I wish that my time now allowed me to write a longer post on the problems with models of monopolistic- (and monopsonistic-) competition. I believe these models to be fundamentally flawed and misleading. But time constraints oblige me to put off the longer treatment that I envision. (I do have a paper coming out soon, in a volume edited by my GMU colleagues Pete Boettke and Todd Zywicki, in which I develop these ideas a bit more fully.) Nevertheless, I toss out a couple of relevant thoughts below the fold.
First is the language. The term “monopolistic competition” – when heard by the public, by politicians, by bureaucrats, and by young economists yet to develop the maturity required to wisely use (that is, to not misuse) theories – suggests that there’s some real monopoly in play in the realities supposedly captured by this model. But no such true monopoly exists. There are no barriers to entry. There are, it is true, costs that prevent the creation of a world as ideal as that which a clever theorist can conjure in his or her imagination if he or she succeeds in assuming away the perfectly innocent costs that give nearly all firms some ability to raise their prices without losing all of their customers. But cost is not monopoly, and so there is nothing remotely like genuine monopoly in so-called “monopolistically competitive” markets. Entry is free, and the process of competitive rivalry eventually reduces all firms’ profit rates to “normal” – that is, to rates consistent with competition.
In monopolistically competitive markets, no consumer is harmed – where “harm” is defined to mean ‘made worse off than he or she would be given unavoidable resource and technology constraints.’ The consumer who chooses not to buy a Mercedes because the market price of that Mercedes is much higher than is the price of a Toyota is not “harmed”; the consumer who would buy a Mercedes but for the fact that a tariff raises the price of that car higher than it would otherwise be is harmed. The former result is the consequence of the unhappy reality of resource scarcity (or unavoidable cost); the latter result is the consequence of government-created monopoly protection for domestic auto merchants.
It is a tremendous conceptual and practical error for economists to have gone along with Edward Chamberlin’s term “monopolistic competition” to describe real-world firms that face downward-sloping demand curves. The fact that a true monopolist does face a downward-sloping demand curve for the product it sells is not the essence of monopoly. Therefore, the fact that nearly all real-world competitive firms face downward-sloping demand curves for the products they sell does not in any way mean that those firms have even as much as a smidgen of monopoly power. A downward-sloping demand curve is neither a source nor a necessary sign of monopoly power.
What is true of monopolistic competition is equally true of monopsonistic competition. The upward slope of an input- (e.g., a labor-) supply curve – while a characteristic of the market confronted by genuine monopsonists – is neither a source nor a necessary sign of monopsony power. Yet because even firms that must compete vigorously to hire workers typically confront upward-sloping supply curves for labor are conventionally said by economists to possess some “monopsony power,” uncritical economists mislead themselves (and others) into supposing that the real world is infected nearly everywhere with genuine and harmful monopsony power. The word (“monopsony”) – applied uncritically and carelessly to real-world markets – generates ‘theoretical’ and empirical conclusions about reality that are wildly mistaken.
Second (and related to the above remarks). Imagine two employers of low-skilled workers. Each of these firms has some ability to lower the hourly compensation it offers to workers without having all of its workers immediately – or even eventually – quit. The textbook-infatuated economist says “That is an instance of monopsony power. That’s how we economists define it! Therefore, this labor market is infected with some monopsony power.”
What, though, is the source of this alleged monopsonistic power? In the monopsonistic-competition model the source is neither collusion nor government-erected barriers to the hiring of workers. Instead, the source is simply the reality that it’s costly (although not remotely impossible or even necessarily difficult) for every worker to quit his or her existing jobs if employers slightly lower these workers’ pay. In the same way, however, it’s costly for employers to fire current employees if these employees slightly increase, on their own, the amount of leisure they take on the job. Should we conclude from this latter fact that employees possess some monopoly power in the labor market? If so – and consistency with naive textbook models suggests that we should indeed reach such a conclusion – we’re left with the odd result that employers have some quantum of monopsony power over workers while workers have some quantum of monopoly power over employers. That’s an odd and questionable result.
Suppose that one of these two hypothetical employers of low-skilled workers sometime in the past improved its working conditions explicitly because it wanted to make itself a more attractive place for people to work. Perhaps it hired a more friendly supervisory staff, built a spacious and nicely appointed lunch room, or adopted a policy of being more lenient with workers who are occasionally late to work. The other employer did no such thing.
What will be one of the effects of this first employer’s worker-friendly steps? Answer: a greater willingness of workers to remain employed with that firm if that firm cuts its monetary hourly wages a bit. In other words, this firm that improves its working conditions will gain (or so many economists say) some “monopsony power” over workers! (Note the analogy with monopolistic competition, under which the source of the ‘monopoly’ part is product differentiation – that is, changed product attributes that cause consumers to prefer the new “differentiated” version of product over the former version of the product.)
See the oddity of this result? A firm that takes steps that every sensible person would describe as competitive – here, a firm that makes itself a more attractive place for people to work – gets some “power” to then, if it wishes, lower its monetary wages without risking losing all of its employees. And such a firm is then described as possessing “monopsony power” even if the long-run consequence of its competitive efforts to improve its working conditions do nothing to raise its rate of profit above the competitive rate.
Of course, this firm’s rivals will, in response, also likely try to improve their own working conditions in order to better compete for workers with this first firm. Such non-price competition will eventually reduce the first-firm’s “monopsony power” (as it increases the quantum of “monopsony power” held by the first-firm’s rivals who successfully improve their own working conditions). Such improvements in working conditions, though, do not happen instantaneously, and different firms will experiment – with different degrees of success – with different modes of improving their working conditions. At any period of time as this unending competitive process plays out, nearly every firm will have some ability (“power,” as it is misleadingly called) to cut its monetary wages somewhat without risking the loss of all of its employees. The textbook-literal economist then concludes, unwisely, from this fact that monopsony power is a common feature of labor markets.
More generally, changing jobs, even for low-skilled workers, is not costless. And competition by employers on any or all of the available margins for workers is also not costless. Firms X, Y, and Z will not compete as quickly or as successfully for workers as will firms A, B, and C. Even for low-skilled workers, potential employers will not be perfect substitutes for each other, if only because the costs of creating firms to meet each of the many different preferences of workers is never zero. It is a fundamental error, however, to mistake the consequences of such unavoidable costs – along with the consequences of the fact that competition plays out, in reality, in real time and is not instantaneous – for monopsony or monopoly power.