most areas of manufacturing, a domain in which productivity has increased considerably, also have an impact on the larger picture?
Krugman and Mishel argue that “inequality” is the problem, and if there had been no increase in “inequality” over the period under discussion, the compensation picture would look very different. But this strikes me as a non sequitur. If manufacturing employment has declined and service employment has polarized between knowledge-intensive, high-end services and less-skilled service work, it is at least possible that we’re stacking the deck. That is, the big productivity gains may well have come in helping a relatively small number of privileged workers do their jobs more effectively than in the past through new combinations of technology and labor.
To put this somewhat differently, no one is surprised to learn that manufacturing employment has decreased as a share of total employment, and that less-skilled U.S. manufacturing workers are competing with less-skilled workers in other countries. This will tend to depress compensation levels, all else being equal. Productivity in manufacturing to at least some extent reflects inputs from other countries, as Michael Mandel has argued. So basically, the representation of highly-productive manufacturing workers (who are highly-productive because of their access to expensive capital goods and, perhaps, because they are processing lower-cost inputs as part of a global supply chain) as a share of all production/nonsupervisory workers has presumably fallen over time. At the same time, the share of workers specializing in outsourced household production may have increased over time in response to the increase in the number and the affluence of two-earner households.