Why Aren’t Wages Rising? The Answer Sounds a Lot Like Monopoly (Episode 42)

Date: 22 Mar 2018 | posted in: Building Local Power, Podcast, Retail | 0 Facebooktwitterredditmail

If unemployment is so low, why aren’t workers getting cut in on the deal? That’s a question that our guest, economist Marshall Steinbaum, has been trying to answer. In a new study, which was featured in the New York Times, Steinbaum and his co-authors find that one of the main reasons that wages have not kept pace is that most local labor markets are highly concentrated. Only a few companies are hiring and, as a result, these dominant firms have the power to set wages below the rate that people would earn in a more competitive labor market.

(The technical term for this is “monopsony,” a close cousin of monopoly, which refers to a situation in which a buyer — of labor, in this case — has the power to dictate the price.)

Steinbaum, who’s Research Director at the Roosevelt Institute, joins Building Local Power host and ILSR co-director Stacy Mitchell to discuss his research and how elected officials can fix the broken market for labor.

“Antitrust can do a lot in the labor market to make it more competitive, and it’s absolutely appropriate to talk about things like merger review on the basis of labor market definition, and [to] take wage reduction seriously as a potential threat of anti-competitive conduct,” says Marshall Steinbaum

Photo Courtesy of Roosevelt Institute

“But antitrust alone cannot solve the labor market’s problems, and specifically the wage-setting power of employers or just employer power more generally. This is why historically we have had labor market regulations, why we have protections for collective bargaining, because we recognize that the employer/employee relationship is inherently one of unequal power.”

Our guest, Marshall Steinbaum, provided a book recommendation, and we’ll also share some of his original research:

Map courtesy of The Roosevelt Institute with this caption: “A map of the average concentration of the 200 occupations that appear most frequently in the Burning Glass data, by Commuting Zone.”


  • 6 Ways to Rein in Today’s Monopolies — Monopolies are strangling competition and cutting off opportunity. In this feature for The Nation, we show how to stop them.
  • The New Brandeis Movement: America’s Antimonopoly Debate — This article by Lina Khan in the Journal of European Competition Law & Practice details the work that a coalition of academics, policymakers, and thought leaders have done to change the conversation about monopoly in America.
  • Listen: Stacy Mitchell on “We the Podcast” with Rep. Keith Ellison and Lina Khan — Taking on Amazon. It’s a big subject. But in this episode of Rep. Keith Ellison’s podcast, “We the Podcast,” ILSR’s Stacy Mitchell and Lina Khan of Open Markets join Rep. Ellison to talk about how to do just that — and why we need to. The three also discuss how Amazon is part of a trend of market concentration more broadly, and what that means for the middle class and communities.
  • Report: Monopoly Power and the Decline of Small Business — This report from ILSR’s Stacy Mitchell details how the United States is much less a nation of entrepreneurs than it was a generation ago. It suggests that the decline of small businesses is owed, at least in part, to anti-competitive behavior by large, dominant corporations.
  • Beating the Monopolies: Barry Lynn Explains How We Will Win (Episode 30) — In this episode, Stacy Mitchell interviews anti-monopoly researcher Barry Lynn talks about how we fight against our economy’s most prominent monopolies, such as Amazon, Facebook, and Google.
Stacy Mitchell: Hello and welcome to Building Local Power. I’m Stacy Mitchell of the Institute for Local Self-Reliance. A few weeks back there was this really great story in the New York Times by Noam Scheiber and Ben Castleman. It was a story about a guy named Matt Gies who lives in rural Wisconsin.

Gies grew up on a farm and he spent a lot of time as a kid tinkering with farm equipment. When he grew up, he decided to become a farm equipment mechanic, repairing tractors and the like, and he really loved the work. He worked for a John Deere dealership, owned by a Wisconsin company called Riesterer & Schnell, which owns 12 different John Deere dealerships across Wisconsin.

But Gies actually left that job because he said they demanded far too many hours of work for too little pay, so three years ago he quit. The problem is that all these years later, despite being surrounded by farms, and despite looking for work, he hasn’t been able to find another job repairing farm equipment. Most of the seven John Deere dealerships within an hour’s drive of his house are actually owned by Riesterer & Schnell, the same company that he left.

They have a monopoly on the supply of jobs repairing farm equipment. It turns out that this is a common situation in most regions of country. Many occupations are dominated by just a few employers at most. As a result, there’s no real competition for labor. Economists have recently began to study this problem and what they’re finding is quite striking. Concentration has played a significant role in holding down wages over the last 20 years. It’s one of the key reasons American workers have not gotten a raise, even as productivity has soared.

Today on the show, our guest is Marshall Steinbaum, one of the economists whose been at the forefront of this research and whose work was featured in that New York Times story. Marshall is Research Director at the Roosevelt Institute. Marshall, welcome to the show.

Marshall Steinbaum: Thank you for having me. It’s my pleasure.
Stacy Mitchell: Why don’t we start with you telling us a little bit about what’s been happening to wages in recent years?
Marshall Steinbaum: Yeah, so there’s a longstanding question among labor economists about why wages have been stagnant, a issue that has now lasted for several business cycles. Normally economists would expect there to be a decline in wages, or at least a slowing in the growth rate of wages when there’s a recession, and then there’s a catch up period during the boom that follows the recession, in which wages grow faster than average and bring the whole economy back, such that the share of total income that goes to workers is more or less constant over time.

Even as the economy grows, there’s an equal division of the pie among the owners of labor and capital over time, and consequently, the total absolute amount that’s going to workers would be rising over time as the economy gets larger.

What has been going on now since 2000 at least is that the share of the pie that goes to workers has been in decline, and it’s been in decline in a very specific way. That share of the pie declines when there’s been a recession, as there was in 2000-2001, and then again in the Great Recession, starting in 2008. And then, it’s basically flat during the resulting recovery that follows the recession.

That is the phenomenon that I think is getting increasing attention from labor economists, because it cannot be explained by any of the data, by any of the observables that economists would typically think of as causing long run changes in wages for individuals, and for the economy as a whole. Notably, that would be education, so the view among economists is that what determines how much individual workers get over the course of their lives is what their skills are, and you can tell what their skills are from their level of education and their level of experience in the labor market.

Increasingly, education is just not a very good way of discerning who gets what in the labor market. We had already known this before my paper came out, about monopsony specifically, because there’s an increasing inequality in interfirm earnings among workers. What that means is the company you work for matters more in determining your wages than it previously did. Rather than it be your own qualities, your own data that’s like education and race or gender, that is relevant to knowing what your total income is, it matters more who you work for and their position in the economy, and their position vis-a-vis workers.

I think that had already primed the economics scholarly community to look for explanations for who gets what that are at odds with received wisdom, at least as its existed for the last couple of decades in economics.

Stacy Mitchell: Yeah, there’s a lot of interesting stuff in what you just said, and let’s unpack it a little bit. But let me first just go back to what you were saying near the beginning, which is that in recessions, what we’ve seen in the last couple of recessions is that the share of the overall economic activity that’s going to wages has been declines, and that’s sort of normal in recessions.

But then what’s been really unusual is that in the recovery periods, workers are not seeing their wages go up. And I think most people who are listening will recognize that problem maybe in their own lives. What’s really striking right now is I think, if I read this correctly, unemployment is now at a 17 year low. I mean, we would expect, am I wrong in that kind of condition that we’re experiencing right now, wouldn’t we expect wages to really be being pushed right up?

I mean, if there’s that much demand for work overall, that would have this effect on wages, but we’re really not seeing that and what signs of growth we are seeing in wages are tepid. Is that right?

Marshall Steinbaum: Yeah. I think so. There’s definitely a puzzle that results from, “Well, if unemployment is so low, why aren’t workers getting cut in on the deal?” So to speak. I want to drill down directly to one of the statements that you just made, which is the reason why unemployment is low because demand for labor is high. I think that’s the thing that economists would typically have assumed that when unemployment is low, that is because demand for labor is high.

But that is potentially one of the mechanisms that is broken in the economy. I think one of the causes of measured low unemployment is that a lot of people have left the workforce and you can see that both at younger ages, people spending more time in school, people going back to school, people getting credentialized, and also at older ages, people finding ways to take what amounts to early retirement, or taking early retirement as opposed to waiting for full retirement. All of those mechanisms are eating away at the labor force from both ends of the age distribution, and even in the middle, you see people exiting the labor force for good, and becoming discouraged.

I think that matters a great deal for whether, to what extent monopsony is the ultimate cause of these problems, because monopsony’s a problem that would cause firms to demand fewer workers and thus, lower wages for the workers that they already have, and that could macro level, exactly be at the heart of these issues of, “Well, why has labor force participation seemingly declined among workers at every level of the labor market life cycle?”

Stacy Mitchell: Yeah, and we’ve got an opioid crisis out there, and people dying earlier than they used to. I think that feeds into maybe part of this despair and being outside of the job market that you’re talking to. It’s all maybe wrapped up together. You used a word a couple of times now, monopsony. Now, most people are wondering, “Are you misspeaking? Did you mean monopoly?” ‘Cause that’s a word people know, but what is monopsony?
Marshall Steinbaum: Monopsony is like monopoly, except it’s like when the power is held on the buyer side of any market. You can have a monopsony in say the market for agricultural products, as I think we probably do have. It’s especially important in the labor context because the view is that in general, workers have less power in the labor market than do employers. Certainly the findings that we have in our power would seem to suggest that, as does the findings on interfirm earnings inequality, and lots of types of evidence point in the direction of widespread monopsony power in labor.

What that means, in the nerdiest context, is that employers have power to set wages. In the standard, competitive model of the economy, individual firms do not have the power to either set prices in their market for their output, so that’s monopoly, nor do they have power to set wages in the market for one of their inputs, which is labor. They just go to the market and if they need a new worker, the market sets the price for that work.

When there’s monopsony power, firm’s decisions change because they affect the wages that all of their workers make when they decide to employ or not employ a worker at the margin. So, in the most orthodox model of labor market monopsony, firms will choose to hire fewer workers than they do in a competitive labor market, because of the effect that hiring fewer workers has on the wage that they pay to all of their workers.

That’s the fundamental story that’s going on not just in our paper, but in all of the theoretical work about why would we think that monopsony would have an effect on all of these labor market outcomes we’ve been talking about. Not just wages, but people exited the labor force, as you pointed out, the opioid crisis. Lots of what I would consider labor market pathologies can potentially be explained by widespread monopsony power.

Stacy Mitchell: Labor market pathologies. That’s a new phrase I haven’t heard before. A lot of our listeners are probably familiar with monopsony in the context more that we’ve talked about it on this show before, where a company is so powerful that it can set the price that it pays for a certain input. We’ve talked about this, like for example, in agriculture, where farmers are trying to sell their hogs or their chickens, and the processing industry is so concentrated into a few hands that those companies basically can say, “We’re not going to pay you very much and you have no competition.”

This is kind of the same thing, but it’s about wages and what workers can get in the marketplace. So you’ve done together with a couple of other economists, Jose Azar and Ioana Marinescu, have done a couple of studies that really delve into this and have been getting a lot of attention and I want to talk a little bit about the first one that you did which came out back in December. I’ll say to listeners, you can go to the show page for this episode and we’ll post links to these studies and to the New York Times article, and you’ll find the show page at ILSR.org.

But starting with that first study, one of the most visually arresting maps I’ve seen in a while is in that study, and it’s this map of the United States broken up, I think by county, and it’s a measure of how concentrated the labor market is in each county. That is, the notion that if you’re in a particular occupation, there may only be one or two, or three companies within your commuting distance that you could apply to for work. It’s a highly concentrated market, so you use red to show extremely concentrated and most of the map is red.

And then, some other parts of it are orange and yellow, which are highly and moderately concentrated. That’s pretty much the rest of it, orange, yellow, and red. And then there are few islands of green. Those are unconcentrated markets, but they’re very few. It’s Minneapolis, Denver, Boston, Miami. It’s mostly cities. Then, just surrounded by this sea of highly concentrated labor markets.

Tell us a little bit about what’s behind this map. What kinds of occupations are you looking at? What does this really mean?

Marshall Steinbaum: Yes. So, the dataset that we worked on in that first paper is from the online job matching company, CareerBuilder. Basically, employers would pay to post a job ad on that website and then would be applicants can see it, just as they would have in the olden days gone to classified advertising in the newspaper, and submit an application via the website for that job.

That’s a fairly rich dataset relative to the datasets that most labor economists are used to, because you can see at least some information about both employers and employees. So, the classical labor economics that I was referring to before where you study education, race, and gender, that’s because mostly for workers, you just see things about the worker, and also what they get paid, but you don’t see anything about their firm.

Here at least, we do see which firms are posting vacancies and where they’re located, and we can thereby create the map that you were just referring to. What we did is look at the occupations that arise most frequently in this database. I think it’s 20 of the most frequently appearing occupations, and we defined the labor market as the vacancies that are posted for those occupations within a given commuting zone, in a given quarter.

So, that is an important concept, market definition, especially if you’re doing antitrust type analysis, to see whether the market is concentrated or unconcentrated, and to what extent. We think that that market definition is not necessarily exactly the right market definition in every single case, but it is small C conservative, in the sense that when workers are looking for a job, they tend to look for jobs that they think they would have some chance of getting, and if you look at the occupation level that we look at, it’s actually wider than the set of jobs that workers would think they have some chance of getting, looking at that dataset from the perspective of which jobs do the workers actually apply to.

We’re including more job vacancy postings in our market definition than we think are really relevant to individual workers who are looking for a job, so that would tend to underestimate the degree of concentration in that labor market, and yet we still find the results that you refer to, which is that labor markets tends to be highly concentrated.

One last thing to note about how that chart is created, how that map is created, is this whole dataset is about job vacancy postings, insofar as economists have studied labor market concentration at all. Before, it tends to be the concentration of employment, and not the concentration of vacancies, and I think it is correct to look at concentration of vacancies if you can, as opposed to the concentration of employment, generally easier to get data on the concentration of employment, because what’s really relevant to workers who are looking for a job is how many firms are actually hiring.

And again, going back to the idea of pathologies of the labor market, one thing that we observe in this era of slack labor markets is that workers tend to stay in the same jobs for longer because they themselves cannot move up the job ladder as the metaphor that economists typically use, and the flip side of workers staying in the same side for longer and not moving up the job ladder is that any given job is vacated less frequently.

Even if say there’s a number of people currently working as a farm equipment mechanic in the area of Wisconsin, where Matt Gies works, that doesn’t necessarily mean that there are vacant jobs in that area that he can potentially apply to. That’s why I think we show a finding that caught a lot of people’s attention, because they tend to think like, “Oh, well every firm, not every firm has a farm equipment mechanic, but every firm has administrators and secretaries,” and certainly it seems like the healthcare sector employs a lot of people, so you would think that say the market for nurses is relatively unconcentrated, but no, that doesn’t necessarily mean that there are jobs available for people in all of these different occupations, especially not for people who are looking for a job.

There I think there are many fewer options to be had, and that’s why the finding of concentrated labor markets seems to ring so true with people.

Stacy Mitchell: We’ve talked obviously about farm equipment mechanics, and you mentioned nurses. Were there other particular occupations that really jumped out at you in this data as being highly concentrated areas where people were going to just not have a lot of choices in places that might hire them?
Marshall Steinbaum: Well frankly, basically every occupation looks quite concentrated. We have a figure in the paper that shows the level of concentration and there’s a great deal of variation over the 20 occupations that we look at, but none of them look particularly unconcentrated. And specifically, I think a lot of people have the view that labor markets would get more concentrated as you move up the skill hierarchy of the labor market, because as you get more specialized skills, the set of employers that are really options for you diminishes, but there’s still very high labor market concentration among unskilled employees, and frankly, I would say evidence suggests that monopsony is a bigger problem.

Not our evidence. I think that there exists pretty strong circumstantial case that monopsony’s a bigger problem for lower skilled workers, and that’s exactly because they are “more interchangeable,” that their power is therefore diminished in the labor market and what makes workers ever get what they’re worth is the idea that they’re irreplaceable.

Our paper certainly doesn’t suggest that monopsony is a bigger problem for … I should say concentration is a bigger problem for highly skilled people, and in fact, there are occupations in our dataset that most, especially antitrust people, but really anybody who studies the occupational distribution of the labor market might think, “Well, concentration can’t really be a problem for secretaries or administrators or for people who seemingly have skills that are relatively interchangeable across employers.”

I think that’s just not borne out in the data, and the fact that it’s not borne out really should make people rethink their theory of how the labor market works, and how it is determined who gets what in the labor market.

Stacy Mitchell: That’s really interesting. This is a really broad effect, and even if you’re someone who does have skills that ought to be relatively interchangeable, you’re still facing these really concentrated markets where wages are being held down as a result. I would actually ask about the level of the effect of this on wages. I mean, are we talking about people getting a little bit less? Like pennies or nickels on the hour, in terms of the effect on wages? How big of a factor is this?
Marshall Steinbaum: Our preferred specification from the CareerBuilder paper is that we think the best estimate that we have of the effect of concentration on wages is that it reduces wages by on average, 17%. The confidence interval for that is something between 10 and 25%, and I would consider that to be a sizable effect, but I also think it’s also worth pointing out that that is net of a lot of other things that could affect both the level of concentration and the level of wages.

As in any social science endeavor, you want to say, “Okay, well we have documented a concentration and wages, and now let’s see if we can be more robust in saying the variation in concentration causes a variation in wages,” and our estimate for the extent to which the variation in which concentration causes a variation in wages is that if you increase the level of concentration from the 25th percentile labor market to the 75th percentile labor market, you reduce wages by 17%.

But I think the way a layperson might look at that paper, but more generally the question of how big of a problem is monopsony in the labor market, is that you could see all of these different mechanics by which monopsony could affect wages, among other concentration, as not so easily separable. As a social scientist, you want to chop up the potential pathways of causation and say, “Well, this one, specifically concentration, has this estimated effect on wages.”

I think the more general policy question is to what degree does power on the part of employers affect wages? And there are multiple mechanisms by which that could affect wages. I mean, for one thing, let’s say that shareholders or firms are demanding that the firms that they own pay out more money to the shareholders and therefore hire fewer people. I would certainly think that that reflects the increasing power of employers versus employees.

It might have two different effects on the mechanism that we’re talking about. It might cause firms to post fewer vacancies, and it might also cause firms to pay lower wages, but in reality, both of those outcomes are the effect of the same cause, which is the different power that is held by the shareholders to extract what they can from the firm.

This whole long story is about trying to say, “Well, we might say that the point estimate is 17% but we’re trying to net out these other mechanisms that cause variation in concentration, and also cause variation in wages,” where it’s not the variation in concentration that causes the variation wages, but that doesn’t necessarily mean it’s not the rising power of employers that causes the variation in wages. I think that, there’s never going to be any one paper that tells you what the answer to that question is, but I think there’s a lot of research now that suggests that that is really crucial component of understanding what’s going on in the economy and in the labor market.

Stacy Mitchell: I want to turn next to talking about what we should do about this, but first, we’re going to take a short break. You’re listening to Marshall Steinbaum, Research Director for the Roosevelt Institute. I’m Stacy Mitchell with the Institute for Local Self-Reliance. We’ll be right back after a short break.

First of all, I want to thank all of you who made a donation at the end of the year to the Institute for Local Self-Reliance. Your financial support not only underwrites this podcast, keeping it ad free, but also helps us produce all the research and resources that we make available on our website, as well as all of the technical assistance that we provide to elected officials and citizens across the country.

Every year our small staff helps hundreds of communities challenge corporate concentration, and rebuild their local economies. So, please take a minute and go to ILSR.org, and click on the donate button if you haven’t. That’s ILSR.org, and if making a donation isn’t something you can do, please consider helping us in other ways.

One of the best things you can do to help us reach a wider audience with this podcast is to rate and review it on iTunes, Stitcher, or wherever you get your podcasts. That gives us a big boost in visibility, and it’s just hugely helpful when you do that, so thank you so much.

We’re back with Marshall Steinbaum, Research Director of the Roosevelt Institute, and we’re talking about concentration, and the ways in which corporate concentration is actually driving down wages and has really kept people who work for a living from earning what their should from their labors, and from all of the productivity increases in the economy.

Marshall, I want to turn to this question of like how, given what your research and the research of others is showing about these effects, what the answer is. I want to start with one potential answer that I’ve heard some people put out there, which is that workers should just move. We talked about the map and I was noting that there are a few, very few, green islands of actually competitive labor markets, mostly cities. Why don’t people in rural Wisconsin and Eastern Maine and Iowa and everywhere else just all move to Miami or Denver?

Marshall Steinbaum: You know, I’ve heard that claim made basically throughout my career as an economist. I put that in the category of blame the victim for the fact that we screwed up the economy, and frankly I think it’s just completely out of touch with reality. I mean, one of the findings not just in our paper, and with the CareerBuilder dataset, but in this case I can actually cite my own dissertation, is that historically very few workers move. If they do move, they move a pretty short distance.

That is, they want to stay as close as possible to where they originate, and the idea that the economy consists of workers who are highly mobile, any place is on the table, and in particular, if they’re not doing well in the labor market right now, that’s because they’re just unwilling to take the risk, take the plunge necessary and go to a place where they could definitely find a job. That’s just a wrong understanding of how people live their lives, and how people’s work interacts with the rest of their lives.

People don’t want to move. They want to be able to have a job where they live, and it’s not so easy to move because most people get their jobs from their family or their professional network, or I should say their social network, and to say, “Well, leave all that behind and just go to some city where you’ve never lived and you’ll be fine there,” that’s just not true for people. They’re not going to just have a job land in their lap.

I mean certainly if you have a family, you’re not going to take the risk of going some place where you don’t have a job and not being able to support them, and have them live, you’re going to stay where you have a social network and potential family to help you out with all of the things that are necessary in life.

Not only is it just not true that workers move very much. They’ve been moving less and less since 2000, and I would say again that that reflects these labor market pathologies. It is one of the pathologies of the labor market. Moving is something that workers do when they have a job offer that actually makes it okay, makes it worthwhile for them to take the plunge and leave their social network, or because it’s their social network that was actually able to come up with the job offer, even if it’s at some geographic distance from where they currently originate.

It’s indeed a problem that geographic mobility is on the decline, but that’s not a problem that individual workers aren’t doing what they should. It’s a problem that is part of the larger problem of economists and policymakers just not understanding how to do their job.

Stacy Mitchell: It’s interesting, too, ’cause looking at these green islands, I mean San Francisco and Boston, have you tried to buy a house in either of those places?
Marshall Steinbaum: Oh yeah.
Stacy Mitchell: That’s the other thing that’s really striking, and I’ve argued elsewhere that part of the reason that living costs have gone up so much in some of these cities has at least some roots in concentration.
Marshall Steinbaum: Yes, I think that’s undeniable. On the one hand, this issue of high housing prices in the supposedly dynamic parts of the economy is pointed to as the key barrier that is preventing the right amount of reallocation in the labor market, so it would be the case that people should just move from all of the rural areas you named to San Francisco, except that housing prices are too high, and the reason why housing prices are too high is because zoning regulations are too onerous, and if we only liberalize the housing market in San Francisco, then house prices would go down and everyone could move there and that’d be great, so that’s our economic policy.

I think that is again, also overly simplified, and misunderstanding why it is that house prices are high in San Francisco, and what the effect of a policy that would totally deregulate say the housing market in San Francisco, I think that would most likely increase the market power of incumbent owners of housing and of land in San Francisco, and possibly even increase rental costs to the people. That’s more speculative, but it’s certainly not some sort of policy panacea that you could just cause a bonanza of building and construction, that would suddenly bring the place of living in San Francisco down to the level of the rent of living in a much more rural and less densely populated area.

Stacy Mitchell: You’ve mentioned antitrust a few times in this conversation, and you talked about monopoly and monopsony being two different sides of the same sort of problem. Antitrust policy mostly has been thought of in terms of consumers for a long time, and we’ve talked about this on the show before, that when we think about concentration, mainly we think about is it causing prices to go up? And that’s really the only thing that gets considered.

You’ve been among, along with others, been advocating that some of the labor market effects ought to be part of how we think about antitrust, how we think about mergers. Tell me a little bit about what that would actually mean. How would you implement that kind of set of ideas in the context of antitrust policy?

Marshall Steinbaum: Yeah, that’s I think a highly topical policy question. We define labor markets as occupations by commuting zone, by quarter, essentially, and that paper, the CareerBuilder paper, shows that defined that way, labor markets are highly concentrated. The main policy lever that antitrust has been using, especially recently, is merger review.

There are some non-merger cases that at least the federal antitrust agencies undertake, but I would say that’s probably the bulk of their work is reviewing the mergers where two companies say, “Okay, well we want to merge” and the agencies either say, “We have no problem with that” or they bring a case and demand concessions, or even go to trial, as it happening now in AT&T/Time Warner.

The key economic analysis that’s undertaken as part of merger review is to deduce whether the merger of two competing companies in a given market would cause prices to go up or down, and that requires defining what the relevant market is for the parties to the merger.

As you’ve just correctly said, that is in the present time, more or less solely done on the basis of monopoly and pricing of output goods, and it is quite likely that the labor markets are defined fundamentally differently from product markets.

If you think of something like the Maytag/Whirlpool merger that happened in the mid 2000s, that was basically the two major home appliance manufacturers merged. They were allowed to merge unchallenged. I think there’s good evidence to suggest that that merger turned out to be anti-competitive after the fact, in the market for washing machines, dishwashers, other appliances that they produced.

At the time, I would bet that the authorities defined the market for washing machines and other home appliances as being the entire United States, or maybe something slightly less than that. But the point is, everyone gets their washing machine. They might shop for it retail, but there’s a distribution network that is essentially geographically unbounded.

Whereas for labor, the whole point of what we’ve been talking about, and certainly of our paper, is that the market for labor is quite a bit geographically determined. If that merger had been evaluated for its labor market impact, you would have seen the whole economy wide market for washing machines is not going to become less concentrated as a result of this merger.

I mean again, I don’t think that turned out to be correct, but I think that was probably the finding that the agencies came up with, but the labor market for workers at the factories where those companies worked are very likely to get more concentrated, and I think we know for a fact that a couple of factories at one or the other parties was closed directly as a result of that merger, and a lot of people were laid off.

Under the present implementation on antitrust policy, those layoffs and closure of factories would be put forward as an efficiency gained from the merger, because the parties would say, “Well, we’re concentrating our production at our most efficient factories, and this will actually end up benefiting consumers, because we’ll be able to lower prices, as a result of the fact that our production chain is more efficient.” But if you, again, analyze the merger from the perspective of labor market, then what looks like an efficiency is actually just a monopsonization of the local labor market and consequently would be viewed as anti-competitive.

That is but one of the areas in which antitrust policy could change as a result of taking labor markets seriously. There’s been a policy interest in curtailing the use of restrictive labor market agreements, so this would be in the realm of antitrust conduct, not of market structure. Things like no poaching agreements, and noncompete closes.

So, no poaching agreements are the parties are employers, and they would be agreeing to not hire one another workers. That is clearly already against the law, because it’s a horizontal cartel basically, in the labor market, and what movement there is to change the law with respect to no poaching agreements pertains to the use of them in franchising contracts.

So, say two McDonald’s franchisees or any McDonald’s franchisee in the contract it signs with the McDonald’s headquarters will say, “Well, we’re not going to hire the employees of another McDonald’s franchisee.” It’s unclear whether that’s a horizontal or a vertical agreement, but in any case, there’s a policy proposal from Senator Cory Booker to make that also illegal, which is good.

On noncompete clauses, that is an agreement between an employer and an employee for the employee not to work for a different company if they leave their current one. That is more controversial as a matter of antitrust law, because since the takeover of antitrust law by the Chicago School, vertical non-price agreements, that is imposing terms of any contract on the counterparty, is viewed as procompetitve, so if antitrust law were to move to treating noncompete clauses in the labor market as illegal, then that would just be a pretty major departure from existing policy, at least vis-a-vis labor. I’m sorry, vis-a-vis antitrust. Vis-a-vis labor may be more in the spirit of existing labor market policies.

The final point I want to make is a big economic theoretical point, which is that our findings, both in the CareerBuilder paper, and especially in our more recent paper that uses a larger dataset of essentially all online vacancy postings, in that paper we do more of an analysis of what is the right market definition for labor markets? We come to the conclusion, as I was discussing previously, that this occupation by commuting zone by quarter is actually fairly conservative because that overestimates the number of alternatives that are really functionally available to most workers if they’re looking for a job.

Based on our analysis in this paper, and from some other work in the literature, it looks like the right labor market definition is more along the lines of individual firms. That means that for every single firm in the economy, or under this theory, most firms in the economy, have a sufficient amount of wage setting discretion that is monopsony power, such that they are monopsonists in the market for their own workers’ labor. If every firm is a monopsony, then what you’re looking at in extreme antitrust policy world is that every firm should be broken up.

That is not really a feasible antitrust policy. What that suggests is rather that antitrust can do a lot in the labor market to make it more competitive, and it’s absolutely appropriate to talk about things like merger review on the basis of labor market definition, and that takes wage reduction seriously as a potential threat of anti-competitive conduct.

But antitrust alone cannot solve the labor market’s problems, and specifically the wage setting power of employers or just employer power more generally. This is why historically we have had labor market regulations, why we have protections for collective bargaining, because we recognize that the employer/employee relationship is inherently one of unequal power.

Whereas, the premise in antitrust is more or less that you have parties with equal power and then you try to correct the market imperfections, such that where one party becomes overwhelmingly powerful, you pair it back, their power, and limit it in such a way as they can exert it to distort markets completely and remove competition. In labor markets, we already know that one party’s going to be more powerful than its employers, so we have a whole realm of policy that is labor and employment policy, and collective bargaining, and unionization rights. All of those are necessary to restore a proper degree of balance to the economy. Antitrust alone is not going to be able to do it.

Stacy Mitchell: That’s really interesting. You just gave us three different ways to think about ways to approach this from a policy standpoint, and just to go back through those, I have a question about one but I’ll go through them backwards. The third notion that you mentioned is this idea that antitrust can’t carry the weight of everything, and I think that we’ve seen in this other sectors as well, but in the context of workers we also need to think about policies around how easy is it to form a union, for example. Do we let employers misclassify workers as being subcontractors when they’re really actually employees? Amazon and others do that a lot.

There are all kinds of ways in which policy needs to help ensure that people have a fair playing field when they’re bargaining as workers. We see this in other sectors, too. We’ve done a lot of work on dairy farms, for example. If you have a very perishable product which is fresh milk, and it’s very bulky, and you can’t ship it very far away before it goes bad, you’re really hamstrung. It’s hard to bargain with the milk process, and so we’ve always supported this history of federal and regional pricing supports in the milk industry, because it just really is inherently not the kind of competition, doesn’t produce enough competition for milk.

The second thing you mentioned was around these noncompete clauses and these poaching agreements which are just we should have you or someone else on the show at some point to talk more in depth about this, because it’s just so scandalous to me that you’re a fast food worker, and you can’t go take a job at a different McDonald’s, or at another fast food outlet, because you’ve signed this agreement or there’s a no poaching deal.

We could take some direct action using competition policy against those kinds of things, and then the first one you mentioned was around mergers, and this idea of having these labor effects and these labor markets being considered in the context of mergers. I have a question that I want to ask you about that to wrap things up here. Historically antitrust was as much about people as producers of value as workers, as it was about us as consumers.

John Sherman, who lent his name to the very first or one of the first antitrust laws back in 1890 said monopoly is a problem because, “It commands the price of labor without fear of strikes, for in its field, it allows no competitors.” Clearly this issue that you’re talking about has always been there from the beginning, and the question I have for you is how we should think about this. It relates to a really interesting blog post that I read, that Lina Khan wrote recently for a European outlet about … Lina is the Legal Policy Director at the Open Markets Institute.

She was writing about this new movement for a better antitrust, and what should be the principles of it. We’ll post a link to this on the show page, but one of the things she talks about is this idea that antitrust went wrong because we really collapsed everything to this single outcome, which is lower consumer prices. We should be careful, she says, not to collapse everything to other sets of outcomes, like for example, labor effects, that really we need to think about antitrust in terms of using it to structure markets.

What we should be doing is asking, “Does a merger create a market where there’s a lot of … Does it reduce competition too much?” How many competitors are there in this market? Is it a market where it’s easy for a new entrant to come in, a new business to come in? We really should be looking more at structure and not at outcomes, and even if we switch from a consumer outcome to a different outcome, we may be not in really a better place in the long run.

I wanted to ask you a little bit about how you think about that, particularly in the context of these labor issues?

Marshall Steinbaum: Yeah, I think that Lina’s definitely onto something there. I would say that economics as a field, and in particular it’s “policy relevant economics,” I used to have just a totally different conception of what the role of economists and of economics was in setting policy.

I would say we, for ideological reasons, developed the view that the economy structures itself, and hence the sphere in which that policy can affect is relatively narrow, and is about correcting minor deviations from the optimal structure of the economy and where the just field of operations and the field of which policy can affect is constrained from the outset.

Whereas, when antitrust was invented and brought certainly to the federal level, in which there absolutely were economists there at the birth part of the debate about what was the role of this policy alongside other policies like regulation, natural monopolies, utilities, and labor policies. I mean, taxation more broadly, it was that the role of economic policy was to structure the economy. It was up to us to decide what economy we wanted to live in and how it was going to work.

Or, we could just let the big businesses decide, let the powerful decide it. I think there’s a great quote in the book “Social Control of Business” by the economist John Maurice Clark that says the key question is, are we going to let big business control us or are we going to control them? 10 words that are an extremely pithy and absolutely on point summary of the issue that faces us with respect to antitrust and all these other policy areas, and that also point to where economics as a field went wrong, and where just overall policy making has shrunk to a shriveled bare imitation of its former self, in terms of what its ambitions were and what was on the table.

I think that given that it’s now a matter of consensus or nearing consensus that the economy that we live in is not functioning well to the benefit of all, and that really we went wrong somewhere, even if you don’t agree with me that antitrust is a big component of where we went wrong, I think it’s a little hard to deny that economic outcomes are not what they should be.

This whole idea that we really lost something profound and need to win it back I think is going to become a lot more attractive beyond the set of relatively small people like me and Lina who are really focused on this issue of antitrust policy and what the point of that is.

Stacy Mitchell: Thank you so much for joining us today. I want to just close by asking if you have a reading recommendation for our audience?
Marshall Steinbaum: That would be exactly what I just said. “The Social Control of Business” by John Maurice Clark. He was an economist and an “institutional economist.” He wrote that book, or he published that book first in 1926 and I should say I’ve not read just the 1926 edition. He republished it in 1939 when there was the whole experience of The Great Depression. Where did that come from? As well as both the experience of concentrating markets between 1926 and at least the mid 1930s, and then he spends a lot of time recounting why antitrust policy became a lot more active in the mid to late 1930s.

He had seen enough of the move to an active antitrust policy in the second half of the new deal to have something to say about that by the 1939 republication. He also has a lot of very interesting comments about the relationship between what we now call macroeconomic policy or macroeconomics in general, and antitrust and market structure. This is a confluence that has not had a lot of interest from economists in the intervening period up until right now.

I would say now there are papers being published that have that flavor to them, but I think that Clark was absolutely on point in terms of trying to locate the ultimate causes for both macroeconomic failures that happened in The Great Depression, and also why what we might think of a Keynesian macroeconomic policy is a good idea, given how much market power is just a pervasive fact about the economy and how big business and its structure, and market concentration give rise to macro phenomena that Keynes and his contemporaries were concerned about.

That is just, it’s a fantastic book. He covers a ton of ground, and it is, I mean maybe it’s especially meaningful to me ’cause I know how thin and pale and pathetic most economics publications are in the intervening 70 years since Clark was writing, but it certainly makes me think that we went seriously awry somewhere along the way, and we need to rediscover the spirit of John Maurice Clark.

Stacy Mitchell: Excellent. We’ll post a link on the show page to today’s episode to the 1939 edition of that book, so people can go find it, and give it a read. Thanks again, Marshall. It’s been great to have you on.
Marshall Steinbaum: Yes, it’s been my pleasure. Thanks so much. Very interesting conversation.
Stacy Mitchell: Thank you for tuning into this episode of Building Local Power. You can find links to what we discussed today by going to our website, ILSR.org, and clicking on the show page for this episode. That’s ILSR.org. While you’re there, you can sign up for one of our newsletters and connect with us on Facebook and Twitter.

Once again, please help us out by rating this podcast and sharing it with your friends. This show is produced by Lisa Gonzalez and Nick Stumo-Langer. Our theme music is Funk Interlude by Dysfunction_AL. For the Institute for Local Self-Reliance, I’m Stacy Mitchell. I hope you’ll join us again in two weeks for the next episode of Building Local Power.

Like this episode? Please help us reach a wider audience by rating Building Local Power on iTunes or wherever you find your podcasts. And please become a subscriber!  If you missed our previous episodes make sure to bookmark our Building Local Power Podcast Homepage

If you have show ideas or comments, please email us at info@ilsr.org. Also, join the conversation by talking about #BuildingLocalPower on Twitter and Facebook!

Subscribe: iTunes | Android | RSS

Audio Credit: Funk Interlude by Dysfunction_AL Ft: Fourstones – Scomber (Bonus Track). Copyright 2016 Licensed under a Creative Commons Attribution Noncommercial (3.0) license.

Photo Credit: Fibonacci Blue via Flickr.

Follow the Institute for Local Self-Reliance on Twitter and Facebook and, for monthly updates on our work, sign-up for our ILSR general newsletter.

Avatar photo
Follow Nick Stumo-Langer:
Nick Stumo-Langer

Nick Stumo-Langer was Communications Manager at ILSR working for all five initiatives. He ran ILSR's Facebook and Twitter profiles and builds relationships with reporters. He is an alumnus of St. Olaf College and animated by the concerns of monopoly power across our economy.