August 29, 2013
There are a lot of economists who are determined to say that technology is responsible for inequality rather than policy. There are many parts to this story that seem absurd on their face.
Are doctors, dentists and lawyers really whizs at technology? These occupations make up a very large share of the 1 percent. They sustain their income the old-fashioned way, they have the government arrest the competition. When more middle income workers like nurses and computer engineers start to see their pay rise, their employers run to the government whining about shortages so that they can bring in foreign workers to keep down wages.
The financial sector has become a fraud factory. Top executives at banks can pocket tens or even hundreds of millions of dollars off various illegal schemes and never fear more than repaying a portion in penalties. And of course the laws that protect workers’ right to organize unions have become a joke, while the laws that protect employers from organized workers remain sacred. (Union officials who openly support a secondary strike don’t risk a slap on the wrist from the National Labor Relations Board, they go to jail.) I could go on (and do).
But there is a big market for economists who produce stories saying that the problem is technology, so there will be economists who respond to the demand. Brad Plummer gives us a couple of examples in a recent blogpost.
The first is from Catherine Mulbrandon of Visualizing Economics. She gives us a graph which is supposed to show us that the industries that pay in the middle of the wage distribution are shrinking, while industries at the top and bottom are expanding. This is the story of wage polarization, with a disappearing middle.
Perhaps I have a problem with my eyesight, but it’s hard to see that story in the graph. The industry with the biggest increase in employment over the period from 2001 to 2011 is health care and social services, which sits almost directly on top of the line showing the average for all industries. Manufacturing, which is somewhat above the average pay rate, shows a big decline as we all know. However the industry grouping “professional, scientific, and technical services,” which is only slightly higher on the pay scale, is number 3 in job growth over this period.
Retail, which is definitely towards the lower end, is a big source of job loss over this period. Education services, which is closer to the average wage than manufacturing, is a big job gainer. If wage polarization is going on, you won’t know it from this graph.
The second graph, which comes from Joshua Lerner of the Oregon Office of Economic Analysis, does a little better job of telling the story. This one shows occupations by average pay and we do see that the 2nd highest of four groupings is not doing very well in the job creation category.
But even here the story is less compelling that the proponents of the technology did it story might hope. Sales, which is just below the wage cutoff that would make it a middle wage job, saw the second fastest growth for all occupations over this period. It’s also worth noting that people always take bad jobs in a weak labor market. The fact that restaurant employment has been growing rapidly tells us much more about the state of the labor market than the demand for restaurant workers.
The two big losers in the middle wage category were construction workers and teachers. The former look to be far more the victims of a collapsed housing bubble than technology.
The latter were primarily victims of budget cuts at the state and local level that were forced by members of Congress who don’t believe in economics. It’s hard to make this a technology story unless we somehow think that technology is responsible for putting political power in the hands of economic creationists.
This issue has been looked at more systematically. My colleagues and friends, John Schmitt, Heidi Sheirholtz, and Larry Mishel, examined the growth in occupations by wage level, following the work of David Autor and David Dorn. They found that the story of wage polarization fits the pattern of the 1980s to some extent, but it doesn’t fit the pattern of the 2000s at all.
In spite of their demolition of the Autor-Dorn thesis, we can continue to see the wage polarization story pop up in endless different permutations. Clearly some folks have a reason to believe.