Thinking Clearly about Economic Inequality

PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1PA Masthead.indd 1 2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM2/9/06 2:08:34 PM

Recent discussions of economic inequality,marked by a lack of clarity and care, have con-fused the public about the meaning and moralsignificance of rising income inequality. Incomestatistics paint a misleading picture of real stan-dards of living and real economic inequality.Several strands of evidence about real standardsof living suggest a very different picture of thetrends in economic inequality. In any case, thedispersion of incomes at any given time has, atbest, a tenuous connection to human welfare orsocial justice. The pattern of incomes is affectedby both morally desirable and undesirable mech-anisms. When injustice or wrongdoing increases

income inequality, the problem is the originalmalign cause, not the resulting inequality. Manythinkers mistake national populations for “soci-ety” and thereby obscure the real story about theeffects of trade and immigration on welfare,equality, and justice. There is little evidence thathigh levels of income inequality lead down a slip-pery slope to the destruction of democracy andrule by the rich. The unequal political voice of thepoor can be addressed only through policies thatactually work to fight poverty and improve edu-cation. Income inequality is a dangerous distrac-tion from the real problems: poverty, lack of eco-nomic opportunity, and systemic injustice.

Thinking Clearly about Economic Inequalityby Will Wilkinson

_____________________________________________________________________________________________________

Will Wilkinson is a research fellow at the Cato Institute and editor of Cato Unbound.

Executive Summary

No. 640 July 14, 2009

3648_PA640_1stClass:3648_PA640_1stClass 7/6/2009 9:08 AM

Introduction

“We are now living in a new Gilded Age, asextravagant as the original,” says the NobelPrize–winning Princeton economist and NewYork Times columnist Paul Krugman.1 In thedays of Krugman’s youth, “the economic dis-parities you were conscious of were quite mut-ed.” But that America is “another country,”Krugman says. Once, the AFL-CIO was a fix-ture of nature, even Ike liked the New Deal,and lawyers and longshoremen felt themselvesto be peers—as men, and as Americans. Today,the income gap is as wide as it was when therobber barons built lavish mansions and keptsenators as pets. And today’s gap is just asmalign. “The United States doesn’t have Third

World levels of economic inequality—yet,”Krugman warns. “But it is not hard to foresee,in the current state of our political and eco-nomic scene, the outline of a transformationinto a permanently unequal society—one thatlocks in and perpetuates the drastic econom-ic polarization that is already dangerously faradvanced.”2 That we have so far failed to graspthe dangers attests to “the growing influenceof our emerging plutocracy,” which has bus-ied itself denying and obscuring the reality ofthe new era of inequality.3

What is to be done? Economists ThomasPiketty and Emmanuel Saez, experts on themeasurement of income inequality, havefound that “the top 1 percent income share hasincreased dramatically in recent decades.” Theyconclude that “it is obvious that the progres-sive income tax should be the central elementof the debate when thinking about what to doabout the increase in inequality.”4 ForKrugman, a corresponding bump in the open-handedness of the welfare state is also recom-mended, as is the resuscitation of the mori-bund labor movement—not to mention acampaign of disapproval aimed at the balefulchanges in culture (precipitated by the propa-ganda campaigns of so-called “movement con-servatives”) that have permitted executivesalaries to soar with neither resentment nor

shame.5 Krugman surely speaks for manywhen he argues that democracy is itself atstake:

Even if the forms of democracy remain,they may become meaningless. It’s alltoo easy to see how we may become acountry in which the big rewards arereserved for people with the right con-nections; in which ordinary people seelittle hope of advancement; in whichpolitical involvement seems pointless,because in the end the interests of theelite always get served.6

This is a dark portrait—a nightmare for lib-erals of any stripe—and its realization is to bepassionately resisted. It is in fact realized inmuch of the world, and it is a time-testedrecipe for misery and strife. But is this reallywhere we’re headed if the income gap does notcontract? For many citizens, politicians, and celebrat-

ed scholars (such as Krugman), high and ris-ing levels of income inequality are just wrong;they obviously pose a danger to the ideal of anopen-textured liberal society where disadvan-tages need not be permanent, where advan-tages of birth and good fortune do not createa self-sustaining structure of supremacy andhumiliating subordination, and where all citi-zens enjoy the respect due to free persons whoare equal under the law. But should this seem so obvious? Are the

millions who nod along with Krugman cor-rect? Is American income inequality really anexistential threat to the democratic values atthe heart of our political culture? Does itthreaten imminently to transform the UnitedStates into an irreversibly stratified illiberalregime, dominated generation after generationby the rich and well-connected?Well, no. It should not seem obvious that

American income inequality imperils justiceor threatens to gut our democracy, because itisn’t true. You don’t have to be duped by theplutocracy to find Krugman’s line of thinking,so representative of the views of left-leaningAmericans, badly misguided. Paul Krugman is

2

Is Americanincome

inequality reallyan existentialthreat to the

democratic valuesat the heart of ourpolitical culture?


without question a brilliant economist, and heis perhaps the most talented communicator ofeconomic ideas of our era. He’s the most rig-orous and forceful public intellectual oftoday’s American left. But in this paper I’llargue that Krugman-like conceptions of thereality and immorality of economic inequalityin America reflect a tangle of conceptual errorsand a mix of questionable moral assumptions. The public discussion of inequality in the

United States, and no doubt elsewhere, ismarked by a lack of clarity and care. Publicdeliberation and debate about it are thereforeconfusing, and a lot of people are confused.Few commentators—even among those whoare professional economists—speak clearlyabout what the various measures of economicinequality do and do not tell us. And it is rarelymade clear how these measures relate to whatis valuable about equality as a political ideal.What is it that is supposed to make economicinequality in general—or income inequality inparticular—so deeply worrying? Much likeconservatives who warn firmly that legalizinggay marriage will destroy the American family,many liberals warn firmly of the disasters thatunchecked income inequality will bring—without pausing to explain how the cause willcreate the evil effect. We can do better, and theaim of this paper is to show how.In what follows, I’ll seek to clarify the main

ideas involved in thinking about the realityand morality of economic inequality. Thepoint of this is to help us to

• get the descriptive story straight aboutinequality in America;• evaluate inequality according to reason-able, broadly liberal standards that areaccepted by most Americans; and• clarify the relationship between econom-ic inequality and the freedom and well-being of the least advantaged.

There are limits to what I can do in a singlepaper. For instance, I can’t offer a full accountof the value of equality, the harm of inequality,or the relevance of economic inequality tosocial justice. Even if I could, moral and politi-

cal concepts may be “essentially contestable,”which is to say that we’re going to fight aboutthem forever. One reason the fight never endsis that free societies inevitably produce wilddiversity in thinking about morality and poli-tics. If a slightly new way of filling in the mean-ing of “liberty,” “equality,” or “justice” becomespopular, it will tend to change our politics, andhundreds of millions of lives, in a way thatsome will celebrate and others will resent. As afree society, we keep fighting because it mattersto us. This paper is meant as a small sally in that

perpetual contest. Although I can’t hope tosettle the big questions, I will offer a numberof arguments meant to support one way ofthinking about equality and political moralityand to call into question another, more wide-spread way of thinking about them. My goal isto illustrate as forcefully as I can that there is amorally deep and analytically rigorous alterna-tive to the conventional way of thinking aboutthese questions. At the very least, I call intoquestion the cogency of some remarkablycommon assumptions about equality andpolitical morality that appear again and againin textbooks, media reports, and public dis-cussions.Right now, we are in the depths of a reces-

sion, and our attention is riveted to the ques-tion of reviving economic growth, so theissue of how the fruits of growth are distrib-uted has receded in salience. But it remains offundamental importance, as attitudes aboutrising inequality during the “Long Boom”will have a huge impact on the way in whichthe restructuring of economic policymakingnow under way proceeds.One last thing: Paul Krugman’s recent best-

seller, The Conscience of a Liberal, contains anaccount of the rise of what he calls movementconservatism. In effect, Krugman’s story poi-sons the well from which this paper is drawn, soit seems necessary to say something about it.“Movement conservatism,” Krugman main-tains, “is financed by a handful of extremelywealthy individuals and a number of major cor-porations, all of whom stand to gain fromincreased inequality. . . . Turning back the clock

3

The public discussion ofinequality ismarked by a lackof clarity andcare.


on economic policies that limit inequality is, atits core, what movement conservatism is allabout.”7 Krugman goes on to mention the CatoInstitute as part of the malign infrastructure ofmovement conservatism.Setting aside the many other shortcom-

ings of Krugman’s historical narrative, hisaccount of the rise of institutions like mineseems to leave no room for authentic moralmotivation or a sincere interest in the truth.It is disappointing to see Krugman take thisroute, to call into doubt the possibility ofhonest disagreement. Still, he says nothing toimply that those with malicious hearts can-not speak truly. It is probably best to let ourlives prove our hearts, and so let me simplyinsist that an argument is good or it isn’t.And here come some arguments.

The Trend of EconomicInequality

Just about every newspaper article, editori-al, and blog post about inequality you haveever read discusses inequality in annualincomes. And it’s true that a variety of mea-sures show a sizable increase in incomeinequality since the 1970s. What exactly doesthat mean? If you take each individual’s wagesand salaries in a given year, or each house-hold’s total annual income, and plot a curvethat shows how many individuals or house-holds earned a given income, the curves fromrecent years will look stretched out comparedto the curves from a generation ago. The righttail of the curve will extend further out to theright than it once did (i.e., there are more peo-ple with extremely high incomes than in therecent past). Also, the overall shape of thecurve will look flatter (i.e., a smaller portion ofthe population is bunched at the middle ofthe curve than in the recent past).As a result, the income of the top X per-

cent of earners (take your pick: the top 10percent, 5 percent, 1 percent, 0.1 percent, oreven 0.01 percent) accounts for a higher per-centage of total national income than in thepast. The rich are getting richer. And if the

poor aren’t getting poorer, they’re at leastfalling farther and farther behind the coun-try’s income leaders.But looking at the dispersion of annual

incomes isn’t the only way to measure trendsin economic inequality. In fact, if we’re inter-ested in trends in overall material well-being,income statistics can provide a surprisinglydistorted picture.Suppose you made a million dollars last

year and put all but $50,000 of it in a shoebox.Now imagine you lose the box. What good didthat $950,000 do you? Maybe it purchasedsome temporary peace of mind. It’s certainlyreassuring to know that you have resources atyour disposal. But it likely did rather less foryour well-being than did the $50,000 youspent on housing, food, entertainment, healthcare, transportation, gadgets, toys, and so on. Why do we want income at all? So that we

can acquire things we value. The good ofincome is almost entirely in the good of con-sumption. We eat bread, not paychecks. Now,consumption tends to be measured in termsof the amount of money individuals or house-holds spend over some period of time. This isnominal consumption. It is very important tograsp that nominal consumption does notnecessarily track the value of the consumptionto a consumer. We may discover that someonehas spent a dollar, but that does not tell ushow much satisfaction, security, health, orhappiness was gained. If we’re interested in the overall material

well-being of a life, what we really want to knowis the quantity of goods and services a personhas consumed over the course of his lifetime,and the value to that person of all those goodsand services. It turns out that snapshots ofannual income just aren’t very reliable proxiesfor lifetime consumption or overall well-being.That’s because a person’s income varies a greatdeal over his life (low at the beginning of acareer, typically rising over time, then falling offin retirement). And it is quite common forincomes to fluctuate a good deal from year toyear—because of bonuses, temporary jobless-ness, a spouse entering or exiting the work-force, or the receipt of an inheritance.

4

If we’re interestedin trends in

overall materialwell-being,

income statisticscan provide a surprisingly

distorted picture.


By contrast, our consumption fluctuates con-siderably less. Because we can save, draw downsavings, or run up debt, we are able to engage inwhat economists call “consumption smoothing.”As a result, consumption in a given year tends totrack, not our income in that particular year, butour expectations regarding our long-term futureearning prospects—our “permanent income,” aseconomists call it. Accordingly, annual consump-tion figures have the potential to give us a morerepresentative picture of overall economic in-equality than do annual income figures.8

The conceptual argument for favoring con-sumption over income as a measure of eco-nomic well-being is decisive. The practical argu-ment is a bit less so. Consumption data aremore difficult to collect than income data, andthe available data sets are less comprehensive.Together with the fact that estimates of povertyand inequality tend to be significantly loweraccording to most studies that rely on con-sumption figures, this has made the interpreta-tion of consumption data a sometimes heatedsubject.9With that caveat, the weight of the evi-dence shows that the run-up in consumptioninequality has been considerably less dramaticthan the rise in income inequality. “Has U.S. current income inequality in-

creased over the period 1989–2003?” ask DirkKrueger of the University of Pennsylvania andFabrizio Perri of the University of Minnesotaand the Federal Reserve Bank of Minneapolis ina summary of their recent research. “Looking at[the Consumer Expenditure Survey] data sug-gests that yes, it did.” However, they continue,“The consumption data suggest . . . that theconsequences of this increase have not causedan increase in the dispersion of the distributionof lifetime resources; if it did it would haveshowed an increased consumption inequality.Consumption inequality, however, has re-mained substantially stable.”10

In an influential book, University of Texaseconomist Daniel T. Slesnick finds that dur-ing the 1990s consumption inequality didn’trise at all:

The widely reported U-turn in inequali-ty in the United States is an artifact of

inappropriate use of family income as ameasure of welfare. When well-being isdefined to be a function of per equiva-lent consumption, inequality eitherdecreased over the sample period orremained unchanged.11

How can income and consumption in-equality diverge? A good portion of the disper-sion of annual incomes reflects temporaryfluctuations in income; in other words, life-time or permanent income inequality shouldbe significantly lower than annual incomeinequality. Accordingly, if incomes from year toyear are growing more volatile (and there issome evidence that this is the case) but the abil-ity to engage in consumption smoothing iskeeping up (for example, through improvedaccess to credit), or if the ability to smooth con-sumption races ahead of changes in incomevolatility, then consumption inequality willgrow more slowly than income inequality.12

Nominal consumption numbers may offera less distorted picture than do income statis-tics, but they may conceal as much as they illu-minate. Records of nominal consumption cantrack only the dollars spent, but not the value—the pleasure or health or well-being—gainedthrough the spending.13 A stable, or even ris-ing, trend in nominal consumption inequalitycan mask a narrowing of real consumptioninequality—the inequality in the utility gainedfrom consumption. That is to say, real materi-al standards of living may become more equaleven if consumption inequality stays stableand income inequality rises. The difficulties involved in using either

income or nominal consumption as a reliableproxy for real economic well-being are pro-found and have motivated a large number ofeconomists to attempt to measure welfaremore directly through surveys and other self-reporting techniques. Recent work in “happi-ness research” shows that inequality in self-reported happiness, or “life satisfaction,” hasbeen shrinking over the past several decades inwealthy market democracies—the UnitedStates included.14 In a fascinating recent study,University of Pennsylvania economists Betsey

5

The weight of the evidenceshows that therun-up in consumptioninequality has been considerably lessdramatic than therise in incomeinequality.


Stevenson and Justin Wolfers find that “in-equality in happiness has fallen substantiallysince the 1970s” in the United States, cuttingacross the trend in income inequality.15 Theynote that the trend toward greater equality inhappiness stalled and began to reverse coursein the 1990s, due in part to widening inequali-ties in happiness (and wages) between individ-uals of unequal levels of education.The truly striking fact is that for decades the

level of happiness inequality in the United Statesfell simultaneously with rising levels of incomeinequality, and it remains significantly lowertoday than it was 30 years ago during the lowpoint of American income inequality. Stevensonand Wolfers plausibly attribute much of the nar-rowing in the happiness gap to the rapid gainsin social and economic status enjoyed bywomen and African Americans since the early1970s, highlighting that the sociopolitical settle-ment underpinning the much-lauded, mid-cen-tury “Great Compression” achieved significantequality mainly among white men. Self-report-ed life satisfaction is plausibly a more direct andaccurate indicator of psychological welfare thaneither income or spending, and these findingsshow that, on the whole, the quality of livesacross the income scale have become morealike—not less—since the unraveling of the GreatCompression.16 We would expect to see theopposite were real standards of living driftingoceans apart.That real inequality might remain stable, or

even decline, while the income gap explodes iscertainly counterintuitive, but it’s consonantwith both theory and fact.17 I’ve already dis-cussed how incomes and nominal consump-tion can diverge, but there are other factors atplay as well. First, if inexpensive goods improvein quality more rapidly than expensive goods,the typical bundle of goods and services con-sumed by poor families will come to moreclosely resemble the bundle typically con-sumed by rich families. To put if more breezily,if cheap stuff gets better faster than expensivestuff, the gap between cheap and expensivestuff narrows, which in turn narrows the gap inthe quality of life between rich and poor.Second, if the goods and services typically con-

sumed by poor families rise in price more slow-ly than those typically consumed by rich fami-lies—if the rich face a higher effective rate ofinflation—gaps in incomes will not reflectequivalent gaps in real consumption.18

You can see leveling in quality across theprice scale in almost every kind of consumergood.19 At the turn of the 20th century, onlythe mega-rich had refrigerators or cars. Butrefrigerators are now all but universal in theUnited States, even while refrigerator inequali-ty continues to grow. The Sub-Zero PRO 48,which the manufacturer calls “a monument tofood preservation,” costs about $11,000, com-pared with a paltry $350 for the IKEA EnergiskB18 W. The lived difference, however, is rathersmaller than that between having fresh meatand milk and having none. The IKEA modelwill keep your beer just as cold as the Sub-Zeromodel. Similarly, more than 70 percent ofAmericans under the official poverty line ownat least one car. Despite a vast difference inprice, the difference between driving a usedHyundai Elantra and a new Jaguar XJ is practi-cally undetectable compared with the differ-ence between motoring and hoofing it.20 Asimilar compression has occurred for food,clothing, and shelter. John Nye makes the gen-eral point powerfully:

Just as spices like vanilla and pepper arenow so trivially cheap that we forget thatfortunes were once made importingsuch treasures to the West, we come todenigrate if not simply ignore the vastnumber of things that ordinary peoplecan afford because they have become socheap. In some sense, fixating on mone-tary income will always overstate thesedifferences.Thus, whatever the measured gap

between the rich and the poor intoday’s world—the real (utility-adjust-ed) gap in incomes and wealth is liableto be substantially smaller than that ofa century or so earlier, even when mon-etary measures tell us otherwise.21

The vast spread of prices, and the widening

6

For decades thelevel of happinessinequality in theUnited States fellsimultaneouslywith rising levels

of incomeinequality.


range of incomes, can distract us from anoften narrowing range of experience. Thepoint is not that in America the relatively poorsuffer no painful indignities, which would beinsulting and absurd. The point is that, overtime, the everyday experience of consumptionamong the less fortunate has become in manyways more like that of their wealthier compa-triots. This is a huge egalitarian triumph. Awidescreen plasma television is a delight, but acheap 19-inch TV is enough to allow a viewerto laugh at Shrek.Unfortunately, changes in the quality of

consumer goods are difficult to measurewith great precision. Some economists aretrying hard to do this by employing sophisti-cated new statistical methods. Mark Bils ofthe University of Rochester finds that con-ventional measurement techniques haveunderestimated the quality growth of manymundane consumer durables, such as cars,televisions, furniture, home appliances, andmore.22 Bils says nothing about variations inquality, or rates of increase in quality at dif-ferent price levels. But we do know that ordi-nary folks spend a higher percentage of theirbudget on these kinds of things than do thesuper-wealthy, who spend more on travel,luxury goods, and personal services. Thesequality changes are therefore likely to meanrelatively more to lower- and middle-classconsumers, and constitute a compression inthe range of material experience. On the other side of the equation, in her

book Deluxe: How Luxury Lost Its Luster, journal-ist Dana Thomas complains that luxury goods,once made by old-world artisans according tothe highest standards of craftsmanship, havebecome shoddy, mass-market commoditieswith a huge price tag.23 They just don’t makeHermes like they used to.This compression is a predictable conse-

quence of innovations in production and dis-tribution that have improved the quality ofmany goods at the lower range of prices fasterthan at the top. New technologies and knock-off fashions now spread down the price scaletoo fast to distinguish the rich from the aspir-ing for long. Indeed, increasingly speedy down-

ward diffusion of once-dear consumer goodsinterferes with the ability of the wealthy to setthemselves apart through “conspicuous con-sumption.” The general effect of the democra-tization of luxury is to increase demand amongthe wealthy for nonmanufacturable, inherent-ly scarce “positional goods” whose signal of rel-ative socioeconomic status will not be so swift-ly diluted by broad mass-market diffusion.Think of real estate with ocean views, or IvyLeague diplomas, or goods like yachts (whichare so large and complex that they cannot bemade broadly affordable). Such goods, ofcourse, are insanely expensive. Economistssuch as Robert H. Frank tend to decry the futileinefficiency of efforts spent in zero-sum com-petition over positional goods.24 But John Nyeargues compellingly that positional competi-tion can amount to

a positive force for the democratizationof the benefits of economic growth.Thus the spending by the wealthy onmany positional goods acts as a curioussort of natural taxation. The richest (ormost ambitious) must work harder andpay more for virtually the same goods asyesteryear while their productive invest-ments (necessary to stay on top of theincome distribution) benefit the entireeconomy.25

Holding the quality of goods fixed, realconsumption inequality can decline simply asa consequence of the changing prices of thingsthe rich and poor tend to buy. To take onerecent example, Jerry Hausman of the Massa-chusetts Institute of Technology and EphraimLeibtag of the United States Department ofAgriculture show that Wal-Mart’s move intothe grocery business has driven down foodprices.26 Lower prices are found not only atWal-Mart stores, but practically everywhere, ascompetition from the retail giant forced thetraditional grocery chains to increase efficien-cy and sacrifice profit margins. And this hasn’tcome at the price of lower quality. On the con-trary, many stores attempted to competeagainst Wal-Mart by offering higher-quality

7

Over time, the everydayexperience ofconsumptionamong the lessfortunate hasbecome in manyways more likethat of theirwealthier compatriots.


fare, only to see the Bentonville behemothmatch up with goods of comparable quality,but cheaper. Because the poor spend a larger portion of

their budget on food than any other incomegroup, lower prices have benefited them themost. As a rule, when the prices of food, cloth-ing, and basic modern conveniences drop rela-tive to the price of labor-intensive services andluxury goods, real consumption inequalitydrops, too. That’s why developments thatheighten demand for (and therefore raise theprice of) status goods coveted by wealthierconsumers can counterintuitively act as anequalizing force on real standards of living. Hausman and Leibtag’s findings suggest

that poorer consumers have recently faced alower rate of inflation than have the rich,moderating the divergence of real incomes.This possibility has received further supportfrom new work by University of Chicago econ-omists Christian Broda and John Romalis.27 Itis difficult to summarize their findings morepithily than did their University of Chicagocolleague Steven Levitt, of Freakonomics fame:

How rich you are depends on twothings: how much money you have,and how much the stuff you want tobuy costs. If your income doubles, butthe prices of the things you consumealso double, then you are no better off.When people talk about inequality,

they tend to focus exclusively on theincome part of the equation. Accordingto all our measures, the gap in incomebetween the rich and the poor has beengrowing. What Broda and Romalis quiteconvincingly demonstrate, however, isthat the prices of goods that poor peopletend to consume have fallen sharply rela-tive to the prices of goods that rich peopleconsume. Consequently, when you mea-sure the true buying power of the rich andthe poor, inequality grew only one-thirdas fast as economists previously thoughtit did—or maybe didn’t grow at all.28

In particular, Broda and Romalis credit trade

with China, which has massively increased itsexports to the United States by supplying waresto huge discount outlets like Wal-Mart andTarget. “We are underestimating the gains fromtrade,” Broda said in a recent interview, empha-sizing the profound importance of what mightseem like esoteric questions of economic meth-odology:

The current statistical interpretationignores the fact that a poor householdtoday can access goods that, in the1960s, they could not—microwaves,DVDs—and, more importantly, that theprices of the staples that lower-incomehouseholds consume have also gonedown dramatically.. . . The bottom line with our study is

that we may have won the war againstpoverty without even noticing it.29

Wealthier Americans, Broda and Romalisobserve, spend a much smaller portion oftheir budgets on the things for sale at Wal-Mart and a much larger portion on servicesprovided by local labor such as home clean-ing, lawn care, psychotherapy, and yoga class-es. Because the prices of such services are rel-atively unaffected by the rise of competitiveglobal markets or advances in manufactur-ing and distribution technology, these land-mark developments in recent economic his-tory have done less to improve the bang of awealthy person’s buck. To compound matters, economist Enrico

Morreti at the University of California–Berkeleyhas found that college graduates, who tend tobe wealthier than nongraduates, prefer to live inrelatively expensive cities, which further reducesthe real purchasing power of their incomes.According to Morreti, this pattern implies that“college graduates are increasingly exposed to ahigh cost of living and that the relative increasein their real wage may be smaller than the rela-tive increase in their nominal wage.” Morretifinds that half the increase in the college wagepremium disappears when the housing costsborne by college grads are taken into account,and suggests that “the increase in well-being

8

Hausman andLeibtag’s findings

suggest that poorer consumers

have recentlyfaced a lower rateof inflation thanhave the rich,

moderating thedivergence of real

incomes.


inequality between 1980 and 2000 is smallerthan the increase in nominal wage inequality”on the basis of this fact alone.30

Fresh findings like those of Broda, Romalis,and Moretti constitute compelling evidencethat inflation-adjusted consumption inequali-ty has risen very little, if at all. Moreover, thereare good technical, historical, and experientialreasons to suspect that quality improvementswithin many of the kinds of goods that loomlarge in the budgets of poorer consumersremain underestimated by the prevailing quan-titative methods. That inequalities in real mate-rial conditions may be trending downward overtime is suggested not only by recent evidencefrom happiness research, but also by the dra-matic long-term narrowing of other, more easi-ly observable inequalities between rich andpoor, such as the inequalities in height, lifeexpectancy, and leisure. Robert William Fogel, aNobel prize–winning economic historian, hasargued that nominal measures of economicwell-being have often glossed over enormouschanges in the conditions of life. “In every mea-sure that we have bearing on the standard of liv-ing . . . the gains of the lower classes have beenfar greater than those experienced by the popu-lation as a whole,” Fogel observes.31

Taken together, the preceding considera-tions show, at least, that real economicinequality has grown far less than the incomefigures suggest. At most, they show that wehave become in many ways a more economi-cally egalitarian society, even as the range ofincomes has widened. It should now be clearthat income statistics can be a source of pro-found distortion and unnecessary confusion.Once we adopt the habit of surveying the eco-nomic landscape through the lens of real con-sumption and real standards of living, moraloutrage over income inequality and the relatedpush for a renewed regime of corrective redis-tribution simply look like mistakes. None of this is meant to suggest that

America is all roses and rainbows. Some wor-rying inequalities—for example, inequalitiesin access to a good education or to qualityhealth care—may indeed be widening, arrest-ing economic mobility and denying decent

opportunity to the least fortunate. But PaulKrugman can spare himself the night sweats,because today’s new-style Gilded Age incomegaps simply do not imply old-style GildedAge lifestyle gaps. Many enterprising Americans have indeed

accumulated vast fortunes turning out everhigher-quality goods at ever lower prices. Andthey have widened the income gap by doing it.But in the process they have also minimizedsome of the material inequalities that mattermost. If we are worried about inequalities ineducation and health care, as we should be, wemight stop to consider that these are preciselythe areas we have chosen to shield most jeal-ously from entrepreneurship and marketcompetition. Allowing profit-seeking innova-tors to compete on price and quality, andthereby to put better and more affordable vitalservices within reach of the poor, might makesome people really, disgustingly rich. And itmight also make a healthier, better-educated,more egalitarian America. If we care, weshould consider it.

Mechanisms of Inequality

Let’s pause a moment to review. If we’reconcerned about economic inequality, incomeinequality isn’t the only way, or even the bestway to measure it. The most credible definitionof economic inequality refers to the gap inoverall material well-being. By that definition,it is clear that we are far from a new era of dan-gerous invidious inequalities. With all dueattention to the Lamborghinis, NetJets, andcavernous mansions flaunted by today’s super-wealthy, the real, lived difference betweentoday’s rich and poor does not approach theshocking contrast of garish opulence and bare-foot misery that marked the real Gilded Age.Still, something is going on. As well as we

can measure, differences in people’s earningpower—in the market value of their labor—have gone up considerably over the past gen-eration. Which brings us to the questionabout income inequality: So what? Why isthis a problem that we should care about?

9

If we are worriedabout inequalitiesin education andhealth care, wemight considerthat these are theareas we havechosen to shieldmost jealouslyfrom entrepre-neurship andmarket competition.


Paul Krugman has stated that “the UnitedStates doesn’t have Third World levels of eco-nomic inequality—yet.” It turns out that, byhis own lights at least, he isn’t being pes-simistic enough! Income inequality in theUnited States is higher than in any otherwealthy nation and just slightly higher than incountries such as Russia and Burkina Faso. Asmeasured by the standard metric, the Ginicoefficient, U.S. income inequality is about thesame as in Ghana.32 If you believe that incomeinequality is a rough measure of the justice ofa nation’s social and economic institutions,then it would appear that the United Statesand Ghana are roughly on par.Yet the UN Human Development Index—a

relatively comprehensive measure of averagewell-being—ranks Ghana 136th out of 177nations, while the United States is ranked12th.33 This yawning gulf in well-being betweenthe average American and the average Ghanaianis the product of starkly contrasting systems ofsocial, economic, and political institutions.Because the United States and Ghana have thesame level of measured income inequality, wecan be certain that starkly contrasting systemsof institutions, which produce dramatic differ-ences in wealth, health, education, and longevi-ty, can also produce the same mathematicalratio of incomes between the rich and the poor. This suggests that a nation’s level of

income inequality, in isolation, tells us verylittle. It would be analytically convenient if allpossible causes of income inequality weremorally undesirable, and equally so. But itturns out that the world isn’t like that. Somecauses of inequality are less bad than others,and some are good. Indeed, because incomeinequality can be the effect of so many differ-ent causes, noting that a country’s level ofincome inequality is high or low logicallyimplies nothing at all. What we presumably want to know is

whether people are doing as well as they couldbe doing, whether people are being treatedfairly, or whether people are given what theyhave coming to them as human beings—andinequality measures alone simply don’t tell usthat. Other measures of welfare or well-being

are more likely to be informative for the pur-poses of moral evaluation and deliberationover policy. Harvard economist Louis Kaplowobserves:

A country with low inequality may haveimplemented effective policies aimed atthe poor or may have destroyed theincentives and wealth of the upper class-es, to the detriment of the poor. If onereported social welfare measures instead,one would know more. Focusing oninequality rather than welfare obscuresthe situation.34

It’s important to emphasize the point thatthe level of income inequality within a countrymay or may not be a byproduct of wrongdoingor injustice, depending on the mechanismsthat have produced it. Consider countries likeGhana. According to Branko Milanovic, chiefeconomist at the World Bank, high levels ofincome inequality in many African nations arethe result of a traditionally hierarchical socialstructure, which was reinforced by colonialism.This structure has persisted through indepen-dence, despite the egalitarian rhetoric of manysocialist African leaders in the 1960s, such asGhana’s Kwame Nkrumah. Milanovic argues“that the historically hierarchical structure ofthese societies has reasserted itself, and that thenew leaders—even those who use a ‘progressive’rhetoric—have simply reverted to the old-fash-ioned patrimonial state where concentratedpolitical power is used to acquire economicgains.”35

Now consider the United States. It is notlike that. No one thinks that the level ofAmerican income inequality was caused bysystematic political predation enabled by tra-ditional patrimonial social norms. Thoughthe level of income inequality in the UnitedStates is the same as Ghana’s, it is generatedby entirely different, and less evidentlyexploitative, institutional mechanisms. The low informational content of measure-

ments like the Gini coefficient is powerfullyillustrated by a path-breaking new study byBranko Milanovic, Peter H. Lindert, and

10

A nation’s level of income

inequality, in isolation,

tells us very little.


Jeffrey G. Williamson.36 The authors note thatthe higher a population’s mean income, thehigher the possible income inequality. Theidea, in a nutshell, is that a generally wealthypopulation is a sweeter target for plunder bythe ruling political class—the people withaccess to the coercive instruments of govern-ment—than is a generally poor population. Ifthe ruling class were able to strip a formerlyrich population of all but the means for baresubsistence, the increase in inequality wouldbe stupendous. But if the general populationwas already at or near subsistence, maximumpredation would yield relatively little increasein inequality. The authors call this upper limitof potential inequality the “inequality possi-bility frontier.” Their second key idea, the“extraction ratio,” is the distance between themaximum possible level of inequality and theactual measured level. Because potential andactual inequality can be separated, the actuallevel of inequality as measured by traditionalmethods tell us rather less than we havebecome trained to think, especially if weintend to use these numbers as a basis for themoral evaluation of a country’s institutions.Milanovic, Lindert, and Williamson write:

This new measure of inequality maycapture our notions of inequality moreaccurately than any actual measure.For example, Tanzania . . . with a rela-tively low Gini of 35 may be less egali-tarian than it appears since it has ahigh extraction ratio. On the otherhand, Malaysia . . . may have a muchhigher Gini (almost 48), but its elitehave extracted only about one-half ofmaximum feasible inequality.Another implication of this approach

is that it considers jointly inequality anddevelopment. As a country becomes rich-er, its feasible inequality expands. Con-sequently, if recorded inequality is stable,the inequality extraction ratio must fall;and even if recorded inequality goes up,the ratio may not. Thus, the social conse-quences of increasing inequality under condi-tions of economic growth may not entail as

much relative impoverishment or perceivedinjustice as the recorded Gini might suggest[emphasis added].37

This speaks eloquently to the importance ofthe mechanisms that produce inequality. Ifincome inequality in the United States is symp-tomatic of injustice, the problem is unlikely tobe the level of inequality as such, but the insti-tutional mechanisms or social norms—such aspredation by political elites or the systematicexclusion of ethnic minorities from economicopportunities—that tend to generate incomeinequality. If you believe that American incomeinequality does reflect injustice in the structureof its institutions, then it is important to iden-tify precisely where and how the system isunjust instead of simply fixating on the factthat there is inequality. If the level of inequalityis a knock-on effect of a more fundamentalinjustice, then we should focus our attentionon the original site of wrongdoing. The fire isthe problem, not the alarm. To make the point clearer, let’s look at an

example of a possible mechanism of risinginequality. A number of theorists, such as PaulKrugman and MIT’s Frank Levy and PeterTemin, point to changes in laws that havemade the organization of labor unions pro-gressively more difficult and argue that suchchanges explain part of the rising trend inincome inequality.38 Suppose for the sake ofargument that they are correct about the facts.In that case, they have successfully identified aset of institutional mechanisms that explainsome part of the increase in income inequality.Have they also identified a mechanism ofinjustice? Maybe, and maybe not. Whether you think they have depends on

what you think about the moral case for laborunions. If a certain level of state-backed bar-gaining power on the behalf of their membersis something that unions ought to have—something that unions and their members aremorally due—then demonstrating erosion inthe laws that shore up union bargaining pow-er will indeed amount to demonstrating a fail-ure to give certain people and their associa-tions what they have coming to them.

11

If the level ofinequality is a knock-on effect, then weshould focus ourattention on theoriginal site ofwrongdoing.


However, it is far from obvious that this is so.Indeed, a powerful argument from justice canbe mounted against the laws that enhancelabor bargaining power.39 If this argument iscorrect, then the decline of union power maybe a sign of a gain in justice.The point is not to rehearse or resolve this

debate, but simply to illustrate that it’s notenough to identify a mechanism of risinginequality. An additional argument is requiredto show that there is some kind of injustice orwrongdoing involved. In this case, and inmany others, there is heated, longstanding dis-agreement among well-meaning, intelligentpeople on the substantive moral question.40

It is important to recognize that, in manycases, the fact of ongoing disagreement helpsexplain the persistence of the mechanismthat accounts for some bit of inequality. Afuller consensus that the erosion of unionpower is unjust (to stick with that example)would likely be reflected in public opinionand public policy. If a mechanism of inequal-ity persists, despite well-known and well-advertised arguments that it is unjust, it maywell be because many or most people remainunmoved by the arguments to that effect.This suggests that it would be more fruitfulfor economic egalitarians to redouble theirefforts at persuasion at the level of the allegedinjustice rather than continuing to point outthat income inequality is high and rising, asif that fact speaks for itself. Here is the point at which some are tempted

to argue that interested parties have been suc-cessful in manipulating public opinion tostand on the side of injustice. For example, theattempt to depict movement conservatism andits influence as the creation and instrument ofscheming wealthy elites is an exercise in rhetor-ical needle-threading. It is an attempt to con-demn public opinion while exonerating thedemocratic public. This kind of “false con-sciousness” argument tends either toward ageneral indictment of democracy as an en-abling condition for injustice (i.e., voters can’tbe trusted to do the right thing) or toward theunverifiable claim that people’s true democrat-ic preferences would emerge were many of the

author’s other controversial policy preferenceswidely adopted. For example, some have arguedthat voters would democratically supportgreater redistribution if only the precedent ofgreater redistribution could be establishedthrough some other, perhaps nondemocratic,means—and that democratic support, unlikestatus quo public opinion, would reflect thevoters’ authentic preferences.41 Of course, any-one of any ideological persuasion can make asimilar argument when frustrated to find thathis convictions are in the minority. The “liberalmedia” often plays this role for the right, andthat is why such arguments are generally awaste of everyone’s time.42 It is both more hon-est and more charitable to suppose that the pol-icy you prefer lacks sufficient support becausemost people have yet to be convinced by thearguments for it. Even a bare majority of sup-port is a significant achievement. There is agreat variety of moral convictions in a free,diverse society such as ours. Even a broadlyappealing argument must rely on an implicitordering of values with which a sizable numberof people will reasonably disagree. The general lesson, then, is that the level of

economic inequality is a reliable indicator ofneither individual well-being nor social justice.A society’s least-privileged class can fare verywell in a highly unequal society (such as in theUnited States) and fare dismally in a highlyequal society (such as Ethiopia). Either a highor low level of economic inequality may beconsistent with justice—with people gettingwhat they are due as free and morally equalmembers of society—or it may be a side effectof injustice. In the case of injustice, the impor-tant thing is not the side effect—some level ofinequality—but its primary causes: the injus-tices where they have occurred. Take a moment to imagine a society where

even the poor do very well and there is no evi-dence of systematic injustice in its basic insti-tutions. People are free and equal under thelaw, and treated with respect simply by virtueof being people. The rules of the game are notrigged against any one group of people andeveryone has access to educational, social, andeconomic opportunities sufficient to take an

12

It’s not enough to identify a

mechanism of rising inequality.An additionalargument is

required to showthat there is somekind of injusticeor wrongdoing

involved.


active part in public life and enact a dignifiedand meaningful life. Sounds pretty good, does-n’t it?Now, suppose we discover later that this

society also contains a number of immenselywealthy people who have a great deal moremoney than the average person. Have we sud-denly discovered injustice? Is the United States an example of this kind of

society—a model of perfection with a fat dollopof inequality? Sadly, no. There is overwhelmingreason to believe that in the United States thedeck really is stacked against some people. As aconsequence, many millions of people are doingmuch less well than they might be. Legions ofinner-city kids consigned to abysmal publicschools are systematically denied a fair chance todevelop the capacities need to participate fully inour institutions, or to enjoy their potentiallyample rewards. The United States imprisons alarger share of its citizens than any country onEarth, literally disenfranchising hundreds ofthousands of men and women (though they aremostly men) and leaving hundreds of thousandsmore dispirited and damaged. Undocumentedimmigrant workers increasingly constitute a per-manent economic underclass explicitly deniedmany of the basic legal protections of citizens,which invites both government and privateabuse. And at the level of culture, patterns of pri-vate discrimination continue to constitute formillions a web of real, seemingly inescapable bar-riers to opportunity and achievement and helpto generate self-reproducing patterns of dimin-ished expectations and wasted potential. Weshould focus all our attention and energy on thetask of rectifying these vicious injustices. Maybefixing all this would decrease the variance innational incomes. But the idea that fixing all thissomehow requires “fixing” the pattern ofincomes is an excellent way to avoid the realproblem and fix nothing.

Economic Patterns andDistributive Justice

Too often those who write about incomeinequality assume that it is unfair or unjust

simply because it is inequality. If it’s badintrinsically, the evaluation of the mecha-nisms that have brought it may seem besidethe point. One common source of this con-fusion about the moral status of the incomedistribution is the ambiguity of the word“distribution.”Talk of the “income distribution” mixed

with talk of “redistribution” encourages thethought that there is someone or something(perhaps the government or a cabal of interna-tional bankers) who decide what teachers,plumbers, computer programmers, and bas-ketball players will be paid each year. This silly,but sadly widespread, misimpression is com-pounded by talk of the median worker’s dwin-dling “share” of the “national income,” as if theUnited States of America was a super-sizedfirm with profits to be bargained over anddivvied up or “distributed” among the interest-ed parties—this much to labor, this much tomanagement, this much to capital improve-ments, etc. It is simply impossible to conduct ameaningful public discussion about inequalitywithout an upgrade in conceptual and linguis-tic clarity. The income distribution is nothing more

or less than an ordered list of numbers, whereeach number represents the money value of anindividual’s or household’s annual earnings.For a list of 100 numbers, we find the topdecile by counting down to the tenth numberfrom the top and drawing a line under it. Ifour list has 300 million numbers on it, wecount down to the 30 millionth number fromthe top, and so on. You can have a distribution of anything you

can put a number on. Take height. Andre is 68inches tall, Beatrice is 70 inches, and Carlos is80. Let’s say they are members of a club. If welist their heights from tallest to shortest, thenwe have the height distribution of the club.Notice that no one distributed their heights tothem. The distribution is simply the pattern ofheights in the population. This pattern mighthave any number of interesting properties. Ifwe like, we can add together their heights (it is218 inches) and see who has what percentageof total club height. It turns out that Carlos,

13

There is overwhelmingreason to believethat in the UnitedStates the deckreally is stackedagainst some people.


who is only one-third of the club, has a full 37percent of the total height. (Is that unfair?)Imagine they add a member, Daria, who is amere 50 inches tall. The average height of theclub has just dropped significantly. Also, clubheight inequality has notably increased. But noone has become a whit shorter.The income distribution is like that. It is a

pattern. Income inequality is a property ofthe pattern of incomes. Every time a penni-less immigrant walks across the border, hechanges the pattern. Income inequality mayhave marginally increased thereby, but noone became poorer for it.The pattern of incomes emerges from bil-

lions upon billions of individual choices andtransactions. Every time you buy a candy bar,a pair of shoes, or a ticket to a concert, youhave made a tiny change in the pattern ofincomes. Nicole Kidman is fabulously wealthybecause millions of individuals have chosen tosee a movie with Nicole Kidman in it insteadof a non-Kidman movie, or instead of goingbowling. Of course, these myriad choices takeplace within a framework of political, legal,economic, and social institutions—includingcultural conventions and norms—all of whichaffect the eventual pattern of incomes. TheConstitution of the United States, workplacesafety regulations, contract law, family con-ventions, and tipping norms are all part of thebasic framework of institutions and all shapethe choices that determine the pattern ofincomes.The exception to the idea that the pattern of

incomes is not “distributed” by anyone, butemerges from countless individual choiceswithin the basic framework of institutions, isgovernment redistribution. It would be a mis-take to think of redistribution as a redo of aprior round of active distribution, which leftsomething to be desired. That would makesense only if there was a prior round to do over.But there are no rounds; the music never stops.The pattern shifts continuously, unceasingly,as a byproduct of normal human social life.Instead of thinking of redistribution as a redo,think of it as an active intervention into andrearrangement of the dynamically emerging

pattern of incomes. Think of a gardener prun-ing branches and grafting them elsewhere on atree. Redistributed income and benefits are dis-tributed. They are distributed by the govern-ment, which came to have those resources bytaking them away from someone.The confusion over what an income distri-

bution is goes beyond the common assump-tion that the pattern reflects some kind ofintentional top-down division of incomes, andbeyond the hasty inference that a rise inincome inequality reflects injustice when itmay simply reflect benign or beneficial pat-terns of voluntary transfer. Even worse, there isendemic confusion over the appropriate scopeof the distribution, which tempts us to seeinjustice where there is none while blinding usto injustice under our nose.Consider immigration. Looking at that

issue through the prism of conventional eco-nomic analysis or liberal egalitarian politicalthought tends to simply take for granted whatmight be called “analytical nationalism.” Afterall, income statistics are kept by governmentson a national level. Of course, the mere factthat most useful economic data are collectedby nation-states about individuals and fami-lies within their physical jurisdictions is irrele-vant to the task of determining the morallyrelevant pattern of incomes. If you focus onlyon the shifting pattern of incomes amonglegal residents within the statistics-keepingjurisdiction (the United States), you can easilylose track of the real story of human welfareand social justice.Consider a discussion of the effects of

immigration on income inequality from threeeminent political scientists: Nolan McCarty,Keith T. Poole, and Howard Rosenthal, intheir recent book Polarized America: The Danceof Ideology and Unequal Riches:

The new immigrants are predominantlyunskilled. They have contributed great-ly to the economy by providing low-wage labor, especially in jobs thatAmerican citizens no longer find desir-able. They also provide the domestic ser-vices that facilitate labor market partici-

14

The pattern ofincomes emerges

from billionsupon billions of

individual choicesand transactions.


pation by highly skilled people. On theother hand, immigrants have also increasedinequality both directly, by occupying the low-est rungs of the economic ladder, and indirect-ly, though competition with citizens for low-wage jobs. Yet as noncitizens they lack thecivic opportunities to secure the protections ofthe welfare state. Because these poor peoplecannot vote, there is less political support forpolicies that would lower inequality by redis-tribution [emphasis added].

This is a sadly typical example of the dis-tortions of analytical nationalism. If we wereto assume a natural and mundane moral per-spective, from which all people involved aretaken into account and assumed to have equalworth—that is, if we assume the perspective ofmoral egalitarianism—what we would see is aprofound reduction in both poverty and eco-nomic inequality. If the question is “Whathappened to the people in this scenario?” thenthe answer is “The poorest people became con-siderably wealthier, narrowing the economicgap between them and the rest.” But whatactually happened seems either invisible orirrelevant to the authors, which certainly sug-gests that their analytical framework leavessomething to be desired. Here’s how the pas-sage I highlighted might be more accuratelystated:

Immigration decreased inequality bothdirectly, by sharply increasing the wagesof low-skilled, foreign-born workers,and indirectly, through remittance pay-ments to low-income relatives at theimmigrants’ places of origin. Due to thewidespread opposition of American vot-ers to liberalizing immigration, verylarge additional reductions in povertyand inequality have been forgone.

Reading allegedly social-scientific accountsof inequality by most celebrated economistsand political scientists, one would simply notknow that nation-states are not giant firmswith profits (“national income”) to be parceledout to various constituencies, or that political

boundaries defined by histories of colonialaggression, war, and dumb luck do not definethe natural and inevitable domain of moralevaluation. These are not trivial conceptualgaffes. Once committed, they distort almostevery judgment about political morality andsocial justice. Society is not a set of people shar-ing a legal status set by local law, or a set of peo-ple inside the borders of a political jurisdiction,but the international system of cooperation weact within every day. Global air traffic patternsor shipping lanes limn the shape of society bet-ter than a civics class map of the 50 states.When you buy socks made by strangers in a far-away factory, you have entered into societywith them. You made a tiny ripple in the distri-bution, in the pattern, of national and interna-tional income and well-being. You can chooseto ignore the ripple once it crosses the border,but that doesn’t mean that questions of socialand distributive justice stop at the border, too. Analytical nationalism has serious real-world

consequences. It leads well-meaning people tocountenance, or even support, acts of injusticeagainst fellow members of our transnationalsociety—restrictions on the free movement ofpersons across political boundaries—in thename of combating the illusory injustice of anuptick in the national Gini coefficient. Thesegaffes lead Paul Krugman, for example, to tie hisconscience in a liberal knot. “I’m instinctively,emotionally pro-immigration,” Krugman con-fesses.43 But he is also instinctively, emotionallycommitted to the moral relevance of nation-lev-el income inequality statistics. Thus does a mod-est rule that tells the Census Bureau where tostop counting come to tell Krugman whose wel-fare really counts. “We’ll need to reduce theinflow of low-skill immigrants. Mainly thatmeans better controls on illegal immigration,”Krugman concludes. After all, “the net benefitsto the U.S. economy from immigration, asidefrom the large gains to the immigrants themselves[emphasis added], are small.” Of course, national jurisdictions matter.

Borders define the physical scope of legal andeconomic institutions. Differences in the qual-ity of institutions explain, among other things,the large degree of economic inequality be-

15

Political boundariesdefined by histories of colonial aggression, war,and dumb luckdo not define thenatural andinevitabledomain of moral evaluation.


tween, say, Americans and Mexicans. (It alsoexplains why Mexicans don’t worry about massAmerican immigration.) Mexican immigrationnarrows the income gap between Mexicansand Americans. Trade with China narrows theincome gap between the people of Guangdongand the people of Tulsa. Both slightly widenthe gap between some Americans and others.Why should that gap matter more? Of course,citizenship matters. Americans stand in a spe-cial relationship with other Americans byvirtue of sharing and sustaining their commoninstitutions. If governments are going to offerpublic protections and benefits, then it is nec-essary to define the relevant public. And thereare profound moral questions about what citi-zens owe to one another as citizens. But thecorrect answers to those questions cannotimply that the welfare of those with whom weare in society, but whose passports were issuedby a different political authority, matters lessthan our own.

The Inequality Roadto Serfdom

In 1944, Friedrich Hayek’s The Road toSerfdomhit the shelves in England and America.The following year, Reader’s Digest published anabridged version that brought Hayek’s caution-ary tale to an enormous audience, foreverchanging the shape of the American debateover economic policy. The Road to Serfdom is anegalitarian work penned by a liberal about thegrave danger of political inequality, among oth-er things. Rational economic planning by cen-tralized government authorities was much invogue among “respectable” intellectuals in the1940s. Hayek pointed out that this kind ofplanning necessarily requires power to be vestedin a small elite. Excellent results could perhapsjustify this concentration of power. But, Hayekargued, no matter how comprehensive theirdata gathering or rigorous their models of theeconomy, the planners would never have ade-quate information to pinpoint the most effi-cient allocation of resources. Conflicts amongthe planners would surely break out over which

plan to impose, implementation would beinconsistent, and since no plan could possiblybe adequate to the task, all would fail. The pub-lic would attribute compounding failure to alack of a unitary authority with the power tosettle on a single course of action and just get itdone. Popular demand for a “strongman”would lead to totalitarian dictatorship—thelimiting case of inequality under the law and,therefore, the utter death of liberal rights. Theargument may seem melodramatic today, butin the era of Hitler, Stalin, and Mussolini, itseemed a lot like life. Why bring this up? Because Hayek’s slip-

pery-slope logic thrives in the thought of con-temporary egalitarian liberals—from the pagesof academic journals to the editorials in localnewspapers. There is more than one road toserfdom and, according to egalitarian liberals,an excess of income inequality sets us on one ofthem. Supreme Court Justice Louis Brandeissums up the core what I will call the “InequalityRoad to Serfdom” argument when he said,“We can have a democratic society or we canhave great concentrated wealth in the hands ofa few. We cannot have both.” The late John Rawls, the dominant liberal

political philosopher of the last half of the20th century, emphasized that uncheckedeconomic inequalities will lead to politicalinequalities and status inequalities. This can,Rawls argued, threaten the liberties of theleast well-off both directly and indirectly:directly, by denying them the conditions forequal democratic representation; indirectly,via the demoralizing effect of a low relativesocial position, which may lead to a sense ofresignation and political disengagement.44

The entry on “distributive justice” in theStanford Encyclopedia of Philosophy nicely cap-tures both Rawls’ influence and the currentlydominant view in academic liberal politicalthought:

Very large wealth differentials maymake it practically impossible for poorpeople to be elected to political office orto have their political views represented.These inequalities of wealth, even if they

16

Hayek’s slippery-slopelogic thrives inthe thought ofcontemporaryegalitarian liberals.


increase the material position of theleast advantaged group, may need to bereduced in order [to secure “the fair val-ue of the political liberties”].45

The same basic point is stated more urgent-ly and concretely in Krugman’s claims aboutthe threat posed to democracy by “the emerg-ing plutocracy.”46 “Either democracy must berenewed, with politics brought back to life,”writes the journalist Kevin Phillips in his bookWealth and Democracy, “or wealth is likely tocement a new and less democratic regime—plu-tocracy by some other name.”47 This narra-tive—from income inequality, to the eviscera-tion of true democracy, to the tyranny of therich—is the contemporary liberal’s version ofThe Road to Serfdom. “Plutocracy” or a “bananarepublic with nukes” may not be as harrowingas totalitarian dictatorship, but it’s still a liber-al nightmare.48

It’s important to see that the InequalityRoad to Serfdom argument is not merely con-ceptual or philosophical, but in fact makes anumber of falsifiable empirical claims. Itmakes predictions. The chief prediction is that,past a certain threshold level of economicinequality, the democratic process will tend tolock-in and even exacerbate trends in inequali-ty by successfully resisting redistributive policy.The way this is supposed to work, in theAmerican context, is that the success of theparty most strongly supported by the poor,and which favors greater redistribution—theDemocratic Party—will be systematically un-dermined as inequality gets out of control.The advocates of the Inequality Road to

Serfdom argument therefore need to accountfor the success of the Democratic Party in the2006 congressional elections and BarackObama and his party in the 2008 election.What’s the story here? One possibility is thatAmerica luckily averted the Gini coefficienttipping point that would send it slidingtoward oligarchy. The likelier possibility isthat key assumptions of the Inequality Roadto Serfdom argument are false. Perhaps the main implicit assumption is

that wealthy Americans will throw their heft

behind politicians and policies that will opposeprogressive redistribution. The election ofBarack Obama makes it increasingly clear thatthis assumption is false, especially when werecall Obama’s admirable clarity in his inten-tion to raise both income and payroll taxes forthe wealthiest Americans, and the intensity ofthe McCain campaign’s tireless advertisementof Obama’s desire to “spread the wealth” andattempt to brand him as a “socialist.” Never-theless, according to exit polls, 52 percent ofAmericans with incomes of $200,000 or highervoted for Obama. In 2004, by contrast, only 35percent of high-income voters supported JohnKerry.49 And what about campaign contribu-tions? The story here is a similar one. Tradition-ally Republican candidates have enjoyed afundraising advantage over their Democraticrivals. Yet in the recent presidential contest,Barack Obama raised an eye-popping $660 mil-lion while John McCain managed only $375million.50

The trend is evidently moving in the wrongdirection for the Inequality Road to Serfdomargument. The evidence that the rich, as aclass, are about to gang up and rig the politicalsystem in their favor is thin. In particular, thefact that the wealthy as a class should be drift-ing steadily toward the more progressivelyredistributive party as income inequality hitshistorical peaks should be almost enough tolay the Inequality Road to Serfdom argumentto rest.51

What’s the mechanism—the chain of causeand effect—that is supposed to take us downthe Inequality Road to Serfdom? The way ris-ing income inequality is supposed to endangerdemocracy is through the conversion of un-equal economic resources into unequal politi-cal resources—the means to affect the out-come of the democratic process. Yale politicaltheorist Robert Dahl lays it out as clearly asanyone:

Because market capitalism inevitablycreates inequalities, it limits the demo-cratic potential of [the best kind of lib-eral] democracy by generating inequal-ities in the distribution of political

17

The evidence thatthe rich, as aclass, are about togang up and rigthe political system in theirfavor is thin.


resources. Because of inequalities inpolitical resources, some citizens gainsignificantly more influence than oth-ers over the government’s policies,decisions, and actions.52

Yet, according to Dahl, political resourcesare a varied lot that include status, honor,respect, affection, charisma, prestige, infor-mation, knowledge, education, communica-tion skills, access to the media, organization,legal standing, persuasive influence over doc-trine or belief, votes, and more. When wefocus on the complex and varied nature ofpolitical resources, it becomes easier to seehow the Inequality Road to Serfdom argu-ment founders.First, it becomes evident that it is not very

easy to convert economic resources intopolitical resources. Many items on Dahl’s listcannot be purchased, as the relative electoralachievements of Steve Forbes, a “connected”heir, and Barack Obama, the son of a schoolteacher and an African immigrant, amplyillustrate. Second, if we think carefully for amoment about the actual distribution ofpolitical resources, it becomes immediatelyclear that the distribution is tightly correlat-ed with economic resources up to a certainlevel of income. But it is very difficult to seehow income in excess of the threshold neces-sary to receive a high-quality education addsmuch to most people’s pool of politicalresources. Given this, rising income inequali-ty should have very little effect on the abilityof the wealthy to influence the outcome ofthe democratic process, beyond financingothers’ attempts at political persuasion. Indeed, financing the operations of politi-

cal action committees, campaigns, thinktanks, advocacy organizations, and money-losing ideological publications is likely thebest that most wealthy Americans can hope todo in converting their money into politicalinfluence. And beyond relatively small-scalegiving to campaigns and causes, most wealthypeople do not spend their money this way.Even when they do, ideologically motivatedwealthy Americans are limited by the menu of

preexisting organizations, prevailing ideas,and the supply of ideologically congeniallabor. No amount of money can buy you athink tank with your politics if there is no onewith your politics to work in it. When Paul Krugman or Wall Street Journal

columnist Thomas Frank emphasizes the roleof free-market research and advocacy organi-zations in their attempts to explain the demo-cratic failure of their favored policies, theyunwittingly undermine the assumptions ofthe Inequality Road to Serfdom argumenteven as they attempt to make it. By emphasiz-ing the importance of the climate of opinionon policy, and the role of writers, commenta-tors, and policy analysts in affecting the cli-mate of opinion, they make it clear that thepolitical resources they think matter the mostare communicative and persuasive resources.These resources are held by the researchers,writers, and media personalities themselves. Inthe best case, an infusion of money canexpand the supply of persuasive talent overtime by supporting the market for it. In mostcases, it amplifies preexisting voices, few ofwhom are especially rich. Now, if it were possible to plot the distrib-

ution of persuasive and communicativeresources on the model of the distribution ofincome, we would not find that they areheavily concentrated in the hands (or brains)of people with anti-distributive politics. Themechanism that is supposed to lead us downthe Inequality Road to Serfdom does notappear to be especially effective. Left-leaning commentators on inequality

have made a great deal of the influence of a fewfree-market think tanks and advocacy organiza-tions, and point to this as evidence of the waymoney can buy persuasion. But there is almostno evidence that right-leaning policy and advo-cacy groups are better-funded overall than left-leaning groups. It’s no secret, after all, that thegreat philanthropic foundations such as Fordand Pew have for decades channeled enormousresources into left-leaning institutions. But thebig story about big money and political activismin recent years is a story of the left. For example,a large, somewhat secretive group of extremely

18

It is not very easy to convert

economicresources into

politicalresources.


rich liberals, calling themselves The DemocracyAlliance, has come together with the specificgoal of limiting, and even overtaking, the influ-ence of right-leaning think tanks and advocacyorganizations by creating an enormous pool offunds to be strategically disbursed to similar left-leaning groups.53 The point here isn’t to argueover whether the “right” or “left” enjoys morebillionaire largesse, but just to observe that atthe high-water mark of the trend in risinginequality—the trend which Krugman and somany others fear will spell the demise of genuinedemocracy—the rich have not come together toconsolidate the influence of the right-leaninginstitutions allegedly designed to guard theirincreasingly vast riches against the hoi polloi.On the contrary, a league of liberal billionairesand an extremely well-financed “movement pro-gressivism” have emerged to check and evenoverwhelm the “movement conservatism” thatallegedly set us on the Inequality Road toSerfdom.54

In any case, gifts from the wealthy are morelikely to be directed toward universities andcolleges than to think tanks. And policy ideasare disproportionately drawn from the work ofacademics supported by those institutions,who also often serve directly as advisers topoliticians and policymakers. Academics as aclass command enormous influence overwhich policies are put on the table for broadpublic consideration. Consider the roles thatacademics like University of Chicago econo-mist Austan Goolsbee and Harvard politicalscientist Samantha Power played in BarackObama’s successful campaign. Academic econ-omists such as Robert Reich, Larry Summers,and Ben Bernanke are routinely appointed tohigh positions in government with enormousdirect influence over economic policy. Thework of legal academics routinely affects thecourts’ interpretation of the laws, and they areroutinely recruited into services as judges onstate and federal courts. Many of these scholarswield an influence over policy, both directlyand indirectly, through their influence on pub-lic deliberation, that few billionaires could everhope to match. Of course, academics over-whelmingly favor the Democratic Party and a

more progressive redistributive policy. A recentstudy by Daniel Klein and Charlotta Sternfound that Democrats outnumbered Republi-cans among political and legal philosophers bya ratio of nine-to-one; among political scien-tists by a ratio of over five-to-one; and amongeconomists by a ratio of about three-to-one.Eighty percent of academic Democrats favoredmore highly redistributive policies.55

The mass media has an enormous influ-ence on how the public perceives political can-didates and public policies, and the question of“media bias” is a perennial source of controver-sy. What is beyond dispute is that, according todata from Gallup, “journalists are still morethan twice as likely to lean leftward than thepopulation overall.”56 Sure, Rush Limbaughand Bill O’Reilly have a lot of influence, as doesRupert Murdoch. But then, so do NationalPublic Radio, Keith Olbermann, and the Ochs-Sulzberger family. A cursory survey of the factson the ground just makes it exceedingly hardto credit the idea that those with the greatestcapacity to affect public opinion and publicpolicy are disproportionately arrayed against amore redistributive, social-democratic UnitedStates, or that rising inequality has created theconditions for its own consolidation. Paul Krugman, both an academic and a

media superstar, is himself an outstandingexample of intensely concentrated politicalresources. (As a matter of fact, he is a rich man,but that’s more an effect than a cause of hispersuasive power.) He has been a staff memberof the White House Council of EconomicAdvisers, in addition to being an enormouslyinfluential trade economist, bestselling author,and columnist for one the world’s most influ-ential and prestigious newspapers. The lowapproval ratings of the Bush administrationand the surging popularity of progressive poli-tics are a testament to the powerful influenceof liberal thinkers like Krugman and a stingingrebuke to the fantastic economic determinismof the Inequality Road to Serfdom argument. Of course, even if powerful opinion lead-

ers and policymakers do favor more redis-tributive politics, the poor themselves maystill be sorely lacking in political resources.

19

A cursory surveyof the facts makesit exceedinglyhard to credit the idea that rising inequalityhas created theconditions for its own consolidation.


But in light of the significant politicalresources arrayed in favor of redistribution,we cannot just assume that the interests ofthe poor are endangered by inequalities in“voice.” In his book Unequal Democracy: ThePolitical Economy of the New Gilded Age,Princeton University political scientist LarryBartels shows that congressmen are relativelyunresponsive to their lower-income con-stituents.57 This may lead us to worry thatthe interests of the least well-off will gounprotected by the democratic process, thatthe “cash value” of their formally equal polit-ical rights will be too little, leaving them espe-cially vulnerable to abuse and neglect by thedemocratic process. But, as Bartels pointsout, his own analysis points to “bright spotsin an otherwise gloomy picture”:

First, the correlation between class posi-tions and political views is not so sub-stantial that support for egalitarianpolicies is limited to “those mired inpoverty.” Just as many poor peopleespouse antipathy to redistribution andthe welfare state, many affluent peoplesupport egalitarian policies that seeminconsistent with their own narrowmaterial interests. Insofar as the politi-cal activism of affluent egalitarians“does perform as advertised,” policy-makers may be much more generousthan the political clout of the poorwould seem to warrant.58

Indeed, if we care about the welfare of theleast privileged members of our society, afocus on equality of “voice” may actually becounterproductive. The issue is improving thewelfare and opportunity of the poor. That is,the issue is whether policy intended to do this“performs as advertised.” It is not surprisingthat the poorest Americans are generally theleast well-educated and have the least access toinformation about politics and policy. But itwould be surprising if those citizens who areleast likely to know the names of candidates,least likely to know the policies that candi-dates support, and least likely to have the kind

of education that would allow them to evalu-ate the effectiveness of alternative policy pro-posals would be able to use democratic partic-ipation effectively to advance their interestsand promote their values. Everyone shouldhave the means to make informed and effec-tive democratic decisions. It would be idealwere each and every citizen to have the incomeand education typical of well-informed, moti-vated voters. To get closer, we need policiesthat will actually work to promote broaderprosperity and a fuller realization of basichuman capacities. We may want to equalize“voice,” but then we need to know what wouldmake that happen, and the democratic publichas to vote for it. In this regard, the danger of “capture” in

democratic politics is not primarily a matterof systemic conflicts of economic interestbetween those occupying different strata ofthe income distribution. Rather, the problemis that political power in democracies flowsto those able to put together winning elec-toral coalitions, and this ability necessarilyinvolves maintaining the loyalties of specialinterests whose demands may not be in thepublic interest. Suppose, for the sake of argument, that the

most effective solution to self-reproducingpoverty is radical structural reform in our sys-tem of primary and secondary education. Inthat case, a party that depends on interestgroups that are violently opposed to anythingmore than marginal reform in this area mayfind itself unable to maintain a cohesive coali-tion, and may also fail to enact the kinds ofpolicies that would actually best ensure thateveryone has the educational and economicmeans to full and effective democratic partici-pation. In such circumstances, we mightexpect that party to minimize the importanceof this kind of reform and instead emphasizepolicies, such as progressive redistribution,that may be much less effective in the longrun, but are also much less threatening to theintegrity of its electoral coalition. Likewise,peace may be strongly in the public interest,but a party coalition that includes powerfulspecial interests that stand to benefit from war

20

If we care aboutthe welfare of theleast privilegedmembers of oursociety, a focus on equality of“voice” may

actually be coun-terproductive.


is likely to overestimate both external threatsto national security and the benefits of mili-tary intervention. The logic of the Americansystem of democracy all but guarantees thatthe government will not be able to reliably pro-duce ideal public policies. Nevertheless, it is possible to do better. But

we’re unlikely to make real progress in improv-ing the quality of public policy if otherwisesophisticated minds continue to be surprisedby the fact that the party promising securitymay leave us less secure, or that the partypromising to lift up the poor may leave themstranded. Strong partisan identification isdangerous because it can pressure even thebest and brightest into accepting that the poli-cies best for the electoral success of theirfavorite party—a fragile and contingent con-sortium of often conflicting interests—willsomehow turn out best for the country. As we’ve seen, the Inequality Road to

Serfdom barely takes one step before stum-bling. We are not easing on down that road.Our democracy has not been captured by therich and turned to the consolidation of theiradvantages. But that by no means guaranteesthat our democracy is well-suited to acting inthe interests of our society’s least privilegedand least powerful members. It is not enoughfor the privileged and the powerful to wishwith their whole hearts to make ours a soci-ety in which all people have a real chance tomake the most of their liberties and lives.Our democracy has to deliver the policiesthat can actually make this happen. But justas special interests can capture democraticcoalitions, our coalitional minds can be cap-tured by democratic politics. What the poorneed is not party faith, but good faith in theeffort to find policies that really deliver.

Equality, Opportunity, andLiberation from PovertyLet’s take stock. Income statistics do not

provide a reliable measure of material well-being. Nominal consumption numbers are abit better, and the weight of evidence favors the

idea that nominal consumption inequality hasrisen much less over the past several decadesthan has income inequality. But nominal con-sumption numbers can be misleading, too.What we’re after is real consumption—real stan-dards of living. The weight of evidence supportsthe idea that there has been no increase in realconsumption inequality. Further, the possibili-ty that standards of living have actually becomemore equal is supported by several strands ofevidence, including the decline of inequality inlife satisfaction since the 1970s. Fixating onincome inequality may have caused us to missone of the biggest stories of modern times:America may have become materially moreequal. And no one noticed. Income inequality is an abstract mathemat-

ical property of a distribution of incomes, andmeasures of income inequality, such as the Ginicoefficient, convey exceedingly little informa-tion relevant to the moral evaluation of socialand political institutions. The same level ofinequality can be the consequence of either justor unjust or moral or immoral influences on apattern of incomes. If there is injustice orimmorality in our social, political, or economicsystem, we should root it out—independent ofits effects on the dispersion of incomes.However, there is little agreement over the injus-tices in the American system, which helpsexplain why some allegedly inequality-causingmechanisms or trends have not been “correct-ed” by democratically determined policy.Moreover, the fact that income statistics are col-lected by government bureaus does not meanthan national-level patterns of income are espe-cially relevant to the moral evaluation of ourpolicies and institutions. Society is an interna-tional network of cooperation and reciprocity,not a nation-state or an exclusive citizenshipclub. A nation-state isn’t a giant firm. “Nationalincome” is an accounting fiction, and notsomething to be divided, either fairly or unfair-ly, among stakeholders, like corporate profits.“Analytical nationalism” can blind us to theimpact of our policies and institutions onnoncitizens, warp our sense of social justice,and lead us to prefer policies that benefit moreadvantaged people over less advantaged people.

21

What the poorneed is not partyfaith, but goodfaith in the effortto find policiesthat really deliver.


Emphasis on abstractions like national-levelincome inequality can distract us from identify-ing and combating injustice and wrongdoingwithin our institutions and can also contributeto further injustice. High levels of income inequality do not

threaten to gut the protections of democraticgovernment and send us down the InequalityRoad to Serfdom. Economic resources are noteasily converted into political resources. Whenthe wealthy attempt to do so, there is little seri-ous evidence that they are attempting to bene-fit themselves at the expense of the poor andmiddle classes. The ability to affect publicopinion and policy is heavily concentrated inacademia and the media, groups that skewdecidedly left. The dizzying levels of economicinequality recently achieved have coincidedwith a movement of the wealthiest Americanstoward the party explicitly in favor of highertaxes and a more progressive redistributionpolicy. But assuring the value of politicalrights by equalizing democratic “voice” re-quires policies that actually work to improvethe education and economic prospects ofpoorer Americans. Our worry should not bethat the wealthy and well-educated will twistdemocracy to their narrow advantage, but thatthe incentives of electoral coalition buildingwill twist the wealthy and well-educated—whodo disproportionately affect the workings ofour democratic institutions—into supportingpolicies that do not really help the people whoreally need it. In their book Why Welfare States Persist, polit-

ical scientists Jeff Manza and Clem Brooksfind, not so shockingly, that welfare states arelarger in countries in which they are morepopular.59 Americans just don’t want Westernor Northern European levels of redistribution,which is a constant source of frustration tothose who would like the United States tolook more like Western or Northern Europe.As I’ve noted, it’s tempting to look for signs ofconspiracy or false consciousness, but it’s easi-er simply to acknowledge that Americanmoral and political culture reflects our pecu-liar political institutions and history. AsHarvard economists Edward Glaeser and

Alberto Alesina conclude in their comparativestudy of American and European strategies forfighting poverty:

Ultimately, we believe the welfare statein the United States did not develop asmuch as in Europe because of Americanpolitical institutions, such as majoritar-ianism (as opposed to proportional rep-resentation), federalism, and checks andbalances, American ethnic heterogene-ity, and different beliefs about the caus-es of poverty in the United States.60

As income inequality has climbed in theUnited States since the 1970s, the populardemand for increased redistribution has barelybudged. This may reflect, in part, a realistic viewof the trends in inequalities in real standards ofliving. But, in the face of constant media reportsabout inequality, it more likely reflects a dis-tinctively American emphasis on poverty andopportunity, as opposed to inequality. It is, of course, a commonplace idea that

Americans endorse “equality of opportunity”but not “equality of outcome.” But this ideaseems to rest on the notion that equality ofopportunity and equality of outcome can beeasily separated. This is a mistake for reasonsPaul Krugman ably identifies:

What it all comes down to is thatalthough the principle of “equality ofopportunity, not equality of results”sounds fine, it’s a largely fictitious dis-tinction. A society with highly unequalresults is, more or less inevitably, a societywith highly unequal opportunity, too. Ifyou truly believe that all Americans areentitled to an equal chance at the startingline, that’s an argument for doing some-thing to reduce inequality.61

There is indeed an internal relationshipbetween opportunity and results. Wealth isjust distilled opportunity, and a child’s oppor-tunities are partly a result of the parent’s levelof economic achievement, that is, of theirresults. However, this suggests that those who

22

Americans just don’t want

Western orNorthern

European levelsof redistribution.


reject equality of results as an ideal have a sim-ilar reason to find equality of opportunityundesirable. If opportunity is so closely tied toresults, then equalizing opportunity will alsorequire constant coercive “corrections” of theemergent pattern of holdings. And that’s themain objection to trying to maintain equalityof results, or any particular pattern of goods,for that matter.Literal equality of opportunity is so undesir-

able that no successful society attempts it. Ofcourse, when most Americans endorse “equalityof opportunity,” they don’t really want the mas-sive intervention that would be literally neces-sary to equalize opportunity. Combing throughpublic opinion data, sociologists Leslie McCalland Lane Kenworthy find that “Americans doobject to inequality and do believe governmentshould act to redress it, but not necessarily viatraditional redistributive programs.”62 Accord-ing to McCall and Kenworthy, as public worriesover income inequality have heightened, moreAmericans have come to support increasedspending on education. This suggests a widely-shared sense that many children are not provid-ed a sufficient opportunity to develop the capac-ities that would serve them well in the presenteconomy. The idea that each person shouldhave access to a baseline level of opportunity is asensible interpretation of the ideal of equality ofopportunity—one that does not imply radicalleveling.Krugman reaches for a familiar metaphor

when he talks about “an equal chance at thestarting line,” but it’s a turn of phrase thatreally confuses people about the positive-sum nature of liberal market societies. Thereis no “starting line” for Americans becauselife in civil society is not a race. It is, insofar asa society is decent, an exercise in cooperation.David Schmidtz lucidly explains that a goodfooting, not an equal footing, is what peopleneed, because what people need is to do wellin life, not to keep up or to “win”:

In a race, equal opportunity matters. In arace, people need to start on an equalfooting. Why? Because a race’s purpose isto measure relative performance. Meas-

uring relative performance, though, isnot a society’s purpose. We form societieswith the Joneses so that we may do well,period, not so that we may do well rela-tive to the Joneses. To do well, period,people need a good footing, not an equalfooting. No one needs to win, so no oneneeds a fair chance to win. No one needsto keep up with the Joneses, so no oneneeds a fair chance to keep up with theJoneses. No one needs to put the Jonesesin their place or to stop them frompulling ahead. The Joneses are neighbors,not competitors.63

If what we’re really concerned about is suf-ficient opportunity, then the link betweenopportunity and income inequality comesundone. Our goal is to make sure that peoplehave meaningful opportunities to make themost of their lives. In a positive-sum society,the fact that some people have fantasticopportunities doesn’t make our goal harder toachieve. How are a poor, inner-city kid’s lifechances affected by the fact that some Webentrepreneur makes billions of dollars asopposed to just millions? If we’re interested inimproving opportunity, we need to focus onthings like intergenerational poverty and fail-ing schools—in other words, things that actu-ally have something to do with the level ofopportunity for people who are struggling. Atbest, income inequality is a distraction.

Conclusion

Income inequality can indeed be reducedin a stroke by taxing the wealthy more heavi-ly. It was, very likely, reduced in a stroke bythe recent financial collapse. But just as thereis no point in wanting a lower Gini coeffi-cient just for its own sake, there’s no point incheering when the income gap narrows, sincethe income gap was never the problem. Theproblem is that too many people in our soci-ety do not have reason to be glad that theylive under these institutions rather than oth-ers. Too many Americans struggle to find

23

Our goal is tomake sure thatpeople havemeaningfulopportunities tomake the most of their lives. Atbest, incomeinequality is adistraction.


decent work, and struggle to raise their fami-lies without a toxic sense of physical and eco-nomic insecurity. Too many Americans areheld captive by the state for acts that shouldnot be crimes. Too many migrant workers areabused because our laws leave them vulnera-ble to abuse. Too many live in fear of losingwhat they have achieved by courageously ven-turing far from home to find opportunity.Too many children are denied the opportuni-ty to develop the intellectual skills and habitsof mind necessary to take advantage of thestupendous variety of opportunities thatwould otherwise be available to them. This isnot okay. And it is not okay for intellectualsand policymakers to waste time and energyworrying that some people, who have had theopportunity to make the most of our institu-tions, have done too well. It doesn’t help. Nordoes it help to encourage people to concen-trate on differences in income, or to resentthem. Demoralization and resentment arenot what people need. As the philosopherHarry Frankfurt put it in his profound essay,“Equality as a Moral Ideal”:

To the extent that people are preoccu-pied with equality for its own sake, theirreadiness to be satisfied with any partic-ular level of income or wealth is guidednot by their own interests and needs butjust by the magnitude of the economicbenefits at the disposal of others. In thisway egalitarianism distracts people frommeasuring the requirements to whichtheir individual natures and their per-sonal circumstances give rise. It encour-ages them instead to insist upon a levelof economic support that is determinedby a calculation in which the particularfeatures of their own lives are irrelevant.How sizable the economic assets of oth-ers are has nothing much to do, after all,with what kind of person someone is. Aconcern for economic equality, con-strued as desirable in itself, tends todivert a person’s attention away fromendeavoring to discover—within hisexperience of himself and his life—what

he himself really cares about and whatwill actually satisfy him, although this isthe most basic and the most decisivetask upon which an intelligent selectionof economic goals depends. Exagger-ating the moral importance of econom-ic equality is harmful, in other words,because it is alienating.64

This exaggeration threatens not only toalienate us from ourselves, but to further alien-ate us from the institutions that have so dra-matically increased prosperity and opportuni-ty. Until we are better able to grasp how it ispossible for well-functioning market institu-tions to narrow gaps in health, longevity, hap-piness, and real standards of living by unleash-ing the entrepreneurial energy and competitivespirit that can also lead to unfathomable for-tunes, our democracy will continue to fail todeliver the conditions most likely to provideeach person sufficient opportunity, a fairchance, to thrive.

Notes1. Paul Krugman, “For Richer,” New York TimesMagazine, October 20, 2002, http://query.nytimes.com/gst/fullpage.html?res=9505EFD9113AF933A15753C1A9649C8B63.

2. Paul Krugman, “The Great Wealth Transfer,”Rolling Stone, November 2006, http://www.rollingstone.com/politics/story/12699486/paul_krugman_on_the_great_wealth_transfer.

3. Krugman, “For Richer.”

4. Thomas Piketty and Emmanuel Saez, “Responseby Thomas Piketty and Emmanuel Saez to: ‘TheTop 1% . . . of What?’ by Alan Reynolds,” http://elsa.berkeley.edu/~saez/answer-WSJreynolds.pdf.This is a reply to Reynolds’s article that appeared inthe Wall Street Journal, December 17, 2006, http://www.opinionjournal.com/extra/?id=110009398.

5. Paul Krugman, The Conscience of a Liberal (NewYork: W. W. Norton, 2007).


7. Krugman, The Conscience of a Liberal, p. 10.

8. See Daniel Slesnick, Consumption and Social

24

It is not okay forintellectuals andpolicymakers towaste time andenergy worryingthat some peoplehave done toowell. It doesn’t

help.


Welfare: Living Standards in the United States and TheirDistribution (New York: Cambridge University Press,2001).

9. Bruce D. Meyer and James X. Sullivan, “FurtherResults on Measuring the Well-Being of the PoorUsing Income and Consumption,” University ofChicago, Harris School Working Paper Series 07.19,August 2007. They conclude that on balance theconsumption data are more reliable, and point outthat the problems with the income data are veryserious when used to evaluate poverty and inequal-ity. “The bottom deciles of consumption exceedthose for income, suggesting underreporting ofincome. There is a high and rising underreportingrate for government transfers, a source of incomethat is particularly important at the bottom.”

10. Dirk Krueger and Fabrizio Perri, “Inequality inWhat?” Cato Unbound, February 16, 2007, http://www.cato-unbound.org/2007/02/16/dirk-krueger-fabrizio-perri/inequality-in-what/. They are re-porting findings from Krueger and Perri, “DoesIncome Inequality Lead to Consumption Inequal-ity?” Review of Economic Studies, March 2006.

11. Slesnick, Consumption and Social Welfare, p. 154.However, Orazio Attanasio, Erich Battistin, andHidehiko Ichimura, who use a different method totackle consumer expenditure survey data, find anincrease in consumption inequality similar to theincrease in income inequality. See “What ReallyHappened to Consumption Inequality in the U.S.?”in E. Berndt and C. Hulten, eds., Measurement Issues inEconomics—The Paths Ahead: Essays in Honor of ZviGriliches (Chicago: University of Chicago Press, 2005).

12. It is possible that, to some extent at least, whatappeared at the time to be consumption smooth-ing was actually excessive and unsustainable bor-rowing. We have to await further data to see if con-sumption inequality trends change in the wake ofthe current recession and tighter credit standards.

13. To see the difference between nominal and realconsumption, imagine twin sisters who purchasean identical bundle of goods and services over thecourse of a year. One lives at the bottom of a moun-tain and the other lives at the top. Because it costsa bit more to transport goods to the top of themountain, everything’s a bit more expensive there.At the end of the year, the sister at the top will havea spent a good deal more than the sister at the bot-tom. So she will be listed as having consumed moreas measured by dollars spent. But, by hypothesis, theyhave consumed an identical bundle of goods andservices.

14. Jan Ott, “Level and Inequality of Happiness inNations: Does Greater Happiness of a GreaterNumber Imply Greater Inequality in Happiness?”Journal of Happiness Studies 6 (2005): 397–420.

15. Betsey Stevenson and Justin Wolfers, “Happi-ness Inequality in the United States,” IZA Discus-sion Paper no. 3624, July 2008, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1198694.

16. Some might be tempted to argue here thathappiness inequality has not risen with incomeinequality simply because, after a certain thresh-old has been met, income has a weak effect on lifesatisfaction. That may or may not be true. Butconsider the tension in insisting that money haslittle to do with happiness and inequalities inmoney are of paramount importance.

17. Anna Yurko at the University of Texas has set outa model in which higher income inequality leadsfirms to compete on quality and price in a way thatreduces variation in product quality over time (i.e.,makes what the richest and poorest buy more alikein quality), and leaves the lowest-quality goods atboth a lower price point and a higher level of qualitythan would be the case in a system with a lower lev-el of income inequality. “Thus,” she argues, “aggre-gate consumer welfare is higher in less egalitarianeconomies.” Like all economic models, Yurko’s con-tains a number of simplifying assumptions. Theimportant thing to note is that fairly straightfor-ward economic theory shows that increasingincome inequality is not only consistent with stablelevels of real inequality, but can even enable reduc-tions in real inequality. See “How Does Income In-equality Affect Market Outcomes in VerticallyDifferentiated Markets?” Working Paper, Universityof Texas Department of Economics, April 2008,http://www.eco.utexas.edu/~yurko/yurko_verti-cal_april08.pdf.

18. Strictly speaking, these two mechanisms areprobably equivalent. The single issue is trends inrates of inflation—trends in the real marginal utili-ty of a dollar, given an income level. But as long asmethods for determining precisely the effect ofquality change on price change remain underdevel-oped and unsatisfactory, it will remain useful toseparate thornier claims about quality changefrom more easily measurable changes in the pricesof the things the rich and poor tend to buy.

19. This section borrows heavily from “The New(Improved) Gilded Age,” The Economist, December19, 2007, of which I am the unattributed author.

20. I borrow this point from John V. C. Nye, “Eco-nomic Growth and True Inequality,” Library of Eco-nomics and Liberty, January 28, 2002, http://www.econlib.org/library/Columns/Nyegrowth.html.

21. Ibid.

22. Mark Bils, “Measuring Growth from Better andBetter Goods,” NBER Working Paper no. 10606,July 2004. “Average quality for durables has likely

25


been understated by about 2.5 to 3 percent per yearduring the past 15 years. This is two to three timesthe magnitude argued for durables in the Boskincommission report.” Among other things, Bils’sfindings imply that the oft-repeated factoid thatmedian real wages have not increased over the pastseveral decades is a mistake based on inadequatemeasurement techniques.

23. Dana Thomas, Deluxe: How Luxury Lost Its Luster(New York: Penguin, 2007).

24. Robert H. Frank, Luxury Fever: Money andHappiness in an Age of Excess (Princeton, NJ: Prince-ton University Press, 2001).

25. John V. C. Nye, “Irreducible Inequality,” Libraryof Economics and Liberty, April 1, 2002, http://www.econlib.org/library/Columns/Nyepositional.html. For another critique of arguments aboutpositional competition by Frank and others, seeWill Wilkinson, “Out of Position: Against the Poli-tics of Relative Standing,” Policy, Spring 2006.

26. Jerry Hausman and Ephraim Leibtag, “CPIBias from Supercenters: Does the BLS Know ThatWal-Mart Exists?” National Bureau of EconomicResearch Working Paper no. 10712, August 2004.

27. John Broda and Christian Romalis, “Inequalityand Prices: Does China Benefit the Poor inAmerica?” Working Paper, University of ChicagoGraduate School of Business, March 2008.

28. Steven Levitt, “Shattering the ConventionalWisdom on Growing Inequality,” New York Times,May 19, 2008, http://freakonom ics.blogs.nytimes.com/2008/05/19/shattering-the-conventional-wisdom-on-growing-inequality/.

29. Heather Wilhelm, “How China Helps America’sPoor,” The American, September 2008, http://www.american.com/archive/2008/september-october-magazine/how-china-helps-america2019s-poor.

30. Enrico Morreti, “Real Wage Inequality,” NBERWorking Paper no. 14370, September 2008.

31. Robert William Fogel, The Escape from Hungerand Premature Death, 1700–2100 (New York: Cam-bridge University Press, 2004), p. 29.

32. United Nations Development Programme,Human Development Report 2006 (New York: Pal-grave Macmillan, 2006), p. 335.

33. Ibid., p. 283.

34. Louis Kaplow, “Why Measure Inequality?” TheJournal of Economic Inequality 3 (2005): 65–79.

35. Branko Milanovic, “Is African Inequality Really

Different?” World Bank Policy Research WorkingPaper no. WPS3169, 2003.

36. Branko Milanovic, Peter H. Lindert, and Jeffrey G.Williamson, “Measuring Ancient Inequality,” WorldBank Policy Research Paper 4412, November 2007.

37. Branko Milanovic, Peter H. Lindert, and JeffreyG. Williamson, “Measuring Ancient Inequality,” Vox,December 7, 2007, http://www.voxeu.org/index.php?q=node/772.

38. Krugman, Conscience of a Liberal, pp. 149–52;Frank Levy and Peter Temin, “Inequality and Insti-tutions in 20th Century America,” MIT Depart-ment of Economics Working Paper no. 07-17, June2007, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=984330.

39. Richard A. Epstein, “In Defense of Contract atWill,” The University of Chicago Law Review 51, no. 4(Autumn, 1984): 947–82. Also see Richard A. Epstein,Free Markets Under Siege: Cartels, Politics and SocialWelfare (Stanford, CA: Hoover Institution, 2005).

40. The debate over the role of CEO pay in risingincome inequality has a similar status. The evi-dence that CEO pay explains some portion of ris-ing inequality is compelling. But the case that itreflects a moral failure is contested and equivocal.Gesturing toward a shift toward a competitivesuperstar market and away from egalitarian com-pensation norms merely invites a debate aboutwhat, if anything, is wrong with that. I am per-sonally inclined to accept that there is somethingdeeply wrong in the structure of corporate gover-nance and that, to some extent, CEO pay is afunction of mediated self-dealing via a web oftightly networked executives. If that is correct,though, what matters is the injustice—and itsdeleterious effects on shareholders and corporatebehavior—not the incidental impact on inequalitystatistics.

41. For an especially stark example of this line ofthinking, see Cornel West, Democracy Matters:Winning the Fight Against Imperialism (New York:Penguin, 2004). The trouble with this view is thatif the policies that would release the people’s“true” or “authentic” preferences are impossiblein the actually existing democratic order, thenthey must be implemented nondemocratically, ornot at all. This suggests that one must eitheraccept the democracy of actual voters or accept anelitist, paternalist alternative. The idea that a briefperiod of benevolently authoritarian rehabilita-tion of “true” preferences may precede the rein-statement of an improved democracy now able toreflect the “authentic” general will is utopian,illiberal, and ridiculous.

42. False consciousness arguments are not neces-

26


sarily bad. It is, again, a matter of identifying themechanism behind it. If a society has only state-controlled media, which is able to control theinformation citizens receive, false consciousnessmay be quite likely.

43. Paul Krugman, “North of the Border,” NewYork Times, March 27, 2006.

44. For the clearest statement of Rawls’ views onthe various harms of inequality, see Justice asFairness: A Restatement (Cambridge, MA: HarvardUniversity Press, 2001), pp. 131–32.

45. Julian Lamont and Christie Favor, “DistributiveJustice,” The Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/justice-distribu tive/.


47. Kevin Phillips, Wealth and Democracy: A PoliticalHistory of the American Rich (New York: Broadway,2003), p. 422.

48. Krugman attributes the “banana republic withnukes” crack to “a friend” in his blog post “Demo-lition Accomplished,” New York Times, http://krugman.blogs.nytimes.com/2008/09/26/demolition-accomplished/.

49. Pew Center for People and the Press, “InsideObama’s Sweeping Victory,” http://pewresearch.org/pubs/1023/exit-poll-analysis-2008.

50. Campaign total receipts were obtained fromwww.fec.gov.

51. Krugman notes a general trend toward pro-redistributive public opinion. He explains this byappealing to the ill-confirmed idea, neverthelesspopular among economists and political scien-tists, that self-interested lower- and middle-classvoters will demand greater redistribution as thepercentage of “national income” held by the richincreases. This sits very uneasily with the idea thatincreasing inequality is self-reinforcing. Krugmantakes the fact that Republicans have won elec-tions despite the trend of rising inequality as evi-dence of conspiracy. But the trend appears to havebeen a portent of massive Republican defeat. SeeConscience of a Liberal, pp. 174–76.

52. Robert Dahl, On Democracy (New Haven: YaleUniversity Press, 2000), p. 167.

53. On the Democracy Alliance, see Jim VandeHeiand Chris Cillizza, “A New Alliance of DemocratsSpreads Funding,” Washington Post, July 17, 2006,http://www.washingtonpost.com/wp-dyn/content/article/2006/07/16/AR2006071600882.html.

54. As I was writing this, John Podesta, the presi-dent of the Center for American Progress, a five-year-old progressive think tank that has receivedlarge sums from the Democracy Alliance, washeading Barack Obama’s presidential transitionteam. That’s influence!

55. Daniel B. Klein and Charlotta Stern, “Profes-sors and their Politics: The Policy Views of SocialScientists,” Critical Review 17 no. 3–4 (2005).

56. Pew Center for Excellence in Journalism, “TheAmerican Journalist: Politics and Party Affiliation,”October 6, 2006, http://journalism.org/node/2304.

57. Larry Bartels, Unequal Democracy: The PoliticalEconomy of the New Gilded Age (Princeton, NJ:Princeton University Press, 2008), Chapter 9.

58. Ibid., p. 289.

59. Jeff Manza and Clem Brooks, Why Welfare StatesPersist (Chicago: University of Chicago Press, 2007).

60. Alberto Alesina and Edward Glaeser, FightingPoverty in the U.S. and Europe: A World of Difference(New York: Oxford University Press, 2004), p. 217.

61. Krugman, The Conscience of a Liberal, p. 249.

62. Leslie McCall and Lane Kenworthy, “Ameri-cans’ Social Policy Preferences in the Era of RisingInequality,” http://www.u.arizona.edu/~lkenwor/americanssocialpolicypreferences.pdf.

63. David Schmidtz, “When Inequality Matters,” CatoUnbound, May 6, 2006, http://www.cato-unbound.org/2006/03/06/david-schmidtz/when-equality-matters/.

64. Harry Frankfurt, “Equality as a Moral Ideal,”Ethics,October 1987.

27


PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2PA Masthead.indd 2 2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM2/9/06 2:08:35 PM

STUDIES IN THE POLICY ANALYSIS SERIES

639. Broadcast Localism and the Lessons of the Fairness Doctrine by John Samples (May 27, 2009)

638. Obamacare to Come: Seven Bad Ideas for Health Care Reformby Michael Tanner (May 21, 2009)

637. Bright Lines and Bailouts: To Bail or Not To Bail, That Is the Questionby Vern McKinley and Gary Gegenheimer (April 21, 2009)

636. Pakistan and the Future of U.S. Policy by Malou Innocent (April 13, 2009)

635. NATO at 60: A Hollow Alliance by Ted Galen Carpenter (March 30, 2009)

634. Financial Crisis and Public Policy by Jagadeesh Gokhale (March 23, 2009)

633. Health-Status Insurance: How Markets Can Provide Health Securityby John H. Cochrane (February 18, 2009)

632. A Better Way to Generate and Use Comparative-Effectiveness Researchby Michael F. Cannon (February 6, 2009)

631. Troubled Neighbor: Mexico’s Drug Violence Poses a Threat to the United States by Ted Galen Carpenter (February 2, 2009)

630. A Matter of Trust: Why Congress Should Turn Federal Lands into Fiduciary Trusts by Randal O’Toole (January 15, 2009)

629. Unbearable Burden? Living and Paying Student Loans as a First-Year Teacher by Neal McCluskey (December 15, 2008)

628. The Case against Government Intervention in Energy Markets: Revisited Once Again by Richard L. Gordon (December 1, 2008)

627. A Federal Renewable Electricity Requirement: What’s Not to Like?by Robert J. Michaels (November 13, 2008)

626. The Durable Internet: Preserving Network Neutrality without Regulation by Timothy B. Lee (November 12, 2008)

3648_PA640_no indicia:3648_PA640_no indicia 7/6/2009 8:59 AM

Untitled-2 2 2/7/06 4:35:00 PM

Thinking Clearly about Economic Inequality

Documents

Transcript of Thinking Clearly about Economic Inequality