Oct 21, 2004
Platonic Productivity
[cross-posted at Austro-Athenian Empire]
Women on the job market make, on average, 75 cents for every dollar men make for the equivalent jobs.
What explains this wage gap? Various possibilities have been suggested. But some Austrians have argued that there is only one possible explanation: women are less productive than men.
The argument goes like this: If employers pay an employee more than the value of that worker's marginal revenue product, the company will lose money and so will be penalised by the market. If employers pay an employee less than the value of his or her marginal revenue product, then other companies can profit by offering more competitive wages and so luring the employee away. Hence wage rates that are set either above or below the employee's marginal revenue product will tend to get whittled away via competition. (See Mises and Rothbard for this argument.) The result is that any persistent disparity between men's and women’s wages must be due to a corresponding disparity between their marginal productivities.
As Walter Block puts it:
I'm not sure why this argument, if successful, would show that worrying about the wage gap is a mistake, rather than showing that efforts to redress the gap should pay less attention to influencing employers and more attention to influencing marital norms. (Perhaps the response would be that since wives freely choose to abide by such norms, outsiders have no basis for condemning the norms. But since when can't freely chosen arrangements be criticised -- on moral grounds, prudential grounds, or both?)
But anyway, I'm not persuaded by the argument, which strikes me as [ominous pause for effect] more neoclassical:-o than Austrian, in that it ignores imperfect information, the passage of time, etc. I certainly agree with Mises and Rothbard that there is a tendency for workers to be paid in accordance with their marginal revenue product, but the tendency doesn't realise itself instantaneously or without facing countervailing tendencies, and so, as I see it, does not license the inference that workers' wages are likely to approximate the value of their marginal revenue product -- just as the existence of equilibrating tendencies doesn't mean the economy is going to be at or near equilibrium. I would apply to this case the observation Mises makes about the final state of rest -- that although"the market at every instant is moving toward a final state of rest," nevertheless this state"will never be attained" because"new disturbing factors will emerge before it will be realized."
First of all, most employers do not know with any great precision their workers' marginal revenue product. Firms are, after all, islands of central planning -- on a small enough scale that the gains from central coordination generally outweigh the losses, but still they are epistemically hampered by the absence of internal markets. (And I'm rather skeptical of attempts to simulate markets within the firm à la Koch Industries.) A firm confronts the test of profitability as a unit, not employee by employee, and so there is a fair bit of guesswork involved in paying workers according to their profitability. Precisely this point is made, in another context, by Block himself:"estimating the marginal-revenue product of actual and potential employees .... is difficult to do: there are joint products; productivity depends upon how the worker 'fits in' with others; it is impossible to keep one's eye on a given person all day long; etc." But Block thinks this doesn't much matter, because"those entrepreneurs who can carry out such tasks prosper; those who cannot, do not." Well, true enough, but an entrepreneur doesn't have to solve those problems perfectly in order to prosper -- as anyone who has spent any time in the frequently insane, Dilbert-like world of actual industry can testify. (The reason Dilbert is so popular is that it’s so depressingly accurate.) A firm that doesn't pay adequate attention to profitability is doomed to failure, certainly; but precisely because we're not living in the world of neoclassical perfect competition, firms can survive and prosper without being profit-maximisers. They just have to be less crazy/stupid than their competitors. Indeed, it's one of the glories of the market that it can produce such marvelous results from such crooked timber.
Even if women are not generally less productive than men, then, there might still be a widespread presumption on the part of employers that they are, and in light of the difficulty of determining the productivity of specific individuals, this presumption would not be easily falsified, thus making any wage gap based on such a presumption more difficult for market forces to whittle away. (Similar presumptions could explain the wage gap between married and single women likewise.)
Hence a wage gap might persist even if employers are focused solely on profitability, have no interest in discrimination, and are doing the level best to pay salary on marginal productivity alone. But there is no reason to rule out the possibility of deliberate, profit-disregarding discrimination either. Discrimination can be a consumption good for managers, and this good can be treated as part of the manager's salary-and-benefits package; any costs to the company arising from the manager's discriminatory practices can thus be viewed as sheer payroll costs. Maybe some managers order fancy wood paneling for their offices, and other managers pay women less for reasons of sexism; if the former sort of behaviour can survive the market test, why not the latter?
I should add that I don't think my skepticism about the productivity theory of wages is any sort of criticism of the market. The tendency to which Austrians point is real, and it means that markets are likely to get us closer to wages-according-to-productivity than could any rival system. (Since neoclassical perfect competition is incoherent and impossible, it does not count as a relevant rival.) If employers have a hard time estimating their workers' productivity (the knowledge problem), or sometimes cannot be trusted to try (the incentive problem), that's no reason to suppose that government would do any better. Employers are certainly in a better (however imperfect) position to evaluate their employees' productivity than is some distant legislator or bureaucrat, and they likewise have more reason to care about their company's profitability (even if it's not all they care about) than would the government. So there's no reason to think that transferring decision-making authority from employers to the State would bring wages into any better alignment with productivity. People in government are crooked timber too, and (given economic democracy's superior efficiency in comparison with political democracy) they're even less constrained by any sort of accountability than private firms are.
Nothing I've said shows that men and women are equally productive; it's only meant to show that, given prevailing cultural norms and power relations, we might well expect to see a gap between men's and women's earnings even if they were equally productive (which is at least reason for skepticism about claims that they are not equally productive).
I would also add that even if there are persistent problems -- non-governmental but nonetheless harmful power relations and the like -- that market processes do not eliminate automatically, it does not follow that there is nothing to be done about these problems short of a resort to governmental force. That's one reason I'm more sympathetic to the labour movement and the feminist movement than many libertarians nowadays tend to be. In the 19th century, libertarians saw political oppression as one component in an interlocking system of political, economic, and cultural factors; they made neither the mistake of thinking that political power was the only problem nor the mistake of thinking that political power could be safely and effectively used to combat the other problems.
As I have written elsewhere:
Women on the job market make, on average, 75 cents for every dollar men make for the equivalent jobs.
What explains this wage gap? Various possibilities have been suggested. But some Austrians have argued that there is only one possible explanation: women are less productive than men.
The argument goes like this: If employers pay an employee more than the value of that worker's marginal revenue product, the company will lose money and so will be penalised by the market. If employers pay an employee less than the value of his or her marginal revenue product, then other companies can profit by offering more competitive wages and so luring the employee away. Hence wage rates that are set either above or below the employee's marginal revenue product will tend to get whittled away via competition. (See Mises and Rothbard for this argument.) The result is that any persistent disparity between men's and women’s wages must be due to a corresponding disparity between their marginal productivities.
As Walter Block puts it:
Consider a man and a woman each with a productivity of $10 per hour, and suppose, because of discrimination or whatever, that the man is paid $10 per hour and the woman is paid $8 per hour. It is as if the woman had a little sign on her forehead saying,"Hire me and earn an extra $2 an hour." This makes her a desirable employee even for a sexist boss.The fact that the wage gap does not get whittled away by competition in this fashion shows that the gap must be based, so the argument runs, on a real difference in productivity between the sexes. This does not necessarily point to any inherent difference in capacities, but might instead be due to the disproportionate burden of household work shouldered by women -- which would also explain why the wage gap is greater for married women than for single women. (Walter Block makes this argument also.) Hence feminist worries about the wage gap are groundless.
I'm not sure why this argument, if successful, would show that worrying about the wage gap is a mistake, rather than showing that efforts to redress the gap should pay less attention to influencing employers and more attention to influencing marital norms. (Perhaps the response would be that since wives freely choose to abide by such norms, outsiders have no basis for condemning the norms. But since when can't freely chosen arrangements be criticised -- on moral grounds, prudential grounds, or both?)
But anyway, I'm not persuaded by the argument, which strikes me as [ominous pause for effect] more neoclassical:-o than Austrian, in that it ignores imperfect information, the passage of time, etc. I certainly agree with Mises and Rothbard that there is a tendency for workers to be paid in accordance with their marginal revenue product, but the tendency doesn't realise itself instantaneously or without facing countervailing tendencies, and so, as I see it, does not license the inference that workers' wages are likely to approximate the value of their marginal revenue product -- just as the existence of equilibrating tendencies doesn't mean the economy is going to be at or near equilibrium. I would apply to this case the observation Mises makes about the final state of rest -- that although"the market at every instant is moving toward a final state of rest," nevertheless this state"will never be attained" because"new disturbing factors will emerge before it will be realized."
First of all, most employers do not know with any great precision their workers' marginal revenue product. Firms are, after all, islands of central planning -- on a small enough scale that the gains from central coordination generally outweigh the losses, but still they are epistemically hampered by the absence of internal markets. (And I'm rather skeptical of attempts to simulate markets within the firm à la Koch Industries.) A firm confronts the test of profitability as a unit, not employee by employee, and so there is a fair bit of guesswork involved in paying workers according to their profitability. Precisely this point is made, in another context, by Block himself:"estimating the marginal-revenue product of actual and potential employees .... is difficult to do: there are joint products; productivity depends upon how the worker 'fits in' with others; it is impossible to keep one's eye on a given person all day long; etc." But Block thinks this doesn't much matter, because"those entrepreneurs who can carry out such tasks prosper; those who cannot, do not." Well, true enough, but an entrepreneur doesn't have to solve those problems perfectly in order to prosper -- as anyone who has spent any time in the frequently insane, Dilbert-like world of actual industry can testify. (The reason Dilbert is so popular is that it’s so depressingly accurate.) A firm that doesn't pay adequate attention to profitability is doomed to failure, certainly; but precisely because we're not living in the world of neoclassical perfect competition, firms can survive and prosper without being profit-maximisers. They just have to be less crazy/stupid than their competitors. Indeed, it's one of the glories of the market that it can produce such marvelous results from such crooked timber.
Even if women are not generally less productive than men, then, there might still be a widespread presumption on the part of employers that they are, and in light of the difficulty of determining the productivity of specific individuals, this presumption would not be easily falsified, thus making any wage gap based on such a presumption more difficult for market forces to whittle away. (Similar presumptions could explain the wage gap between married and single women likewise.)
Hence a wage gap might persist even if employers are focused solely on profitability, have no interest in discrimination, and are doing the level best to pay salary on marginal productivity alone. But there is no reason to rule out the possibility of deliberate, profit-disregarding discrimination either. Discrimination can be a consumption good for managers, and this good can be treated as part of the manager's salary-and-benefits package; any costs to the company arising from the manager's discriminatory practices can thus be viewed as sheer payroll costs. Maybe some managers order fancy wood paneling for their offices, and other managers pay women less for reasons of sexism; if the former sort of behaviour can survive the market test, why not the latter?
I should add that I don't think my skepticism about the productivity theory of wages is any sort of criticism of the market. The tendency to which Austrians point is real, and it means that markets are likely to get us closer to wages-according-to-productivity than could any rival system. (Since neoclassical perfect competition is incoherent and impossible, it does not count as a relevant rival.) If employers have a hard time estimating their workers' productivity (the knowledge problem), or sometimes cannot be trusted to try (the incentive problem), that's no reason to suppose that government would do any better. Employers are certainly in a better (however imperfect) position to evaluate their employees' productivity than is some distant legislator or bureaucrat, and they likewise have more reason to care about their company's profitability (even if it's not all they care about) than would the government. So there's no reason to think that transferring decision-making authority from employers to the State would bring wages into any better alignment with productivity. People in government are crooked timber too, and (given economic democracy's superior efficiency in comparison with political democracy) they're even less constrained by any sort of accountability than private firms are.
Nothing I've said shows that men and women are equally productive; it's only meant to show that, given prevailing cultural norms and power relations, we might well expect to see a gap between men's and women's earnings even if they were equally productive (which is at least reason for skepticism about claims that they are not equally productive).
I would also add that even if there are persistent problems -- non-governmental but nonetheless harmful power relations and the like -- that market processes do not eliminate automatically, it does not follow that there is nothing to be done about these problems short of a resort to governmental force. That's one reason I'm more sympathetic to the labour movement and the feminist movement than many libertarians nowadays tend to be. In the 19th century, libertarians saw political oppression as one component in an interlocking system of political, economic, and cultural factors; they made neither the mistake of thinking that political power was the only problem nor the mistake of thinking that political power could be safely and effectively used to combat the other problems.
As I have written elsewhere:
As students of Austrian economics (see, e.g., the writings of F. A. Hayek) we know that the free market, by coordinating the dispersed knowledge of market actors, has the ability to come up with solutions that no individual could have devised. … [But as] students of Austrian economics (see, e.g., the writings of Israel Kirzner), we also know that the efficiency of markets depends in large part on the action of entrepreneurs; and on the Austrian theory entrepreneurs do not passively react to market prices (as they do in neoclassical economics), but instead are actively alert to profit opportunities and are constantly trying to invent and market new solutions. … [W]e should remember to balance the Hayekian insight against the equally important Kirznerian insight that the working of the market depends on the creative ingenuity of individuals. … I see our role … as that of intellectual entrepreneurs; our coming up with solutions is part of (though by no means the whole of) what it means for the market to come up with solutions. We are the market.We know -- independently of the existence of the wage gap -- that there is plenty of sexism in the business world. (Those who don't know this can verify it for themselves by spending time in that world or talking with those who have done so.) Once we see why the productivity theory of wages, though correct as far as it goes, goes less far than its proponents often suppose, it does not seem implausible to suppose that this sexism plays some role in explaining the wage gap, and such sexism needs to be combated. (And even if the wage gap were based on a genuine productivity gap deriving from women's greater responsibility for household work, the cultural expectations that lead women to assume such responsibility would then be the sexism to combat.) But that's no reason to gripe about"market failure." Such failure is merely our failure. Instead, we need to fight the power -- peacefully, but not quietly.