Jonathan Chait devotes his TRB column in the current edition of the New Republic to a critique of my inequality paper. Jonathan and I recently recorded an online chat about inequality for TNR, but I don't know if it's going to run. (I'm afraid I was rather meandering.) In any case, I'd like to address Chait's main points here.
Chait describes the paper as “a usefully honest and relatively persuasive iteration of the belief system that undergirds right-wing thought.” I don't consider myself a “right-wing” thinker, but I guess it's not really not up to me whether or not I am. My intention was to set out a liberal critique of the common liberal practice of using inequality statistics as rough measure of a society's justice. In the language of contemporary political philosophy, I argue for economic prioritarianism over economic egalitarianism. Which is not to say that I'm not commited to some kind of liberal egalitarianism. I just don't think liberal equality is primarily a matter of economic resources. In an earlier draft of the paper, there is a discussion of the varieties of equality, including the kind of concern for equality appropriate to liberalism. I've put that outtake online for anyone who may be interested in it. The key idea is that liberalism stands first and foremost for the equality in the distribution of coercive political power.
OK. On to the meat of Chait's critique:
Wilkinson begins by pointing out that, while the gap between how much the rich and the non-rich earn has exploded, the gap between how much the rich and the non-rich consume has remained fairly stable. And that's true. But Wilkinson misunderstands the implications of this fact. “Suppose you made a million dollars last year and put all but $50,000 of it in a shoebox,” he writes. (He must have enormous feet.) “Now imagine you lose the box. What good did the $950,000 do you?”
Wilkinson's point–money only has value if you eventually spend it–may be true. Yet most rich people don't put their money in shoeboxes. They invest it so they, their children, or young trophy wives can one day spend even more of it. And, indeed, the gap in wealth (how much money you have) has grown even faster than the gap in income. Meanwhile, the middle class has tried to keep pace with the rich by spending beyond its means, sending average household debt skyrocketing. Tell me why this should make us feel better about inequality?
The shoebox example is meant simply to illustrate the idea that income that is never consumed contributes little to an individual's economic well-being. This is not to say that savings has no utility beyond future consumption. There is certainly some benefit in knowing that you could consume at a higher level now or that you have the means to ensure a decent level of future consumption in the case of a loss of income. But the main benefits of income, and of saved or invested income, flow from consumption.
Chait is right that rich people (and not-so-rich people) don't put their unspent money in shoeboxes; they tend to invest it. Savings (low-risk, low-return investment) and investment mostly amounts to deferred consumption. Maybe some rich people spend most of the gains from their investments on their children and “young trophy wives,” though I doubt it. The important thing to note at this point is that you can make money from your money. That this is so, and the reason this is so, is important for addressing Chait's next argument.
Now, I don't think any of us really know that average household debt increased because “the middle class had tried to keep pace with the rich by spending beyonds its means.” If it's true, the middle class should probably cut it out. I'd note that government inducements to borrow money and buy houses might also have had something to do with it.
Moving on… I argue in the paper that real consumption inequality may have been narrowing (because, in a nutshell, inflation has been higher for the rich than the poor). Chait responds:
Wilkinson is inadvertently bolstering the strongest liberal argument againstinequality: it's inefficient. In case you're unfamiliar with this argument–as Wilkinson seems to be; he doesn't rebut or even mention it anywhere in his paper–it runs like this: Taking money from the rich and giving it to the poor helps the latter more than it hurts the former (at least until you create serious work-incentive effects, a point which most liberals think we're not close to). Wilkinson is saying the rich are getting little (in the case of luxury goods like refrigerators) or zero (in the case of real estate and higher tuition) actual benefit from their rising incomes. So why not take some of that income away and use it to buy extremely useful but currently unaffordable things for the non-rich, like, oh, basic medical care?
I am familiar with the argument from the diminishing marginal utility of consumption. I should have at least included a footnote. If I had, I would have pointed readers to “On the Utility of Equal Shares,” Chapter 23 of David Schmidtz's Elements of Justice. Schmidtz shows that “transferring a dollar from someone who needs it less to someone who needs it more can be unjustified even from a strict utilitarian perspective” — even while granting the assumptions that “(1) marginal utilities smoothly diminish, (2) all are known to have the same utility function, so interpersonal comparisons are easy, (3) redistribution is costless, and (4) there are no incentive problems whatsoever.”
It's a strong argument. It turns on the fact that the next dollar can be devoted to economic production as well as consumption. When the return (in utility) to investment in production is greater than the return from anyone's consumption, utilitarianism forbids using the next dollar for consumption.
Of course, in the real world redistribution is a “leaky bucket.” It costs money to collect a dollar. It costs money to transfer a dollar. And taxes and transfers certainly do affect incentives. The fraction of the dollar left for consumption at the end of the transfer varies a great deal from place to place and depends on a lot of things. But there are many real-world scenarios in which the fraction is so small that even a modest return from investment in production can rule out utility-maximizing redistribution.
Anyway, this is all too abstract to be very useful. The point is that diminshing marginal utility does not clearly imply progressive redistribution in a world of production. As it happens, I'm no utilitarian, and so I don't think redistribution must be utility-maximizing to be justified. The interesting question is whether a particular redistributive program would actually help people who need help at a justifiable cost. My main point in the paper is that arguing about a not-very-useful abstraction like income inequality wastes time and energy that would be much better spent arguing about the merits of specific policies.
I argue in the paper that there is very little evidence that rising income inequality has made it any more likely that the rich, as a class, would capture the democratic process and use it to consolidate their advantages. One piece of evidence I offer is the fact that, during the period of rising income inequality, voters in the top decile became more likely to support Democratic candidates, who are much more likely than Republican candidates to support progressive redistribution. In particular, Obama did better with high-income voters than McCain, despite McCain's attempt to characterize him as a “socialist” on the basis of Obama's explicit intention to “spread the wealth.” Chat replies:
One liberal complaint about inequality holds that it increases the political influence of the rich, thereby locking in even more inequality. Wilkinson scoffs at this prospect, pointing to rich voters' support for Barack Obama over John McCain. Oddly, Wilkinson confines his analysis to campaigning and pays no attention to governing. While it's true that many rich people used their money to help bring about Democratic control of Washington, every day brings a new example of the rich using their money to ensure that Democrats pose the least possible harm to their interests.
I emphasized elections, because liberals tend to emphasize elections. I'm glad to see that Chait doesn't deny that a majority of top decile voters in the last election presidential election supported the candidate who promised greater redistribution. I think our remaining difference is not very big.
The “Inequality Road to Serfdom” argument–the argument that income inequality past a certain critical threshold leads to plutocracy–assumes that economic classes have conflicting political interests, and that many or most votes express the economic class interest of voters. I provided evidence that this is assumption is false. Ronald Inglehart shows that as people become wealthier, they tend to become less likely to vote according to economic self-interest. The move of the rich toward the pro-redistribution Democratic Party illustrates this general point.
I emphatically agree that the major owners and executive managers of large auto manufacturers, investment banks, arms manufacturers, etc. tend to be very rich people who are inclined to deploy the ample resources of their firms to rig legislation and regulation to their firms' (and thus their) advantage. Corporate and interest-group “rent-seeking” is the bread and butter of classical liberal political economy. However, this ugly process isn't well characterized as a situation where the very rich, as a class, coopt state power to consolidate their fortunes against eveyone else. What's going on is that some very rich people are using state power to screw over other very rich competitors. This conflict over the reins of power does tend to make everyone but the winners worse off. And it does tend to make the winners richer than they'd be in a less corrupt system. Rent-seeking and regulatory capture probably do increase income inequality. But income inequality isn't the cause of the problem. The underlying problem is an excess of legislative and regulatory discretion. The solution is to take political power off the auction block by limiting the power of the state to intervene in the economy. If Chait and I have a substantive disagreement, it's probably here.
If I may quote David Schmidtz once more: “If selling X for a dollar is bad, we go after people who sell X, not people who have a dollar.”
I think Chait misunderstands my position in this passage:
The deeper problem with Wilkinson's argument is that it assumes the natural correctness of all market-based outcomes. This is a premise you either take on faith or don't, and which undergirds most of his argument. Wilkinson assumes that inequalities arising from the market are inherently fair. Therefore, he asserts that just about the only unjust forms of economic inequality are those that spring from non-market circumstances: “[I]t's not enough to identify a mechanism of rising inequality. An additional argument is required to show that there is some kind of injustice involved.”
I certainly could have made my thinking more explicit in the paper, but my argument does not assume the natural correctness of all market-based outcomes.
First, I don't consider advanced market institutions to be “natural.” Indeed, I think they are rather unnatural, which is why they've been missing from all but the last few moments of human history. Second, advanced markets are extremely complex structures of interaction governed by law, regulation, social norms, and more. I think it is very hard to make a clear distinction between market and non-market circumstances, which is why I don't make one. Different market systems operate according to different “rules of the game.” These rules make up a large part of the “basic structure” of society. Following John Rawls, I take the “basic structure” to be the primary subject matter of the theory of justice.
Whether a market-based outcome is morally objectionable depends on the justice of the rules that tend to produce those outcomes. I agree with Rawls and Hayek that justice is primarily a question of procedure, not pattern. And I agree with Hayek that it is a basic mistake to think a market-based outcome can be “correct” in any moral sense. I am convinced that we each have strong reason to endorse a market system that tends to allocate resources to their most valuable uses, since that kind of market system is a necessary part of a larger system of institution in which each person has the best chance of prospering. But the pattern that emerges from an ideal system of market rules will largely reflect the vagaries of supply and demand, accidents of fortune, and other morally neutral forces. The constantly evolving pattern of holdings emerging from a good system of market rules may loosely track some regulative ideal of efficiency, but it certainly won't track virtue, desert, or need. If we find that redistribution is required to arrive at a system that does the best it can for everyone, then justice will demand redistribution. But the point of redistribution isn't to correct some flaw (e.g., too much inequality) in the pattern of incomes. The point is to make sure everyone has reason enough to affirm the system in which they will live their lives.
This is where I'm coming from. My commitment to a strongly proceduralist conception of justice is what drives the emphasis in my paper on the justice or injustice of the mechanisms (the rule- or norm-governed regularities in human action) that produce the pattern of incomes.
I'm a bit puzzled that Chait thinks I assume the correctness of market outcomes, or that I think all systemic injustice comes from outside the market setting. I explicitly and rather emphatically deny that the status quo U.S. system delivers justice. For example, I said this:
There is overwhelming reason to believe that in the United States the deck really is stacked against some people. As a consequence, many millions of people are doing much less well than they might be. Legions of inner-city kids consigned to abysmal public schools are systematically denied a fair chance to develop the capacities need to participate fully in our institutions, or to enjoy their potentially ample rewards. The United States imprisons a larger share of its citizens than any country on Earth, literally disenfranchising hundreds of thousands of men and women (though they are mostly men) and leaving hundreds of thousands more dispirited and damaged. Undocumented immigrant workers increasingly constitute a permanent economic underclass explicitly denied many of the basic legal protections of citizens, which invites both government and private abuse. And at the level of culture, patterns of private discrimination continue to constitute for millions a web of real, seemingly inescapable barriers to opportunity and achievement and help to generate self-reproducing patterns of diminished expectations and wasted potential. We should focus all our attention and energy on the task of rectifying these vicious injustices. Maybe fixing all this would decrease the variance in national incomes. But the idea that fixing all this somehow requires “fixing” the pattern of incomes is an excellent way to avoid the real problem and fix nothing.
I wouldn't call the absence of a competitive market for primary education accessible to everyone a “non-market” circumstance. The fact that millions of foreign workers participating in U.S. labor markets are doing so illegally obviously is not a “non-market” circumstance. The ghastly U.S. incarceration rate is largely a result of criminalizing trade in markets for certain substances. And ongoing racist and sexist labor market discrimination is clearly a market phenomenon. I suspect that Chait's determination to see my argument as “right wing” made it hard for him to correctly interpret my actual argument.
I've gone on way too long, so I'll say just one more thing. I don't sense in Chait's piece much of a defense of of the idea that income inequality really is important. He seems to me mostly concerned to defend the permissibility of redistribution. But I'm not arguing against the permissibility of redistribution. If I'm right that income inequality isn't itself a problem, then it does follows that reducing income inequality through redistribution doesn't solve a problem. But if a redistributive program is the best solution to an actual problem, there's a good chance I would support it. I'll say it one more time: We'd do more good arguing about which policies would best rectify existing injustices rather than arguing about epiphenomena like income inequality.