Friday, February 7, 2014

On the marginal productivity of the 1%

The debate rumbles on. The rise in top incomes, particularly striking in the English-speaking democracies, is starting to become a central theme in debates about equality and fairness, shifting attention away from the perennial redistributive conflict between the middle and poor (ably fuelled by the right here in Britain) towards the spectacular gains made by a tiny group of the superwealthy.

As well as hitting mainstream politics - even Obama gave top billing to inequality in his State of the Union address - this debate has penetrated into academic economics. The reaction is interesting to behold. Whilst liberal (in the American sense) economists such as Krugman and Stiglitz argue that the growth in top incomes is in large part the fruit of rent-seeking and overbearing political influence, most economists are reluctant to go so far, invoking instead the role of exogenous forces such as globalization and technological change.

There is a certain irony to this, as economists are usually instinctively awake to the possibility of capture of political institutions for the purpose of subverting markets and winning excessive returns through regulatory manipulation, fiscal tricks and straightforward bribery. Indeed, the traditional skepticism of many economists towards the benefits of government interventionism owes a great deal to their suspicion that government interventions are very likely to favour politically powerful producer groups rather than consumers, a suspicion formalized with great success by the public choice school. So it is puzzling that the doubling of the income share of the richest one hundredth of society over little more than quarter of a century is seen as such a natural development by much of the economics profession.

To the extent that top incomes shares are not seen as a sign of something malfunctioning in our economy, they are generally viewed as broadly reflecting the relative productivity of different groups in the income distribution. Typical of this view is the recent article by Greg Mankiw in the Journal of Economic Perspectives, gamely entitled 'Defending the One Per Cent', which illustrates the risks of meddling with big gains for top earners by positing an imaginary world in which everyone is equally productive and paid the same, until someone like Steve Jobs comes along to shake up the distribution by producing something innovative that everybody wants. Mankiw suggests that 'this thought experiment captures, in an extreme and stylized way, what has happened to US society over the past several decades'. Mankiw therefore concludes that not only would redistribution be bad for incentives, stopping the next Steve Jobs in his tracks, but it would be morally unfair. The earnings of 'superstars' are 'just desserts'.

Mankiw's article provoked an irritated response from Robert Solow, amongst others. There are a large number of ways in which the argument can be challenged theoretically and empirically, and I'll probably deal with them in future posts. But just for now, here's one: the just desserts arguments suggest that earnings have a strong correlation with marginal productivity, and that marginal productivity reflects a combination of innate ability and talent, and hard work. Luck and structural privilege are not supposed to come into the equation very much, or otherwise 'just desserts' start to look hard to justify, and the gains from privilege could probably be taxed without any real loss to the economy, since they don't reflect any real productivity advantage. Except we know that luck and structural privilege are important. For a start, in unequal societies social mobility is lower, suggesting that inequality feeds on itself and prevents talent rising to the top. But to my mind the most striking forms of structural privilege here are gender and ethnic discrimination. Although some diehards still cling to the notion that women or particular ethnic groups are innately inferior, and therefore deserving of lower rewards in the workplace, these ideas are no longer taken seriously by researchers. Yet the one per cent is disproportionately occupied by men, and white men at that. I took a fairly unscientific sample by scrolling down the top 100 in the Forbes rich list of billionaires here. Only ten of them were women.

If women are less likely to be superstars, then either women are inherently less deserving, or there is something other than talent and hard work at play in determining who gets to the top (and no, it is apparently not the choice to spend more time raising a family). And if the latter is the case, then the outsized gains reflect, at least in part, something other than marginal productivity. The typical example would be the CEO of a large company, whose material impact on the company's earnings is generally hard to establish with any reliability. If the CEO were to leave, another would take their place, capturing similar earnings, but it is next to impossible to establish the differences in marginal contribution (and after all, company performance is typically uncorrelated with CEO returns).

Ironically, the debate about desserts seems to stand or fall on how optimistic we are about the power of the market. Conservative economists, strangely, tend to suspend their keen eye for rent-seeking, privilege and market subverting power when it comes to the top one per cent, preferring instead to imagine that the returns to skills over the recent period have just happened to be concentrated in the very top of the income distribution. Yet the obvious riposte is that if the state is so permeable to special pleading, corruption and political manipulation, a wealthy elite should surely be the first to take advantage of this permeability, particularly since the mass organizations - such as political parties and trade unions - that secured redistribution downwards are in terminal decline.

Maybe we are supposed to believe that the one per cent are not only smarter, but so morally unimpeachable that they will restrain from interfering with the market system that has served them so well? I think we should ask Carlos Slim about that one.