Monopsony in motion

Megan McCardle writes that the monopsonistic model of labor markets requires "substantial search friction." I assume that by "search friction" she means the search costs of finding a new job. But while the monopsony model does indeed require that there be important frictions in the labor market, search costs are by no means the only, or necessarily the most important, friction.

Other frictions include heterogeneous preferences -- the fact that you might, for your own personal reasons, prefer a particular job, because you like the type of work, or the hours, or the benefits, or your boss, or your co-workers, or because it offers opportunity for advancement. Another type of friction is mobility costs -- you might want to remain at your job because it's close to home and if you took another job, the commute would cost you more, in time or money.

Those frictions, along with search costs, are the most plausible kinds. Megan argues that the high turnover in the fast food and retail sectors implies that the search costs in those sectors are low, because it's generally easy to find another job in the same sector. That may be true, but heterogeneous preferences still play a role, as do mobility costs, particularly in rural areas.

I was a researcher for a study that looked at low-wage work within a particular retail chain, and one big difference between stores within the same chain were work scheduling practices. In some stores, workers didn't know from week to week what their work schedule would be, while in others employees knew their schedule a month in advance. While the last-minute scheduling wasn't a problem for some workers, others, particularly single mothers, had a strong preference for knowing their schedule in advance. It was important enough to them that they probably would have passed up other jobs that paid them more but practiced last-minute scheduling.

I agree that in some ways the fast food and retail sectors are a closer match to the perfect competition labor model than to monopsony. Yet there still are frictions (if not due to search costs, then due to heterogeneous preferences and sometimes to mobility costs). And also, the other part of the monopsony theory is met in the fast food and retail sectors -- i.e., the employers in these sectors tend to set the wages, with very little role for bargaining or negotiation. So I would argue that monopsony is still a more appropriate model here.

Moreover, certain empirical facts also seem to suggest that monopsony is a better model for these sectors. For example, many minimum wage studies of the fast food sector, such as the famous Krueger and Card studies, have shown that an increase in the minimum wage does not reduce employment. In the perfect competition model, this would never, ever happen. So how do you explain those results? By far the most plausible interpretation is a monopsonistic labor market, because in a monopsony, imposing a minimum wage will not necessarily reduce employment (although it can reduce employment if it's set too high).

Finally, to Megan and others who argue that the perfect competition model is in general more plausible than monopsony, I would ask, if your employer cut everyone's wages by one cent, do you think everyone at your workplace would immediately quit? Because that's what the perfect competition model literally predicts. And yes, of course I'm being a bit unfair here, because all models are stylized representations of reality, and therefore have to simplify matters in ways that may seem grossly unrealistic.

But some stylizations seem to fit the facts on the ground more closely than others. And the monopsony model, which assumes that there are search costs, mobility costs, and heterogeneous preferences that might impede job change, seems to me to be more reasonable than the perfect competition model, which assumes no such obstacles whatsoever.

Comments

[Benjamin Rooney]

This is not a specific response to this post, but a general response to your blog. I'm all kinds of interested in economics, but your recent posts are my first exposure to the entire concept of monopsony.

Which is what I want to comment on. I try to read blogs that expose me to worlds I would not otherwise understand. As a white male born into a wealthy family, I'm about as privileged as it is possible to be, and I find your writing on feminism to be phenomenally eye opening.

Everything else you write about, including your current discussion of labor market models, is fascinating, well researched, and compellingly written. Thank you for the time and effort you obviously devote to sending these writings off into the aether.

[Benjamin Rooney]

Okay, doch, I do I have a question.

Why would the perfect competition model require a reduction in employment to be a result of an increase in the minimum wage? I'm utterly ignorant of the model (sorry), but it seems like an increase in the price paid for labor would only decrease demand for that labor after some breakpoint was reached where the labor being paid for cost more than the return on that investment. In the case of McDonald's, obviously they have an incentive to pay the lowest possible wage, but (having worked in a McDonald's, I know) they have every step of hamburger production scientifically whittled down to create the most efficient hamburger-producing process possible. That being the case, as long as customers continue to be willing to purchase hamburgers at a price that is a reasonable percentage higher than the cost (including labor) of producing them, shouldn't McDonald's be forced to employ the same number of people?

I guess my question is, if the minimum wage rises and McDonald's should theoretically respond by firing Bob, who is going to make Bob's hamburgers?

[tps12]

The perfectly competitive market model should be introduced at the beginning of intro economics courses the way the geocentric universe model is introduced at the beginning of intro physics courses: a significant historical step along the road to scientific insight into the functioning of the real world, and no more.

[Chris Hammond]

"Finally, to Megan and others who argue that the perfect competition model is in general more plausible than monopsony, I would ask, if your employer cut everyone's wages by one cent, do you think everyone at your workplace would immediately quit? Because that's what the perfect competition model literally predicts."

Not so, if wages are cut by one cent, some workers will quit, but not all. The supply curve for labor in a perfectly competitive model is upward sloping - there would still be workers willing to work for lower and lower wages until the wage is lower than the wage demanded by the lowest-wage seeking individual.

[Kathy G.]

Chris, in the competitive model, the labor supply curve at the *market level* is upward-sloping, but the labor supply curve at the *firm level* is flat, and equal to the wage. So at the firm level, the model does indeed predict that if the firm cuts wages, all workers will immediately quit. And I was speaking about what happens at the firm level -- I said "your employer," not "all employers."

[Chris Hammond]

Ah, I did not read carefully enough. Thanks for clarifying -- my mistake.

[Robert]

It is foolish to ask ask for "a model ... that would permit Card and Krueger's finding that employment had risen after a minimum wage increase to be an actual relevant datum rather than an obvious fluke." Some of us know that Card and Krueger did not only examine a natural experiment. They also did a meta-analysis of previous studies.

I am actually published as showing that wages and employment cannot be explained by the interaction of well-behaved supply and demand curves, even under assumptions of perfect competition. I think of my originality as being more in a low-dimension diagram than in this decades-old (but ignored) conclusion.

[A Texan in Bavaria]

Ok, here's an example of monopsony a bit higher on the wage scale: working as software systems engineer/systems administrator/whatever the employer can bill the government the most for calling you for a few specific military simulations systems.

We are well-paid, but not as much as some of the more generalized IT folks. When the contract I work under changed hands, we were all offered our current pay OR LESS to join the new employer. As most of us wished to stay in one of the best-paying jobs a non-EU-citizen can get around here, we accepted - the new firm did not negotiate much.

However, if we all quit, our employer would lose their contract with the military, as we are actually difficult to replace in bulk. If we had all been able to set aside whatever anti-union prejudices many of us have, we might have been able to push for retaining or improving our previous pay packages. One at a time, however, it's not that hard to get someone up to speed, so it's rather difficult to negotiate a significant increase in pay. People with fewer geographic constraints jump to the more lucrative contracts in Poland, Lithuania, etc. as they come available.

[Karmakin]

Good writeup. The problem with trying to fit employment models into conventional economic models, is that quite often they don't fit. When is where the misunderstanding regarding the minimum wage comes in.

In the real world, employers, in order to operate their business require X units of labor (not in terms of time, but in terms of work that needs to be done), and they buy that labor at the lowest cost they possibly can. There are different things at play, slack seasons and busy seasons, etc.

Employers generally don't keep everybody on, especially retail/service economy keep hours flexible, which is bad for workers (and their families), but allows them to cut hours to lower their cost as much as possible.

In any case, the reason why raising the minimum wage adds jobs, is that traditionally, the paid value of labor is much less than it's actual value add. As long as the minimum wage doesn't go over the typical value add, it doesn't make sense to cut the job...especially when the job NEEDS to be done for the business to operate.

Theoretically, business owners have kept extra help in order to maintain high customer service levels, but over the last decade or so, they've been trying to get blood from a stone. Most businesses are running as lean as they can be in the first place.

Won't it result in higher prices? Again, prices are often determined by competition and not cost. As long as the competition doesn't go below the cost price (and so often these days it does in the case of loss leaders...), then prices are in the hands of competition, and probably Wall Street has more to say about prices than anything else.

Finally, Atrios had a comment about specialized knowledge on his blog (that's how i got here)...but it's even worse than that. Quite a few companies are in the process of standardizing/compartmentalizing previously professional work, so it either can be outsourced or done by those down the educational ladder so to speak. As well, your experience is really meaningless when it comes to a different job, as their system is going to be completely different.

I think it'll come to the point where education is less about knowledge, and more about class for 75-80% of the jobs out there.

[crack]

What's bizarre is in successive posts you've critiqued the two women that most make me question my media spending habits.

I can't stand that the Atlantic pays MM. Though she's less offensive than CP qualitatively, she tries to make up for that with quantity.

There are plenty of women out there who are more deserving of being paid by media outlets for their writing.

[hellblazer]

There are plenty of women out there who are more deserving of being paid by media outlets for their writing.

My lord, there are talking horses who are more deserving of being paid by media outlets for their writing.

[horatius]

There are plenty of women out there who are more deserving of being paid by media outlets for their writing.

I wonder if there's a similar model to explain the pathetic incompetence that permeates the current media structure. I demand statistics on the median intelligence of the modern day columnists(for newspapers only). You'll all be surprised by the results.

[crack]

@hellblazer

Really, everyone knows that all the talking horses already write for HotAir.

[mq]

Kathy, look at Card and Krueger's book "Myth and Measurement". As I commented in the other thread, that book contains a reasonable model that combines high turnover with hiring/training costs to get a less than infinitely elastic labor supply curve. All you need is some dependence of the turnover rate on the wage and costs of turnover, and there you are. So high turnover can help lead to a monopsony-like situation.

[Margalis]

Anyone who rags on Camille and Megan gets a thumbs-up from me.

Seriously though, well written and informative post.

I don't know much about economics and I've started perusing some economics blogs. What I've discovered is that Megan is not in any way informative, doesn't seem to know much, and writes meandering nonsensical prose. Then I read something like Delong or Calculated Risk: not in the same ballpark, not even the same sport.

[Genore]

in the competitive model, the labor supply curve at the *market level* is upward-sloping, but the labor supply curve at the *firm level* is flat, and equal to the wage. So at the firm level, the model does indeed predict that if the firm cuts wages, all workers will immediately quit

I'm confused. If the firm is in effect a monopsonist, then wouldn't the supply curve of labor to that firm essentially be identical to the supply curve of labor to the market. If the firm's supply curve is different than the market's supply curve, doesn't that ipso facto mean that the firm is not a monopsonist?