David R. Henderson, an economist at the conservative Hoover Institution, said the existence of additional options outside a worker's current occupation or city made him skeptical that concentration was having an effect on wages. Skilled workers, he said, can seek out opportunities in other cities. Less-skilled workers can change occupations relatively easily.
"Because they're unskilled, they fit in many kinds of jobs, and so you have more employers at the local level," Mr. Henderson said.
Some manufacturing industries, like breakfast cereal and tobacco, are even more concentrated than farm equipment. But since many workers in those businesses are less skilled than farm-equipment mechanics, they may be more interchangeable with workers in other industries.
To the extent that less-urban areas have a problem, Mr. Henderson added, policymakers should make it easier for people to move to cities where there are more opportunities, perhaps by easing building restrictions that drive up housing prices. That suggestion was also raised in a report on employers' market power by President Barack Obama's Council of Economic Advisers.
Because of my 4-part series on monopsony in labor markets in October 2016 in which I dissected a report from President Obama's Council of Economic Advisers, Ben Casselman reached out to me to ask my thoughts on José Azar, Ioana Marinescu, and Marshall I. Steinbaum, "Labor Market Concentration," NBER, December 2017. The ungated version is here.
I took some time to read it and spoke to him the next day. I'm very pleased that he quoted me accurately.
Ben left out one major criticism (I'm not blaming him--as long as he quotes me accurately and doesn't take me out of context, it's good). To measure concentration, the authors used data from CareerBuilder.com. What's wrong with that? Here's the extreme example I gave Ben. Let's say that you have 1,000 employers of a particular kind of labor in one area. What if only 2 of them are hiring? Then the data from Azar et al will show a highly concentrated market. Ben told me that he had wondered about that too and had asked Azar. Azar had answered that that's not a problem because if only 2 firms are hiring, that's what's relevant to job seekers. I told him I didn't think that was a good answer: the worker has 998 more potential employers and the odds are that some of them, within a few months, will be hiring.
I've thought about it more since then. I think my point is even stronger. Let's say that none of them is hiring in the next few months. It's still the case that there's potential competition among employers for workers. It's hard to believe that the wage would be much lower because only 2 firms are hiring if there are 998 other employers of the same type of labor.