Health Care and Income Inequality

Gary Burtless and Pavel Svanton of Brookings add to the pile of reasons income inequality statistics are misleading. Here's the abstract to their paper “Health Care, Health Insurance, and the Relative Income of the Elderly and Nonelderly“:

Cash income offers an incomplete picture of the resources available to finance household consumption. Most American families are covered by an insurance plan that pays for some or all of the health care they consume. Only a comparatively small percentage of families pay for the full cost of this insurance out of their cash incomes. As health care has claimed a growing share of consumption, the percentage of care that is financed out of household incomes has declined. Because health care consumption is more important for some groups in the population than others, the growth in spending and changes in the payment system for medical care have reduced the value of standard income measures for assessing relative incomes across age groups and across the income distribution. More than a seventh of total personal consumption now consists of health care that is purchased with government insurance and employer contributions to employee health plans. In this paper we combine health care spending and insurance reimbursement data in the Medical Expenditure Panel Study with cash and noncash income data in the Current Population Survey to assess the impact of health insurance on the distribution of income and, in particular, on the age profile of income. Our estimates imply that gross money income and disposable cash and near-cash income significantly understate the resources available to finance household purchases. The estimates imply that a more complete measure of resources would show less inequality than the income measures that are currently used. The addition of estimates of the value of health insurance to countable incomes reduces measured inequality in the population and the income gap between young and old. Standard income measures imply that households with an aged household head have significantly lower average and median incomes than households with a head who is less than 55. In contrast, an income definition that includes the value of health insurance implies that aged households have higher incomes than households with a nonaged head. [Emphasis added.]

HT: Peter VanDoren

Poisoned Well of Blood Guilt by Association: A Very Special Full Disclosure

A frenzied Mark Ames at AlterNet attempts to discredit Megan McArdle's writings on health-care reform by pointing out that her father has worked for the government and that she writes for the Atlantic! !!! It's truly hysterical. Ames has done an immense service to critical thinking instructors the world over by piling so many argumentative fallacies into one neat package.

Well, maybe that's the wrong attitude. After all, how can we really evaluate Ames' argument before we know more about his parents and the circumstances of his childhood? And who am I to cast stones? I must admit that I too have parents. I suppose I ought to take this opportunity to expose my own hypocrisy before muckraking geniuses like Ames expose them first. So here we go. This isn't going to be easy. Ahem!

So, my father had a long career as a police officer. There. It's out. During my childhood, the taxpayers of Independence, Missouri and Marshalltown, Iowa put bread on our table and clothes on my back. Indeed, in Marshalltown (and, after I left home, in Council Bluffs, Iowa) my father was the chief of police. Nevertheless, I consider illegitimate many of the laws police officers are charged to enforce, and thus I believe police officers to be guilty of serious moral transgressions that differ from criminal assault, kidnapping, theft, etc., only in the sanction of the state and approval of the bulk of society. Yet, given the fact that policing paid for my piano lessons, any opinions I might have about police abuses of power, or about the state generally, pretty clearly refute themselves.

Perhaps worse, I learned to read and write in public schools, yet I use those very skills at a libertarian think tank (!!!) where I sometimes argue against the status quo system of public education. It is truly a wonder that I am able to sleep at night.

Actually… Doesn't Megan suffer from insomnia? She does. So what more is there to say? Just this: greater government control over the health care system poses no threat to medical innovation or individual liberty. Anyone who would tell you otherwise probably has parents.

New at Cato Unbound: Jorge Castañeda on the War on Drugs in Mexico

Check out the hot-off-the-Wordpress lead essay for the new edition of Cato Unbound by former Mexican Secretary of Foreign Affairs Jorge Castañeda. Here's his gist:

In his lead essay, Jorge Castañeda observes that the consequences of the U.S. drug war fall unevenly on Mexico. The U.S. taste for drugs — and for prohibition — are the chief causes of drug-related crime in Mexico, he asserts. This creates a problem that Mexico cannot solve on its own. U.S. assistance has been insufficient, and Mexican resources are too few to take on the drug cartels effectively. Even if the resources were available, the militarization of life in Mexico would be politically unacceptable to most Mexicans, who have enjoyed a relatively tranquil military in contrast to many other Latin American countries. Another approach to the war on drugs would simply be decriminalization, but again, Mexico cannot unilaterally decriminalize, because it would face severe diplomatic consequences from the United States and possibly become a refuge for addicts. The United States must lead the way toward solving this problem, which is of its own making.

The Value of Savings

Riffing off my response to Chait, Free Exchange's Washington-based blogger writes:

Suppose you made a million dollars and you put all but $50,000 of it in a shoebox. Now suppose that you never lose the box, never spend it, and leave it all to the dog when you die. What good did the $950,000 do you? If one derives pleasure from imagining consumption possibilities but never actually consumes anything, does that count as value derived from consumption? What if the wealth is public knowledge, and it generates an attitude of deference among those who respect the wealthy or hope to profit from association with them? What about the value of security? Does the presence of a large, cash barrier between you and financial disaster count as a gain derived from consumption (given that the barrier represents the ability to consume post-disaster)?

I don't mean this as any kind of criticism of Mr Wilkinson, I'm just wondering to what extent it is true. How much do people enjoy having money just because they enjoy having money?

I've been thinking about this a good deal. One way to look at savings, as Free Exchange suggests, is to see it as just another kind of consumption. This would help make sense of misers who compulsively hoard. When you get a dollar, you can exchange it for something else to consume or keep it and consume the utility of having a dollar over and over and over. I suppose one might say that differences in savings at the same level of income reveals a difference in time preference or risk aversion or estimates of lifetime income or money fetishism or a mix of these. (How hard is it to tell which it is?)

I think all this makes sense, but it's not very helpful. You can't eat dollars, live in them, get an education or much entertainment from them, etc. I think the insurance value of savings is really significant. How do we estimate it? Measure the difference in concentration of stress hormones at different savings levels? I also thing that the status value of savings can be significant, too. But this seems likely to be lower than the status value of effectively signaling wealth, which is just as likely to correspond to huge debt.

I'd guess there's a great deal of variety in peoples' attitudes toward savings and debt. Both sides of the ledger are morally valenced for many people. I'm a weirdo who reads too much economics, so I see my accumulated human capital as serving much of the insurance function of money savings. So, despite the fact that I remain a net debtor in money terms (student loans!), I feel well in the black. Some people are ashamed of debt, because it's debt, and hasten to wipe it out. Others act like credit is free money and run up debt until it explodes in their face. Then they go bankrupt. And then, later, they do it again. Some Spockish types will make minimum payments on debt indefinitely, as long as the interest rate on the debt is lower than the interest rate on investment or the value of present consumption. Some people require a cushion of savings for minimal peace of mind. Others are happy as long as their checking account doesn't dip below zero before the next paycheck. Etc. So I think it's probably hard to draw a really useful generalization about the intrinsic utility and disutility of savings and debt.

Since it remains that most savings is intended to finance future consumption (my account with the largest positive balance is a 401K retirement savings account, and I'd bet that's pretty typical), it seems best to keep the concepts of savings and consumption separate for analytical purposes — even if some of the value flows of savings seems a bit like the value of consumption; even if some forms of consumption have, in addition to the direct value of consumption to the consumer, savings-like value. Here I'm thinking of money spent improving a skill that is fun but pays, or money spent on capital goods that are also a source of enjoyment, like my computer here, or money spent on assembling a meaningful collection that appreciates in real market value. It's complicated! I think it's enough to say that the value of consumption isn't the only source of value in life, and that the value of consumption isn't even the only source of economic value in life. Nevertheless, real lifetime consumption remains far and away the best proxy for lifetime economic welfare.

Why I Love Scott Sumner

I love his long post on my paper. But much more than that, I love that he's willing to write a post denying scientific realism by way of arguing for abolishing the concept of price inflation.

That said, I wish Scott would give up in his Rortyian antirealism, which is false. It may also help to point out that, however entertaining and stimulating Rorty is to read (I am a fan in this regard), he is among the worst possible guides to questions about realism, truth, and knowledge.  As an alternative pragmatism, I would recommend Susan Haack's. I'd espcially recommend to Scott chapter nine of her Evidence and Inquiry, “Vulgar Pragmatism: An Unedifying Prospect,” which gives it to Rorty good and hard. I'd also recommend Michael Devitt's Realism and Truth or his review essay on “Scientific Realism” from the Oxford Handbook of Contemporary Philosophy, embedded below.

That said, I think Scott is largely right that the usefulness of a deflator is a function of what you want to use it for. But it's wrong to say there's no fact of the matter about the average rate of price inflation, though it may be true that is not a very useful fact. I think the fact of “neurodiversity,” as Tyler Cowen calls it in his delightful book, pushes us toward the conclusion that each person faces her own personal rate of inflation. Perhaps increasingly so. One implication is that a single rate calculated on price changes in a single “typical” consumption bundle isn't very informative. This necessarily overlooks the very real distributive consequences of new products or quality improvements which affects some people immensely while affecting others not at all. (For a stark intuition-priming case, think of the value to Alf of a new medical innovation that saves his life but was unavailable a month prior at any pirce. Now think of the value of the same innovation to Betty who suddenly dies when a safe lands on her head.) There are CPI-(U-RS, etc.) “real wages” and then there are really real wages, which rise fastest for people who get the most out of new products and quality improvements, or from new innovations in retail that push down prices (Wal-Mart) or that reduce the costs of matching buyers and sellers in secondary markets (Craiglist's or Ebay). Calculating the average change in really real wages for a group of individuals may be intractable in principle or (as I suspect) it may be a mere technological problem. In any case, the more diverse the preferences and consumption patterns of the group one is averaging over, the less meaningful the application of a single deflator.

I'd like to see more interest in exploring personal inflation rates and individual level changes in really, real wages. It seems to me that a combination of credit and debit cards records together with consumption-focused experience sampling ought to be able to make a good start.

Michael Devitt. Scientific Realism

More on Declining Marginal Utility: Reply to Yglesias and Clarke

Matt Yglesias and Conor Clarke seem to see the same upshot as Chait in the idea that rich people spend way more than poor people for stuff that isn't that much better. Here's Yglesias:

[T]he point here is that the marginal utility of money income declines as it grows. This is also a strong argument for believing that redistributing money from wealthy or high-income individuals to the poor or to public services will be welfare-enhancing. The difference, in welfare terms, between a Sub-Zero refrigerator and an Ikea refrigerator is much smaller than the difference in welfare terms between having health insurance and not having health insurance. So a surtax on high earners that goes to finance expansion of health coverage to the working poor is making people better off. In that case, when we look at statistics indicating skyrocketing income inequality we’re seeing evidence of inefficiency that can be rectified through the policy process.

Here's Clarke:

If rich families are spending an additional $10,650 for fridges that will offer little in the way of “lived difference,” this does not suggest to me that all is peachy in the U.S. of A. It suggests that those rich families have $10,650 lying around that could be redistributed at little “lived” cost!

So, if Will's argument is that income differences in the United States increasingly fail to translate into commensurate differences in consumption, why isn't this just a brilliant argument in favor ofredistributing more income than we already do? Perhaps Will Wilkinson of the Cato Institute should be a hero to liberals everywhere!

I think Conor's right that I should be a hero to liberals everywhere, because I'm an awesome liberal! But seriously. First of all, there's all the stuff I said in my reply to Chait. Everybody go and read David Schmidtz on DMU. If you think DMU is a utilitarian argument for progressive redistribution, you might really be logically committed to a utilitarian argument for higher levels of wealth inequality! The problem might be that the rich consume too much and save too little. It might turn out to be inefficient for the rich to spend ten thousand bucks on a fridge. But a much smaller fridge and several thousand in an index fund might do more for utility overall than a transfer. DMU might be the basis for an argument about tax incidence, not tax levels.

Don't forget about the way luxuries typically become democratized. That rich people were willing to buy plasma TVs for thousands of dollars several years ago has a lot to do with the fact that I was able to buy an even better plasma TV for $600 in June. Goods that can be mass manufactured are by definition non-positional. But the timing of the consumption of newly introduced goods can be positional. Eager early adopters subsidize the rest of us.

Also, Keynesians should be careful not to suddenly forget their beliefs about the downstream demand-side effects of consumption. Progressive redistribution targeted at low hedon-per-dollar luxury consumption should be expected to reduce aggregate demand. The rich will shift away from consumption toward savings. And some portion of the tax collected will be lost in the journey of the leaky bucket.

Most importantly, utilitarianism is false. I don't know about Conor, but I know Matt and I disagree about this. Like Rawls, I think the fact that utilitarianism is completely indifferent to the question of whether an individual's income and wealth is or is not a result of exchange according to fair procedures is one of the main reasons it is false. How we came to have what we have matters. Utilitarianism says it doesn't matter. So utilitarianism is false. As far as I'm concerned, the main reason you can't just take my TV or take the money out of my wallet and give it to somebody who would get more out of it is that it's my TV, it's my money. It's not yours to redistribute.

Of course, both property rights and the power to coercively redistribute property require justification. Suppose we've justified both. It remains that when gains are not ill-gotten, the state's exercise of a legitimate power to redistribute requires justification; it must not be an undue infringement or limitation of property rights. If one among several revenue-equivalent tax schemes harms the taxed less than other schemes, but helps the recepients of transfers just as much, that's a good reason to prefer it. I suspect that a consumption tax, which discourages luxury consumption and encourages productive investment, may be preferable to a revenue-equivalent income tax for DMU reasons. But DMU doesn't deliver the gotcha argument Chait, Yglesias, and Clarke seem to think it does.

Anyway, you've got to love the heads I win, tails I win attitude that seems to be at work here. If I'm wrong, and real consumption inequality has increased a great deal, tax-loving liberals will take that as a reason to tax high-income individuals at a higher rate. Justice demands we close the gap! If I'm right, and real consumption inequality has not increased much, because the effective rate of inflation for the rich has been so much higher than the effective rate for the poor, tax-loving liberals say that just goes to show why we should tax high-income individuals at a higher rate and redistribute more. The principle of utility demands progressive redistribution!

One more point. I know Matt thinks the U.S. spends too much on defense. So why argue for a new tax on the rich to finance health coverage when the program could just as well be financed by changing our budgeting priorities? This would obviously be far superior in terms of both utility and justice.

Response to Jonathan Chait on Inequality

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.