David R. Henderson  

The CEA's Mixed Thinking on Labor Market Monopsony, Part I

Yet Another Reply to Huemer on... Laureate Lessons for Education...

Monopsony graph.jpgThis month, the President's Council of Economic Advisers issued a 21-page report on labor market monopsony. It's titled "Labor Market Monopsony: Trends, Consequences, and Policy Responses." As with many reports issued by economists who are paid by government, it has a mix of good and bad ideas, and good and bad proposals.

Economist Adam Ozimek has a nice short analysis of the report. I'll go through the report seriatim, making my own comments or reporting Ozimek's where I think he is on target, which is virtually all the time.

But I'll do so in a post tomorrow. First, for those who don't understand monopsony, a quick recap of the economics of monopsony. Monopsony is monopoly on the buyer's side. The monopsonist uses his monopsony power to pay less than otherwise. Economists who write about monopsony typically are writing about monopsony in labor market, and the CEA report is no exception. The problem with monopsony, from an efficiency point of view, is that instead of hiring laborers up to the point where the value of marginal product (VMP) equals the wage, the firm hires up to the point where the VMP (assuming the firm has no market power in the output market) equals the marginal factor cost (MFC). And the MFC is above the wage. So, in monopsonistic equilibrium, the firm is not hiring "enough" labor. That is, the wage rate is below the VMP.

See the graph above. Why is the MFC above the wage? Because the supply curve faced by the firm is upward-sloping. Workers further down on the supply curve have lower supply prices than workers higher on the supply curve. (The supply price is the minimum price for which a worker is willing to supply that particular hour of labor.) The firm knows that when it pays an additional dollar per hour, say, to entice that higher-supply-price employee, it will need to, assuming it pays every employee the same, pay a dollar more per hour even to the lower supply price employees. Thus the cost of an extra dollar an hour to that last employee is more than a dollar an hour to one employee. The equilibrium employment is Q* and the equilibrium wage is W*. The efficient level of employment, by contrast, is Qefficient with the corresponding Wefficient wage.

A solution to the efficiency problem would be for the monopolistic firm to price discriminate, paying more to workers with higher supply prices and less to workers with lower supply prices. That way, the firm would come closer to the efficient amount of employment.

So, for example, in the graph above, the employer could pay each employee his supply price plus a penny. Each potentially employee would accept. This is called "perfect price discrimination" and is quite rare. What would the new equilibrium be? An amount of employment equal to Qefficient.

Keep that in mind when we analyze the CEA report.

Short of perfect price discrimination, one way to come close to Qefficient is to segment the labor market, paying a lower wage to the lower supply price segment and a higher way to the higher supply-price segment. Keep that in mind also when we analyze the CEA report.

To be continued.

Comments and Sharing

COMMENTS (4 to date)
Don Boudreaux writes:


I look forward to your follow-up post on this matter.

It's important to note that an employer that truly possessed monopsony power in the labor market would not have to raise the wages of its current workers in order to hire additional workers. The reason is that, well, that employer has monopsony power: its workers have, by assumption, no good alternatives; so if their employer hires another worker just like them but pays that new worker more than it pays to any of them, they will not quit. Perhaps these workers might become more likely to shirk - which would constrain the employer's ability to practice wage discrimination - but at the very least monopsony power calls into question the validity of the assumption of an MFC curve sloping upward so much faster than the supply-of-labor curve. More here.

Miguel Madeira writes:

"It's important to note that an employer that truly possessed monopsony power..."

Think about "monopsonist competition" (several employers, but each one offering different jobs - if only from the psychological point of view of the employees) instead of monopsony.

BC writes:

"Monopsony power" is called "market impact" in financial markets: when a market participant is too large a fraction of the market to avoid affecting market prices from his or her own trading. Is there any *direct* evidence of employer monopsony power, e.g., when America's largest employer Walmart tries to hire more employees, does that by itself raise overall market wages?

Interestingly, market impact is usually viewed as something to be avoided by the market participant when possible. It wouldn't seem favorable to the employer that trying to hire one more worker raises all workers' wages (which presumably is why the employer hires fewer workers than in a competitive market). It seems like an employer would rather see a labor market with many more employees than he needs willing to work at the market wage than one in which he has monopsony "power".

Don Boudreaux writes:

Miguel Madiera: If I correctly interpret the unstated conclusion of your counsel to think about 'monopsonistic competition' in the same way that we think about monopolistic competition, I agree completely that such an exercise is both appropriate and useful. Putting aside any of the many theoretical nits (some big!) that one can pick with the theory of monopolistic competition, it and monopsonistic competition are indeed analogous. One, if it exists, exists in output markets; the other in input markets. They are very much analogous to each other.

Among the insights about labor-market monopsony that we gain by heeding your counsel:

(1) just as, say, a laundry-detergent supplier who is a monopolistic competitor does not necessarily earn excess profits, an employer who is a monopsonistic competitor does not necessarily earn excess profits; whatever excess gains the monopsonistic employer might at first get are competed away if the output market in which that employer operates is reasonably competitive;

(2) therefore, just as in a monopolistically competitive market a government-imposed price ceiling that is perfectly set to lower the output's price to what it would be if the firm had no monopolistically competitive power would inflict losses on some firms and, thus, cause some firms to go out of business, a minimum wage perfectly set to what the equilibrium wage would be if employers had no monopsonistic power will, if those employers have no monopoly power sufficient to keep their output prices above average total cost, cause some firms to go out of business and, thus, might well reduce the employment of low-skilled workers below the level it attains without a minimum wage;

(3) just as the monopolistic-competition model too often misrepresents competitive benefits enjoyed by consumers as pure costs borne by consumers in the form of monopolistically higher prices - such as, for example, when the monopolistically competitive firm 'buys' greater inelasticity in its demand curve by offering something of value to consumers, such as an attractive and quality-bonding brand name - in the monopsonistic-competition model worker benefits can easily be misrepresented as pure costs borne by workers in the form of monopsonistically lower wages. Employers might well 'buy' greater inelasticity in their supply curves of labor by differentiating their workplaces or employment contracts from those of other employers in ways that are so attractive to workers that employers thereby create, in the supply curves of labor that they face, upward slopes, such as in the textbook monopsony way depicted above in David's post.

Comments for this entry have been closed
Return to top