Questions that are rarely asked

If monopsony power is an important feature of the labor market, and monopsony power should be prevalent when firms are bigger and therefore have a larger share of the local industry, then why do big firms pay more than small firms? The small mom and pops should be closest to operating in a competitive labor market and have little bargaining power, but they pay less. Maybe the productivity effects of big retailer outweigh the monopsony effect, but that just is another way of saying it’s not as an important feature of the market.

That is from Adam Ozimek.


Almost as if this never happened -

'Back in January, I wrote about “The Techtopus” — an illegal agreement between seven tech giants, including Apple, Google, and Intel, to suppress wages for tens of thousands of tech employees. The agreement prompted a Department of Justice investigation, resulting in a settlement in which the companies agreed to curb their restricting hiring deals. The same companies were then hit with a civil suit by employees affected by the agreements.

This week, as the final summary judgement for the resulting class action suit looms, and several of the companies mentioned (Intuit, Pixar and Lucasfilm) scramble to settle out of court, Pando has obtained court documents (embedded below) which show shocking evidence of a much larger conspiracy, reaching far beyond Silicon Valley.

Confidential internal Google and Apple memos, buried within piles of court dockets and reviewed by PandoDaily, clearly show that what began as a secret cartel agreement between Apple’s Steve Jobs and Google’s Eric Schmidt to illegally fix the labor market for hi-tech workers, expanded within a few years to include companies ranging from Dell, IBM, eBay and Microsoft, to Comcast, Clear Channel, Dreamworks, and London-based public relations behemoth WPP. All told, the combined workforces of the companies involved totals well over a million employees.'

Immigration will fix that right up!

It's only bad until it can be linked to a lefty cause.

It would be interesting to study whether monopsony power is exerted more against the poorly paid, such as retail workers, or against the relatively well-paid such as baseball players before 1975 or movie stars in Golden Age Hollywood before Olivia de Havilland won her lawsuit or computer engineers while Steve Jobs was running amok. My impression is that people who have what seem like glamorous, high salary jobs can get exploited by employers with some degree of impunity because the public isn't very sympathetic to workers in these situations. The nearly century-long reign of the "reserve clause" in professional baseball, for example, was grotesque, but the usual public response was that ballplayers should be happy to be paid anything to play a kid's game.

Sure it happens to doctors with insurers. Problem is everyone pays. Premium payers need to realize that insurers hassling doctors and making it hard for them to get paid costs them too.

Baseball owners in 1975 look generous in comparison to the NCAA today.

It's the principal/agent problem. Small mom-and-pop operations are managed by the owners, who have a clear incentive to pay as little as possible to their staff. Large companies have multiple layers of management, so staff are paid what the manager thinks they should be paid, not what the owner(s) think. This is particularly relevant when the owners are uninvolved shareholders. The manager himself has an incentive to pay his staff more, since he can then negotiate his own salary higher ("I'm their manager, I should be getting at least 25% more than them").

The question has such a simple, and obvious, answer that one has to wonder why Cowen is asking it. But the reason why the question is rarely asked is because it's a pretty dumb question.

Isn't this also another way of saying that monopsony power is not as an important feature of the market?

Which market? Bureaucratic/principle agent problems might dominate in the market of larger firms, while monopsony power does for smaller.

Interesting thought: there is a real wage effect for unions. However, unions can and do disproportionately target large, well-known firms. Perhaps the union wage effect is at least partially a result of being able to identify places that do have those kinds of principle agent problems? Public organizations being a key example.

Also if the firm underpays, the manager has to work harder to recruit and retain good employees. The manager's bias would generally be to overpay, except where the manager's status (e.g. compensation or continued employment or reputation) is linked to the profitability of his operations net of compensation.

@Andrew M and direct replies to him: Yes.
And more -
Simple problems of scale. Jo's shop of wonders only needs 3 employees out of the 5 million in the metro area - not only can Jo personally attend to this, but an Ad that says "middling low pay for pleasent but boring but very steady job" will get several useful replies out of 5 million. What's more, Jo can fill a reasonable part of the staffing with members of Jo's household - so it's Jo, Pat, their child Sam and Sam's spouse making up 4 of the 6.

Giant corp needs, like, 7,000 entry level people out of that same 5 million. The job will be every bit as boring, but because Giant corp has all of the formal rules and legal constraints of any large entity in the US, it will also be generally unpleasent - meaning the same person doing the same job at Giant corp will likely put up with more random irrelevent crap. For the same boring but steady job. And the Ad that says "boring steady respectable job with lots of pointless headaches, pay is Great!" may not get 7,000 replies even in the city of 5 million. And every person related to everybody in senior management and the BOD will not possibly add up to 7,000 people...

Dale Mortensen, Alan Manning and others in the not literature pride themselves on having a better explanation for the 30 to 50% large firm wage premium. paying higher wages attracts more applicants.

This conclusion rests on random matching rather than balanced matching been the way in which workers and employers are paired together.

The entire monotony literature collapses in a heap if there is balanced matching rather than random matching in the labour market

Right! Matching theories of job search/wage bargaining are superficially like monopsony theories in that they imply that outcomes depend upon the relative bargaining power of the participants, that the actual distribution of bargaining power is often weighted in favor of employers, and that wage floors can offset initial bargaining advantages. The difference is that matching theories are driven by imperfect information, which induces inefficiently high search and bargaining costs and, in extreme cases, missing markets, rather than coercive market power. Wage floors standardize expectations, reducing price-search on the part of both low-wage job seekers and low-wage employers, thereby, increasing productivity and low-end wages without significantly affecting profits. At the same time, floors can be inefficiently high, not just too low. Teulings (2003), for example, argues that wage floors are now inefficiently low – that, indeed, the rise in wage inequality in the lower half of the income distribution in recent times is entirely due to the fall in minimum wages –, but nevertheless acknowledges that there is an optimal level beyond which one
wouldn’t want to go.
Teilings, Coen. 2003. The Contribution of Minimum Wage to Increasing Wage Inequality. The Economic Journal 113/490: 801-833.

See my post above, and now multiply those scale problems by the "random" factor.

The older I get, the more I realize how much "random" connections matter - who you work for, who you hire both in your "day job" and elsewhere, who you have romantic relationships with, etc., are all highly random and often turn on small things even when elaborate searches are going on.

Slugger Mickey Mantle's annual salary under monopsony conditions with the New York Yankees in the 1950s and 1960s is pretty funny. See the bottom table at:

I particularly like how the Yankees cut his salary from $70,000 to $60,000 after his off-season in 1959 in which he fell all the way to third best player in the American League.

Mickey had signed with the Yankees out of high school for less money than other teams were offering. The Yankees told him that he'd make up the difference off his World Series check.

I saw Mr Mantle play. I was told that he was very good at it.

I also like how Mickey in 1957 hit .365 / .512 / .665 and the Yankees gave him a raise from $60,000 to $65,000.

I also like how the Boston Red Sox cut Ted Williams' salary from $90,000 in 1951 to $50,000 by 1956. Then in 1957 when Williams hit .388 / .526 / .731, the Red Sox gave him a raise back to $60,000.


And it's not like 1957 happened to be a year in which the ball was juiced: Mantle and Williams were just playing in their own universe that year.

Ah, the Yankees... In the seven seasons from 1950-1956, during which the Yankees won five World Series, their catcher Yogi Berra was no lower than fourth in the AL Most Valuable Player voting, winning three MVPs. In 1957 he fell to 14th in the MVP, so the Yankees cut his pay. The next season he was 18th so they cut his pay again. Then he clawed his way back to 12th most valuable player in the league, so they let him have the same salary.

The funny thing is that when the arbitrator threw out the Reserve Clause and free agency finally came along in the mid-1970s, many people were convinced baseball would be ruined.

Here's economist Dave Berri's obituary for Player's Union leader Marvin Miller:

Here's one more funny story: In 1953, Augie Busch buys the St. Louis Cardinals and and the St. Louis Browns moved to Baltimore. That season Stan Musial of the Cardinals hit .337 with 30 homers and 113 RBIs, 127 runs scored, 53 doubles, and 105 walks, so the St. Louis Cardinals cut his salary from $75,000 to $57,500. Musial responded by hitting .330 with 35 homers and 126 RBIs, so the Cards cut his salary again.

How could you not make money owning a baseball team under monopsony?

Half a dozen comments in a row? Get your own blog.

He has his own blog--just no one reads it.

Well, they all did make money, obviously, why are you asking? And why are you spending so many pixels decrying something that was so obviously bad, but hasn't been in effect for 40 years?

Is there another side to this argument you are making?

Also, even in the modern era of huge player salaries all the teams make money...maybe not year in and year out (although most do) but every owner sells their team at a very nice IRR when they sell.

It's like posting over and over how crazy it was that at one time we used to have slave labor in this country, can you believe it?

"Is there another side to this argument you are making?"

Read on and you'll find out.

Steve, my suspicion is that the big winners from the reserve clause era were fans. As recently as the 1970s, I could buy a bleacher ticket at Wrigley Field for $1.

I thought you were a man of the people.

Workers still have a lot less bargaining power at mom and pop's grocery store than mom and pop do. Wages for low paid jobs are probably dictated more by social norms and living costs than any kind of negotiation.

This is what I'm wondering.

Let's say you have twenty M&P stores in an area that pay on average $15 an hour. You also have 3 M&P stores that fill niches where the wage is $10 an hour (e.g., location, type of business, etc.).

The big corporation comes in, it competes with the first 20 stores, puts them all out of business, and pays $12 an hour. You still have the remaining three guys paying less, but they were not being competed with. In fact, since the regular wage for workers is not reduced from $15 to $12, the three remaining guys can maybe reduce what they pay down to $9.

In other words, seems to me any business that Walmart actually competed with no longer exists. In some communities I would guess you could say the big corporation pays a ton more than the small Mom and Pop, since the small Mom and Pop doesn't exist and therefore pays a wage of zero.

Is retail ever much of a monopsony? The retail sector is usually about the same limited percentage of the work force in most labor markets across the country, and there usually are numerous retail employers. It's not like Anaconda Copper in Butte, Montana when Dashiell Hammett was a Pinkerton.

The problem retail workers have is less that there are too few retail employers than that there are too many potential retail workers. Below upper management, retail is not a very specialized career (and the semi-specialists such as car salesmen or suit salesmen generally work on commission), so the general supply of labor is the major factor in determining their pay.

In contrast, people with highly specialized skills are more at the mercy of monopsonists, even though their lives are better overall. When Augie Busch cut Stan Musial's pay from $75,000 to $50,000, it's not like Musial had some non-baseball skill that would pay over $50,000 per year.

This is exactly right. Low-wage workers have more roughly equivalent employment options that just about any other segment of the market, because they haven't made significant investments into specialized skills. The monopsony hypothesis is weakest precisely where leftists most want it to be true, and the reality is that low-wage workers have low wages because their marginal contribution to the economy is low.


I could imagine a small town in 1950 where all the shop owners in town get together for a Rotary club luncheon every week and complain about impudent workers wanting to be paid $1.00 per hour or whatever, and thus enforcing a social norm against better pay for retail work. But I'm having a hard time in 2014 seeing how that kind of wage-suppressing cartel would work what with a large fraction of retailers being headquartered out of town and most workers anyway living in large labor markets.

In contrast, the tech industry monopsony case seems more plausible: the workers in question have somewhat specialized skills, a large fraction of the firms are in Silicon Valley; some have massive monopoly power, such as Apple, Microsoft, and Google; the most powerful firms tend to be either allies or potential rivals of other firms; and Apple was headed by a vicious, vindictive business genius who has always felt that laws and morality only apply to other people.

So, the Silicon Valley monopsony cartel organized by Steve Jobs resembles in key ways the most famous, best documented monopsony in American history, baseball before 1975: employees with specialized and potentially highly remunerative skills and a limited number of major employers who cooperate as much as they compete.

In contrast, the increasingly popular hypothesis of a shadowy monopsonistic cartel among thousands of retailers, both national chains and mom and pops, sounds implausible prima facie.

So, now, hopefully you can see the value for understanding the theory of monopsony from posting salary histories under monopsony of Mickey Mantle and some of the other most famous employees in America in 1950-1975. This much more complete data has only recently become conveniently available at Baseball Reference due to the work of economist Michael J. Haupert on the economic history of baseball:

These examples of monopsony at its best documented, where we now know both annual salaries and the employee's detailed on the job annual performance in terms of Wins Above Replacement, suggests under what conditions firms will be more inclined to monopsonistically exploit workers (e.g., tech) and less inclined (e.g., retail).

Well done, thanks. When you leave race out of it you are a really interesting blogger.

Not that there's nothing interesting to say about race, but you (as are many of us) are far too biased to be interesting there.

tyler, have you never heard of "efficiency wage" or is this just the usual disingenuousness?
larger firms presumably also have some degree of monopoly power and can better afford to mark up prices to accommodate the higher wage. they then capture the benefits of stability and higher worker productivity as increased profits. mom and pop have fewer avenues to productivity gains and no pricing power.

I assure you, Tyler has heard of efficiency wages.

That begs the question though, why would only large firms pay efficiency wages? If they really were efficient, both large and small firms would pay them.

Maybe people prefer to work for smaller companies and will do so at a discount? I know I find working for larger and larger companies more frustrating.

Maybe really good people who can charge a premium for their services work at larger companies, or people who are generalists who are not as highly specialized as those at large corporations.

Maybe they do make as much as a percentage of earnings and the small business just takes in less? I'm not sure why monopsony is itself would be that big a player here

in the same vein, maybe small firms also offer other compensation that doesn't show up in wages: greater flexibility, more support in tough times, or simply more chances to convert firm opportunities to private gain? (For example, maybe small firms lose what they save in salaries through 'shrinkage'.)

Have the Card/Krueger results, which IIRC were based on the meticulous records of large firms (fast-food chains), been evaluated against the possibility that minimum wages increased large-chain employment *because* it cleared the market of mom-and-pop competitors?

Because most people would rather be important contributors to a small operation than tiny parts of a huge machine, and getting them to opt for the latter requires monetary compensation?

This is an understandable confusion, which I shared when I was a grad student. I owe the following explanation to Orley Ashenfelter.

Think of a big firm that is locally dominant, e.g., Microsoft in Seattle. If Microsoft wants to hire top talent to commute to Redmond from, say, 60 miles away, they will have to pay enough to make it worthwhile. In the process of paying this next employee more money, Microsoft may also have to pay everybody else that has this guy's potential the same amount, even if their commutes are shorter. Thus, the marginal cost to Microsoft of this new hire will be greater than the new hire's compensation.

A startup tech firm in Seattle would not have to think about this problem.

The author of the article is making a fundamental flaw.

Large firms are in large large markets, there is more competition for labor. Therefore higher prices for labor.

In small markets, where there are small and fewer firms, there could be monopsony effect.

The classic illustration of this story is in agricultural economics.

Imagine a national cheese market...small cheese makers sell into a national market. Cheese prices are competitive nationally. But, look at the price for raw milk. Cheese is produced locally, and supplied by local farmers. Now, look for a cheese plant operating in a valley, in a market where there is insufficient milk supply to support another cheese plant, and where the farmers spend more to ship outside of this milk shed.

You could have a situation where milk prices are lower than prices in neighboring markets, even though cheese prices are competitive.

Anytime you deal with a monopsony issue, look at the input supply market. In this case, large retailers compete in large markets, not in small ones, like the small grocery store in a remote vacation community.

This has got to be a big factor with gas prices being what they are.

It's rarely asked because the answer is obvious: there is no monopsony power in low skilled labor markets. The only reason the argument exists is as a smokescreen for advocates of higher minimum wage. The left doesn't ask the question because they know answering it would demolish their own position. The right doesn't bother asking it because they'd just get accused of hating the poor.

How much of the large firm wage premium over the mom-and-pop store may be due to the former's likely lower discount/risk rate used when contemplating the potential contribution of employees over time?

Small businesses are generally pleasant places to work. Big bureaucratic corporations aren't. Ergo, the large firms have to pay a premium for the same quality of labor.

Small hedge funds pay more for the same workers as huge banks.

Have any of you worked at large firms? For the ones I have joined (always for the pay - never for interesting work) the key to them even BEING large firms is that they have some kind of profit-making "lock" on the market. They are able to get so large by not being in particularly competitive circumstances. Imagine you had the dominant operating system and productivity suite, for instance - the cash just flows in. Or that you have some giant long-term government contracts. There was a time where a good patented drug was also an attractive revenue stream to mooch.

So step 2 is ala Willy Horton, we go where the money is and siphon off that cash to the managers. You still have to kick up reasonable profits to the shareholders, but they are a diffuse group and there's still plenty of lard to dole out internally.

I realize this sounds super-cynical... But I find it weird that TC and Alex are sometimes more comfortable being cynical about elected officials and government workers than they are about folks in the private sector. In my experience modeling with great cynicism is an accurate predictor in both environments.

"the key to them even BEING large firms is that they have some kind of profit-making “lock” on the market."

Right, that's a key difference between studying economics as an undergraduate, where you learn about perfect competition typically using the example of wheat farmers, and studying business strategy as an MBA student: your MBA professors tell you that being a wheat farmer is awful. The goal of business strategy is to get some market power that you can defend and make big profits.

I like how in the cultural long run Willie Sutton and Willie Horton are turning into the same figure.

LOL. The names are too similar.

I wonder what other pairs might get conflated in the future...Tony Robbins and Tom Robbins? Willy Brandt and Willie Brown? Tom Joad and Tom Jones?

Tom Wolfe, Hunter S. Thompson, and Tom Stoppard -- in 200 years they'll all just seem like That 1960s Tom Writer and only grad students will notice any differences.

Whoops... Sorry about that mixup!
On the name topic and true story: the recent obit for Alan Greenberg caught my eye until I realized it was not Alan Greenspan.

There's no reason why it couldn't be the case that large firms are paying less than the wage that would obtain in a competitive market, and smaller firms are paying even less than that. In other words, the wage that smaller firms pay in a non-competitive market has no necessary relation to the wage that would obtain in a competitive market.

If so why would there still be involuntary unemployment?

Big firms have deeper pockets for fired workers to sue, so it pays for big firms to hire higher quality workers since getting rid of bad apples is more expensive for them.

Smaller firms should be able to fill in many profitable spots in local markets. So probably variance is very high, but many small businesses fill in marginal spots in the economy or niches, etc.

HR. Large companies screen out applicants with background issues. The rejected will accept jobs at small companies at a reduced wage as the negative information may not be discovered or revealed by a less formal interview process. This type of information may not be easy to control for in a study as for example "high school diploma" would be. Also, less opportunity for advancement at the small stores will select a less ambitious applicant pool.

A monopsony is not as monopsonistic as a model assumes. It can be broken down in the dynamic real world. Look at MSFT. And the owners may not be complacent, as a static model might imply, they may be uncertain and anxious about the future of the monopsony. Agsin, look at early Bill Gates. By hiring the best talent, even if it has to pay marginally more dollars (including by way of merger premia), the firm hopes to protect or extend its monopsonistic reach, either in scope and/or time. So in a sense the firm is sharing the productivity gains, or the rents, as you may choose to label it. In another sense, it is reinvesting in its monopsony.

The strongest monopsony today is the NCAA but I do not understand why they do not collude on coach compensation.

+1 You can't argue that a coach is a student who you can exploit as part of his training.

Comments for this post are closed