Comment by Ira Katznelson
Evil & politics
In the immediate aftermath of the innocents. Behavior is evil when it events of September 11, 2001, it was hard attacks valued goods proffered by West- to know what to say. We seemed bereft ern modernity. of “a terminology,” as Madame de Staël Disputing the integrity and worth of observed after the Jacobin Terror, in a these goods, critics in other intellectual situation “beyond the common meas- circles have focused instead on the evils ure.” In the days that followed, my own of postcolonialism and the exploitative thoughts turned to Hannah Arendt, and relationships characteristic of global the works she had written in an effort to capitalism. It is these iniquities that grapple with another situation beyond should command our attention, not the the common measure. “The problem of acts of terror they consider in a cooler, evil,” Arendt forecast in 1945, “will be sometimes icy, register. the fundamental question of postwar I ½nd the impulses at play in both intellectual life in Europe–as death responses unsettling. Each group is bet- became the fundamental problem after ter at assuming a posture than develop- the last war.” ing ways of acting and living decently in ‘Evil’ is a word one heard with some a world riven by heterogeneous, inter- frequency in the aftermath of the attacks connected, and sometimes conflicting on the World Trade Center and the Pen- cultures. A rote defense of Western lib- tagon, though rarely deployed with eralism could very well authorize a new Arendt’s precision. Within some intel- brand of colonialism, once again making lectual circles, a denunciation of these many non-Western peoples ineligible for acts as evil has been accompanied by a its core values of rights, toleration, par- far too simple justi½cation of liberalism ticipation, and consent. A wholesale and the Enlightenment as decency incar- rejection of enlightened liberalism as a nate. Evil, in this view, implies more mere ½gment of Western imperialism than doing harm or inflicting pain on could very well license an irresponsible and foundationless antimodernism, reinforcing a mirror-image view of ‘us’ Ira Katznelson, Ruggles Professor of Political Sci- against ‘them.’ Intransigently advanced, ence and History at Columbia University, is the each perspective evades asking how we author of numerous books and articles on Ameri- can shade the sensibilities, deepen the can politics, political theory, and social history, capacities, and address the limitations of including “Liberalism’s Crooked Circle: Letters to the liberal tradition in full awareness Adam Michnik” (1996). A Fellow of the Ameri- that credulous notions of human per- can Academy since 2000, he is completing a new fectibility have been mocked by the book on the New Deal, the South, and the origins global diffusion of human superfluous- of postwar American liberalism. ness, the central hallmark of modern
Dædalus Winter 2002 7 Comment “radical evil,” as Hannah Arendt argued Any meaningful effort to re½ne the by Ira in The Origins of Totalitarianism. language and institutions that a robust Katznelson Writing in the aftermath of total war liberalism requires must move beyond a and the Shoah, Arendt sought both to thin and often misleading claim to uni- apprehend the appearance of “radical versality; it also can gain con½dence evil, previously unknown to us,” and to from a fresh appreciation for the En- transform the eschatology of evil into a lightenment’s rich, though often neg- systematic tool with which to name and lected, lineage of realism and a recogni- explain the terrible cost Nazism and tion of liberalism’s history of invention Stalinism had exacted. By ‘radical evil,’ and transformation. Kant, for example, she understood the project of erasing the worried about the demagogic uses of moral and the juridical person as a prel- reason and the possibility that a new set ude to physical annihilation. Justi½ed by of ostensibly enlightened “prejudices millenarian ideologies and advanced by [can] serve, like the old, as the leading what Arendt called manufactured unre- strings of the thoughtless masses.” He alities, radical evil literally erased human also well knew that demonic violence plurality by stripping large populations has long characterized human affairs. of their rights as citizens, including the Such realism is quite distinct from the right to a name, as a prelude to mass rosy optimism of those eighteenth-cen- killing. Turning innocents into nonpeo- tury philosophes who supposed that sys- ple, both the Nazis and the Soviets thus tematic understanding would trump tor- elided the liberal tradition’s central puz- ture and barbarism, as if to realize the zle of how to make it possible for incom- title of Pierre-Paul Prud’hon’s painting mensurable values and identities to of 1798: Darkness Dissipates as Wisdom and coexist, perhaps even flourish, in a cli- Truth Descend to Earth. Rather than mate of toleration. Prud’hon’s canvas, it is Goya’s etchings Although today’s constellation of of Los desastres de la guerra after the Span- Muslim fervency, fascist-style mobiliza- ish insurrection of 1808 and the Peninsu- tion, and Internet-friendly coordination lar War with Napoleon that might better may be new in some respects, it is mani- be adopted as chastening emblems of a festly as capable of producing radical evil humane realism. as the barbarous offshoots of Western Today’s terror forces, or should force, civilization Arendt addressed, even if an engagement not just with this year’s thankfully it has yet to equal them. Fa- instance of evil but with a proper role for miliar, too, are the challenges that Islam- realistic reason and institutional innova- ic zealotry can pose to the tradition of tion in the face of a persistent human Enlightenment and to the possibilities of capacity for desolation, now enhanced a decent liberal politics. by the legacy and diffusion of twentieth- Given these hazards, we need to century models of radical evil. Times of explore whether the Western liberal tra- turmoil and fear urgently pose two ques- dition can effectively contest radical evil tions: whether liberalism can thrive in without sacri½cing its own best features. the face of determined adversaries and I think it can, though not on its own and what kind of liberalism we should wish only if liberals can ½nd a terminology to have. Answers to ‘what kind’ affect and institutional practices to engage the possibilities for ‘whether’ by offering with nonliberal beliefs and cultures choices not only about doctrines but also without dismissing them too hastily as about institutions and public policies. irremediably antiliberal. The ideals of the liberal tradition, 8 Dædalus Winter 2002 properly appreciated, represent an open nance of an idea may affect its status but Evil & sensibility rather than a ½xed set of not its value or capacity. Even if liberal politics arrangements or ideas. The most impor- political thought is inescapably Western tant moments of innovation and change in origin, it no longer belongs only to the in the modern West’s liberal political West. “Concepts such as citizenship, the tradition have come in circumstances state, civil society, public sphere, human governed by anxiety and alarm. Consider rights, equality before the law, the indi- not only Locke’s institutional formula vidual, distinctions between public and for toleration in conditions of religious private, the idea of the subject, democra- warfare between Catholics and Protes- cy, popular sovereignty, social justice, tants, but his speci½cations for political scienti½c rationality, and so on,” he consent and representation in the con- observes, “all bear the burden of Euro- text of a century of civil war in England. pean thought and history.” These secular Consider, too, the constitutional innova- and universal categories and concepts tions of Benjamin Constant in France were preached “at the colonized and at when faced with a global war and the the same time denied . . . in practice. But collapse of legitimate kingship. Consid- the vision,” writes Chakrabarty, “has er, ½nally, the development of the twen- been powerful in its effects. It has histor- tieth-century liberal welfare state in ically provided a strong foundation on response to depression, class conflict, which to erect–both in Europe and out- and the rise of Bolshevism, Fascism, and side–critiques of socially unjust prac- Nazism. tices. . . . This heritage is now global.” Especially at moments of danger and Even when contradicted by such deep innovation, the liberal tradition has been injustices as slavery and Jim Crow, Euro- neither self-contained nor homoge- pean imperialism, and today’s spectacu- neous. There have been liberal demo- lar global inequalities, struggles based on crats, liberal socialists, liberal republi- these orientations ensue “because there cans, liberal monarchists–and also lib- is no easy way of dispensing with these eral Christians, liberal Jews, liberal Mus- universals in the condition of political lims. In each instance, the absence of a modernity.” Or at least, one might say, partnership with political liberalism has no attractive struggles are possible whol- proved an invitation to oppression. ly outside their frame. Without a commitment to such a cardi- Both liberalism and the Enlightenment nal liberal value as toleration, even a within which it nestles advance a philo- declared democrat may be tempted by sophical anthropology of rational actors despotism. The liberal tradition is thus and rational action, insisting that human necessary to an effectively decent poli- agents develop the capacity to deliberate, tics. But it is not suf½cient. An abstract choose, and achieve sensible goals. In commitment to universal human rights their effort to cultivate such rational citi- by itself, without depth, passion, and zens, liberal regimes in the past have all historical particularity, cannot possibly too often imposed various limits, draw- contend with radical evil. An effective ing boundaries that stunt the capacities liberalism modi½es but does not replace of individuals based on their religion, other commitments. race, gender, literacy, criminality, or col- The more global our world, as Dipesh onized status. But after centuries of Chakrabarty reminds us, the more im- struggle about the dimensions of free- perative it is to register that the prove- dom, enlightened political liberalism
Dædalus Winter 2002 9 Comment today acknowledges no legitimate barri- by Ira ers to reason, hence no legitimate ascrip- Katznelson tive barriers to liberal inclusion and lib- eral citizenship. The result is a deep paradox. The glob- al appeal of an enlightened liberalism cannot help but jeopardize the local attachments, the historical particulari- ties–the human plurality–that consti- tute its most important rationale. Here, then, lies liberalism’s most basic current conundrum: how to broaden its endowments in order to protect and nourish heterogeneity while coping with its perils. As our version of this challenge beck- ons, it is not a war on terrorism that will de½ne the early twenty-½rst century, but a series of battles for the soul–that is, for the content, rules, and respectful inclusiveness–of a properly robust, and realistic, liberalism. This endeavor, rather than a stylized conflict about the merits of Enlightenment, had better be the struggle we make our ½rst priority.
10 Dædalus Winter 2002 Richard A. Epstein
Against redress
One of the hardy perennials of politi- ees. He condemns the society that offers cal theory asks how, if at all, one might nothing but small handouts to mothers justify the inequality of wealth and who raise small children but awards for- opportunity that is so manifest in socie- tunes to go-go dancers. ty. The issue has been with us from the It is easy to recite statistics to show earliest times, but it seems to have that an ever-greater percentage of gained renewed urgency in the past wealth is concentrated in–take your decade or so as economic inequality in pick–the top 1, 5, or 10 percent of the the United States, if not in the rest of the income distribution in the United States. world, seems to have become more Public discomfort with the current situa- extreme with the rise of technology. tion is only magni½ed because this eco- In his essay “Beyond Compassion,” nomic divide between rich and poor Orlando Patterson captures something often tracks profound and enduring of the current anger over inequality racial cleavages. when he laments the perverse distribu- In our multiracial society, it is an over- tion of wealth that allows the ceo of a simpli½cation to treat the inequality of large corporation to pull down wages income and wealth as a racial problem. and stock options that exceed the wages But, that said, there is ample evidence to and bene½ts of a thousand line employ- support the proposition that whites as a group are blessed with both greater wealth and higher income than their Richard A. Epstein, James Parker Hall Distin- black counterparts. In many quarters, guished Service Professor of Law at the University the combination of these two dominant of Chicago Law School and Peter and Kirsten features raises twin concerns about Bedford Senior Fellow at the Hoover Institution, political stability and racial and econom- has been a Fellow of the American Academy since ic justice. 1985. He has written extensively in many legal In light of this sorry state of affairs, it areas and played an especially prominent role in is perhaps too easy for Patterson to con- the ½eld of law and economics. His most recent clude that the market is a “moral non- books include “Principles for a Free Society: Rec- starter.” But denunciation is not quite onciling Individual Liberty with the Common the same as argumentation. The current Good” (1998) and “Mortal Peril: Our Inalien- distribution of wealth in America is not able Right to Health Care?” (1997). just the product of the market. It is also
Dædalus Winter 2002 39 Richard A. the product of the crazy-quilt pattern of United States does not, and should not, Epstein on regulation and taxation that seeps into rank high on the list of legislative priori- inequality every area of life. It is therefore neces- ties for the social and economic reforms sary to disentangle the consequences of the next generation. that flow from regulation from those It is nevertheless this hazardous posi- that flow from the market, which in tion that I wish to defend. return requires some theory of what a In order to show why the redress of society based on the market looks like. inequalities should not rank on our list Frederich von Hayek’s claim that local of legislative priorities, I shall ½rst try to knowledge in a decentralized system will outline a sensible theory that helps outperform the handiwork of state min- explain both the uses and the limitations isters (many of whom control literally of the market. The theory here is rightly tens of thousands of times the wealth of described as libertarian in its orienta- ordinary peasants) remains unrefuted by tion, but it makes no pretense that the history, so much so that virtually all sys- market can discharge all social functions tems of regulation, wise or foolish, treat or indeed operate on its own resources the market almost by default as the start- without the assistance of the state. ing place for analysis. Labor statutes do Rather, it argues that state intervention not prohibit negotiations between man- is needed to supply all individuals with agement and labor to set wages and protection against force, fraud, and working conditions. They just institute a monopoly. system of collective bargaining. Antidis- Once this benchmark is established, I crimination laws do not shut down pri- shall then sketch out in general terms vate employment markets. They only two separate lines of argument used to specify certain grounds on which em- justify government intervention to ployers are not allowed to base their hir- redress economic and social inequality, ing decisions. One can attack or defend and then indicate why both of these fall these institutions for the consequences short of their intended goal. These are that follow in their wake, but the simple arguments about restitution and redis- and inescapable truth today is that when tribution, respectively. it comes to the provision of goods and Finally, I shall indicate briefly a gener- services, nothing beats the market. al strategy for social reform that could, Like it or not–and “moral” or not– and should, be adopted to achieve a with the demise of central planning, the more just society, without the costly and market is our starting point. unneeded by-products of government intervention. Rather than add more lay- It hardly follows, however, that the ers of taxation and regulation, the best market is both the starting and the end- tonic for a healthy society is to free up ing point of the analysis. Questions of entry into a host of markets by removing both social and economic inequality the plethora of taxes and regulations remain with us still. Today’s sorry state already in place. The emphasis should be of affairs, which ½nds prosperity tem- on self-suf½ciency, not transfer; it pered with poverty, surely invites some should be on making the economic pie wholesale reform. Obviously, as a politi- greater, not trying ½tfully to use govern- cal matter, it is hazardous to argue that, ment action to put more goods in the even if we keep our market institutions, hands of those who need them the the redress of inequalities within the most.
40 Dædalus Winter 2002 The question of economic and racial to the transaction is rich and the other Against inequalities plays itself out on a vast can- poor, because the voluntary transaction redress vas. But as with so many large problems, will improve the position of both, it is best attacked by breaking it down regardless of their initial endowments of into smaller problems that may prove wealth. amenable to legal solutions that satisfy This is why a market analysis is, our best moral and political instincts. emphatically, a moral starter for social Before attempting to understand the role theory. of state power in dealing with claims The standard libertarian theory there- between groups, it is far easier to ask fore has ample grounds to draw a sharp when the state should back a claim for line between aggression and deceit, compensation or support brought by which it condemns, and competition, one individual against another. which it praises. Competition expands The ½rst place to turn is the theory of the size of the pie, and of each of its corrective justice that has dominated slices; coercion reduces the size of the our view of human interactions since overall pie, and forces some individuals Aristotle. It takes little imagination to to bear a disproportionate share of the award compensation to the person who loss. is either physically attacked or duped by Many Marxist or left-wing theorists another. dispute this result by insisting that ordi- As an initial matter, it is hard to deny nary market transactions are contami- the proposition that the (indiscriminate) nated by exploitation, which is, when all use of force or fraud seriously diminish- is said and done, a form of theft. es the overall welfare of society. The The term “exploitation” requires, of individual who takes from another course, some explication. Clearly, no one always gains, but that gain pales into is particularly upset when it is said that a insigni½cance beside the loss inflicted skillful halfback exploited an opening in on the other person. This reduction in the defense in order to run for a touch- overall wealth and utility (for in this down. Taking advantage of opportuni- context the two go hand-in-hand) more- ties made available within the context of over has adverse consequences on third the rules is often a good, not a bad, parties, who perceive themselves as at thing. And the ½rm that exploits an risk when force and fraud are allowed opening in the market to introduce a unabated. new widget that displaces its creakier At the same time, Aristotelian, rival deserves our thanks, not our con- Kantian, and utilitarian moral theories demnation. are all hard-pressed to condemn any To the determined Marxist or his form of vigorous economic competition modern sympathizers, however, exploi- that involves neither force nor fraud. tation often carries the more cynical The routine business transactions of connotation that one side of the transac- everyday life produce a common good– tion is left worse off than he would have economic gains for all parties to the been if he had never entered it at all. transactions. That increased wealth in But this view of exploitation offers no turn creates still greater opportunities to explanation as to why someone down on produce more goods for trade through his luck would choose to make a con- third parties. Insofar as this is true, we tract that left him poorer than before. do not need to know whether one party Many contracts are performed on a
Dædalus Winter 2002 41 Richard A. repetitive basis: the ordinary worker can This does not mean, however, that Epstein on quit at any time and yet frequently will exploitation is an empty concept. It inequality return to work day after day. He obvious- receives its best de½nition from classical ly does so because he thinks that this economic theory, which condemns (as opportunity is better than any of his some hard-core libertarians do not) alternatives, and it would be an odd monopoly, even if it grows out of volun- form of assistance to ban him from that tary combination and not government line of work altogether. (or private) coercion. In some cases, the charges of exploita- For the purposes of this essay, I shall tion are re½ned so that they concede the accept the standard ef½ciency-based point of mutual gain by contract, but economic theory that in general seeks to insist that the worker is exploited regulate or outlaw monopoly for the because the ½rm has obtained a dispro- resource losses that it imposes on society portionate share of the joint pro½t from as a whole. Hence it may in principle the transaction. Why they presume the make sense to regulate the rates charged asymmetrical division of this unob- by natural monopolies (i.e., traditional served gain remains something of a mys- water, power, and light companies) that tery. Their intuition is that minimum cannot be divided without fatal losses in wage laws, for example, can boost the ef½ciency. And it may in principle be least fortunate worker’s share of the gain possible to prevent the formation of vol- to a larger, and more just, proportion. untary cartels that seek to divide mar- But the imposition of any such rule of kets or to rig prices. division does more than alter shares But if this theory allows the use of gov- enjoyed by current players. It also ernment force to break up or limit changes the entire landscape. The higher monopoly power, by the same token it minimum wage will induce some takes a very grim view of any state barri- employers to reduce their workforces, ers to entry into various economic or others to change nonwage terms of the social markets. The state that imposes a contract. It will narrow the gap between protective tariff may bene½t some local lower- and higher-skilled employees and industry, but that interference with trade thus reduce worker incentives to invest places a far greater burden on those in their own human capital. other individuals who are blocked from Yet, ironically, the one effect that is not choosing their trading partners. likely is a reduction in the employer’s Within this general framework, the share of the surplus, for the higher the state also commits a wrong against its minimum wage, the more likely that own citizens when it imposes restric- some ½rms will exit from the market. tions against their entering into some Hence, the one con½dent prediction trade or business, unless that restriction we can make is this: any effort to tilt by is clearly calibrated, as most occupation- legislation the contractual wage balance al restrictions are not, to prevent the in competitive markets will block volun- practice of fraud on hapless customers. tary transactions, leaving both sides worse off than before. The Marxist con- We are now in a position to outline ception of exploitation is in the end the relationship between this general undermined by the unjust consequences theory of individual rights and the larger that its application in practice will pro- issues of economic and racial inequality. duce. One way to frame the issue is to ask
42 Dædalus Winter 2002 whether poor people generally, or black today to the descendants of the victims Against people speci½cally, have a claim for resti- of state and private violence. redress tution from society at large. Critical problems arise on both sides To state the question in this way of the line. Who should receive restitu- requires us to observe at least two tion? And who should be made to pay important caveats. The ½rst is that we for it? Let us take these two elements in cannot predicate sound theories of resti- order. tution on bad theories of social justice. First, who counts as a victim? That The case here cannot rest therefore on question was easy to answer in 1865 undifferentiated charges of exploitation when huge portions of the African but must be tied to a demonstration that American population in the United these individuals have been the group States had just been released from the victims of force and fraud, including the bonds of slavery. But it is far harder to imposition of barriers to entry, by other afford victim status, over 135 years later, members of society. to their descendants. No one alive today The second caveat is that in principle suffered the cruelties of past regimes. the question of restitution is not The point here is especially true when restricted solely to the position of claims for restitution are pressed on a African Americans, but could in fact be limited basis. For over twenty years, for asked in connection with American example, black parents in Kansas City, Indian tribes, with Chinese and Japanese Missouri, have pressed claims for resti- immigrants, or indeed with any group tution by claiming that black children in that claims to have suffered injustice at the city had been victims of the vestiges the hands of others. But for these pur- of segregation that survived after the poses at least, I shall concentrate on the 1954 Supreme Court decision in Brown v. black experience precisely because the Board of Education. But no remedy here answer to the question of whether resti- can ½t the ostensible wrong, for today’s tution should be provided may seem to schoolchildren in the district bear no be self-evidently in the af½rmative. relationship to the black children in that After all, there is little question that district whom segregation might have the institution of slavery as practiced in shortchanged two generations ago. the United States before the Civil War Quite simply, the program forces vast and the racial restrictions that lay at the amounts of state tax revenue into lavish heart of Jim Crow and the black codes expenditures in one school district, were wholly indefensible when meas- while the educational needs of other ured against a basic theory of libertarian children, black and white alike, suffer rights. Excluding blacks from participa- from comparative neglect. tion in the political and social life of that The dif½culties are, if anything, time constitutes one of the great stains greater in considering who should pay on our history, made still worse by the these claims for restitution. Here the countless acts of private violence and nub of the dif½culty is that the state is intimidation to which the state turned a not just some disembodied entity with a blind eye. heart and mind of its own. Even more Yet it is one thing to recognize the than the private corporation, it is a com- commission of these past serious posite of huge numbers of individuals wrongs, and quite another to conclude who bring to this sprawling nation their that they support claims for restitution own distinctive pasts. Any program of
Dædalus Winter 2002 43 Richard A. restitution, however, contemplates the its moral compass and provide extensive Epstein on use of tax dollars to bene½t some sub- programs that were designed to remedy inequality class of the population at the expense of some of the past conditions of slavery. everyone else. In all cases, this approach We have had extensive af½rmative- necessarily results in risks of overinclu- action programs; we have had programs sion. that targeted the educational shortfalls For example, the claim for restitution on inner-city youth, predominantly brought against the German state after black; we have extensive welfare pro- World War II necessarily fell with equal grams that bene½t disproportionate weight on resistance ½ghters and the numbers of African Americans. most ardent Nazis. Even so, the state ran As a matter of social cohesion, I the risk of overinclusion because of the believe that we would do far better keep- immediacy and enormity of the wrong; ing some general programs in place that it adopted a two-pronged approach that help those at the bottom than trying to compensated survivors of the Holocaust ½nd ways to pay restitution to blacks and their descendants, and then, gener- rich and poor alike. Owing to the com- ally, the state of Israel. plexities involved, my great fear is that It is hard to see how one could devise any program of restitution will emerge any similar program of restitution for as a twisted jumble of preferences that the descendants of former slaves in heaps a second set of injustices on the America. Too much time has passed to ½rst. have any con½dence that the brunt of these payments will be borne by individ- Having examined the claims for resti- uals who had any connection, direct or tution made by African Americans indirect, with the wrongs of a previous today, I shall next briefly address the generation. Many Americans today question of the best social response to descend from those individuals who inequalities of wealth. gave their lives during the Civil War to Much of the wind would be taken free the slaves. Millions of people have from the sails of the current restitution migrated to our shores from just about movement if the average income of every point on the globe, often to escape black citizens were equal to that of the physical danger and economic whites. But while claims of economic oppression of their own lands. By what inequality only lurk behind restitution right do we ask these immigrants and claims, they become the centerpiece of their children to compensate blacks any claim for the redistribution of whose ancestors have been injured by income and wealth. others when they have done nothing This claim of course runs smack into wrong themselves? the libertarian prohibitions on the use of We could, of course, bite the bullet force and fraud, for it honors claims for and conclude that some substantial redistribution even when the poor per- transfer payment should be made from son concedes that he has no corrective general resources nonetheless. But even justice claim to the wealth of the rich here, we have to consider the complica- person. The question is whether these tions that remain. inequalities of wealth justify some Claims for restitution today do not action for redress when the wealth is occur in a vacuum. The same country acquired, and accumulated, through that saw Jim Crow was able to redirect industry, thrift, and invention.
44 Dædalus Winter 2002 Perhaps the easiest way to make the on the edge of starvation do on average Against case for some redistributive action is to need that next dollar more than the redress appeal to the diminishing marginal utili- fashionable elites choosing between vin- ty of wealth. The point here is that the tage wines. The entire enterprise of value of the additional dollar drops the charitable activities, through churches, more dollars that a given person has. A hospitals, and schools, would be largely perfect system of wealth transfer unintelligible if in fact the marginal dol- between persons could presumably lar of wealth were, and were perceived improve aggregate social utility by tak- to be, worth as much in the hands of the ing dollars from the persons who need rich as in the hands of the poor. Who them the least, giving them to the per- would choose to fund soup kitchens, sons who need them most. The total childhood vaccinations, and scholar- number of dollars could, in some ideal ships under those circumstances? world, remain constant after the trans- So one conceptual objection to redis- fer. Does greater satisfaction from these tribution fails to deliver a knockout (redistributed) dollars justify the coer- blow. How then does one continue to cive transfer? dislodge demands for state-mandated One conceptual obstacle to this argu- redistributions of wealth? ment is that it is fanciful at best and mis- A more promising line of argument chievous at worst to purport to make seeks to demystify the state by treating it these interpersonal comparisons of merely as the agent for those individuals wealth. Clearly no social ruler (pun who in any given situation bene½t from intended) lets us know with certainty its actions. Hence the question of that wealth is worth more in the hands whether the state can take wealth from of a poor person than in the hands of a A and give it to B can be reposed: can B rich person, so the determined econo- demand some part of A’s wealth, solely mist can shipwreck the case for wealth because B needs it more? transfers from the start by denying the At this point, the hard-core libertarian possibilities of interpersonal compar- will dismiss B’s claim as mere theft–a isons of utility. I can assert that wealth is coercive seizure of private property. The worth more to the poor person than to state, therefore, is no better than the the rich person; you can deny that Robin Hood who takes from A and gives proposition. The rich person might use to B. the next dollar to complete work on an This argument looks too glib to be invention that will improve the lives of wholly convincing. There are marked others. The poor person might squander differences between an organized sys- it on a drinking binge. We have no way tem of state redistribution and the iso- of knowing if wealth is more useful to a lated actions of a brigand. State action poor than a rich person. can proceed through the ordinary chan- Still, this hard-edged argument has nels of taxation and thus does not pres- bite only insofar as it cautions us against ent the same threat to peace and social the easy assumption that the marginal order as the actions of the ordinary dollar is always worth more in the hands thief. In addition, the social levies in of the poor person than in the hands of question are not concentrated against the rich person. But it does not in my one person on a whim, but are part of a view show that in general these compar- comprehensive social plan that asks all isons are ill conceived. Homeless people of the more fortunate among us to con-
Dædalus Winter 2002 45 Richard A. tribute something to the support of might argue that whenever the thief has Epstein on those who are least fortunate. This web more use for the stolen goods than their inequality of institutional constraint surely makes owner does, the theft helps to advance state action less of a threat than that happiness. But that shortsighted calcula- individual action. tion ignores the broader dynamics of Or does it? theft. In reply, one could argue that it would If the state were to legalize individual be odd to sanction individual thefts on theft, the scope of these activities would the grounds that the thief took only sharply increase, as many individuals some predetermined amounts of wealth would forsake productive activities for from those individuals who were in a what once passed as a life of crime. In position to pay for it. The interposition response, property holders would be of political majorities does not necessar- forced to hire more armed guards to pro- ily insulate the state’s decision from all tect their possessions. Worse, they might criticism. As James Madison reminds us, avoid theft by prematurely consuming political factions often act and vote in goods that they would otherwise save, ways that allow them to line their own thereby depleting the social store of pockets. It hardly counts as a tribute to wealth over time. And if consumption is the democratic process if a minority of not possible, a property holder can wealthy persons is consistently outvoted always choose to invest resources in and outmuscled by those who enjoy the bricks and mortar, which are harder to advantage of greater numbers, namely, steal than money. the poor. Theft is therefore a losing proposition The objection of theft may not be a on both sides of the ledger. The proper showstopper, but it can hardly be dis- social response is to make it illegal– missed on the grounds that the processes both for individuals and for the state. of deliberative democracy insulate all of The hard social question is how many its decisions from substantive attack. resources should be devoted to its elimi- The owners of private property are enti- nation. Here the idealist might be tempt- tled to nothing more than the protec- ed to hold that the state simply has a tions that deliberative democracy wishes moral and social duty to eradicate all to confer upon them. Outright con½sca- forms of theft, including taxation. For tion is not cleansed simply because it is our purposes, the critical point is that authorized by a majority, or even super- the destructive cycle wrought by individ- majority, vote. Progressive taxation is ual theft may be mirrored when the state not cut from the same cloth as those uses coercive means to redistribute forms of collective action that raise the wealth. standards of wealth and happiness for Thus the wise citizens of Hong Kong, all, which is what the state tries to do by fearful of expropriation after the Chi- supplying certain standard public nese takeover in 1997, invested large goods–military defense, a judicial sys- sums of free cash into their new local tem, police protection, public infrastruc- airport, where it was relatively insulated ture–to all its citizens. from expropriation. Allowing the state This last observation is forti½ed when to steal from the wealthy alters the full one looks more closely at the unhappi- range of productive and consumptive ness created by individual acts of theft. activities–generally for the worse. Here a proponent of redistribution Here again, a note of caution is need-
46 Dædalus Winter 2002 ed. I have no doubt that the strong sense their younger and poorer brethren. Pub- Against that motivates private charitable trans- lic cynicism can mount, as it has mount- redress fers affords some political margin of ed, in response to the transparent efforts error against certain state-mandated to make it appear as though every give- transfers designed to help those in dire away on the map should be extolled in need. Most people who are taxed would the name of the public good. be prepared to devote for religious or Why believe that the total sum of state moral reasons some fraction of their and federal redistributive activities pro- wealth to the alleviation of poverty and vides any net bene½t to the poor, who misery. Once the state undertakes that pay through the nose for every major role, private citizens can reduce their subsidy only to receive relatively paltry amount of private giving to offset the welfare bene½ts in exchange? state exaction. Hence the public system In the struggle between different of support displaces the ordinary system political factions over transfer pay- of private charity, but meets with rela- ments, as with individual theft, two tively little resistance so long as the sides are engaged in either blocking or reductions in private giving are avail- securing wealth transfers. Their com- able. bined activities result in a net diminu- Yet the margin for error in this sce- tion of wealth across the board, whether nario is not in½nite. Raise the level of peanut farmers or tobacco farmers win transfer payments for public services too their vaunted subsidies. high, and the private adaptive response The parallel to individual theft goes will be less charity–such that it will no one step further. Once wealth redistrib- longer be able to offset the increased ution is fair game, people will alter their burden of public taxation. Matters only patterns of consumption and invest- get worse if the transfer payments in ment. They will leave less to the next question have, as is so often the case, lit- generation out of fear that the estate tax tle to do with the alleviation of poverty will gobble up their bequests. And they and hardship in our midst. will hire the ½nest lawyers and planners At this point, Madison’s warning to navigate their private fortunes safely about factions becomes pertinent. How through the arcane niceties of the tax taxes are spent generally depends on the code. kind of bare-knuckled political struggle My conclusion is simple: any effort to that makes Washington politics so ugly secure redistribution necessarily reduces today. Losers from proposed legislation the total stock of wealth. And it is not can lobby furiously against it. But lobby- likely to result in transferring wealth to ing is always a two-way street that the poor. allows well-organized bene½ciaries to mount a political campaign in response. If, as I believe, restitution and redistrib- Once the government halls are open ution are more often than not misguid- for business, anyone can apply for ed, even dangerous, strategies for social grants. Farmers can obtain their special reform, what alternatives exist? subsidies; small-business men can opt I can think of two underappreciated for theirs; corporate welfare can enrich lines of inquiry. The ½rst of these is the well-heeled stockholders; senior citi- use of charitable contributions to assist zens can cash in on a rich set of retire- the poor, even through faith-based ini- ment and medical bene½ts denied to tiatives without direct government sup-
Dædalus Winter 2002 47 Richard A. port. Smaller amounts of state-sponsored vexed these transitional issues, the intel- Epstein on inequality redistribution could give families lectual program is clear: remove barriers stronger incentives to take care of their to entry in the trades and professions. wayward members. In addition, any Removing these obstacles costs the gov- charitable dollar is likely to do more ernment nothing in direct expenditures. good than a government dollar because Indeed, it reduces administrative bloat voluntary contributors have at least and, through it, tax burdens. In addition, some incentive to monitor how their it increases the total level of production funds are spent. Finally, reducing gov- in society. In the midst of all the clamor ernment transfers is likely to increase for redistribution, we should not forget overall wealth, which in turn reduces the our initial point of departure: that ordi- demands on the welfare system. Evaluat- nary contracts produce gains from trade ed by its systematic returns, increased that are shared by all parties. The lower charitable spending is no panacea, but it the level of transactions costs, the higher has none of the drawbacks of coercive the velocity of exchanges that move government programs. resources from lower- to higher-valued The second line of action comes from uses. Open entry and freedom of con- a different quarter. The statute books are tract expand the opportunity set across littered with laws that impose indefensi- the board, and are prey to none of the ble barriers to entry into product and destructive consequences that mark labor markets. It is easy to ½nd all sorts resort to faction or theft. The nineteenth- of regulations that exclude individuals century program of trade and labor lib- from driving jitneys, braiding hair, or eralization makes as much sense in practicing law and medicine. Why is it today’s Internet age as it did in an era that Sears, Roebuck can sell lawn mow- dominated by iron and steel. ers but not legal services if it is prepared John F. Kennedy had it right when he to stand behind both? said that a rising tide lifts all boats. And The political forces behind the status that tide will only rise when we put aside quo are formidable. It may well be the our preoccupation with redress and case that entrenched interests will block redistribution–and agree instead to any quick and sudden shift in political unleash the productive capacities of all fortunes that would block the operation our citizens. of competitive markets. But however
48 Dædalus Winter 2002 Christopher Jencks
Does inequality matter?
The economic gap between rich and household incomes, does not collect poor has grown dramatically in the Unit- good data from the rich, but the Con- ed States over the past generation and is gressional Budget Of½ce (cbo) has now considerably wider than in any recently combined census data with tax other affluent nation. This increase in records to track income trends near the economic inequality has no recent top of the distribution. Figure 1 shows precedent, at least in America. The dis- that the share of after-tax income going tribution of family income was remark- to the top 1 percent of American house- ably stable from 1947 to 1980. We do not holds almost doubled between 1979 and have good data on family incomes before 1997. The top 1 percent included all 1947, but the wage gap between skilled households with after-tax incomes and unskilled workers narrowed dramat- above $246,000 in 1997. The estimated ically between 1910 and 1947, which purchasing power of the top 1 percent probably means that family incomes also rose by 157 percent between 1979 and became more equal. The last protracted 1997, while the median household’s pur- increase in economic inequality occur- chasing power rose only 10 percent.1 The red between 1870 and 1910. The gap between the rich and the rest of America has widened steadily since 1 Estimates of the absolute change in purchas- ing power should be treated with extreme cau- 1979. The Census Bureau, which is tion. The Consumer Price Index suggests, for America’s principal source of data on example, that the purchasing power of the bot- tom quintile did not change between 1979 and Christopher Jencks, Malcolm Wiener Professor of 1997, yet data on food expenditures suggest that the poorest quintile felt it had more discre- Social Policy at the Kennedy School at Harvard tionary income (see Bruce Hamilton, “Using University, has been a Fellow of the American Engel’s Law to Estimate cpi Bias,” American Academy since 1992. The author of several classic Economic Review 91 [June 2001]: 619–630) and works, including “Inequality: A Reassessment of direct measures of housing conditions and the Effect of Family and Schooling in America” other amenities suggest that the poorest quin- tile’s material standard of living rose (see Susan (1972), Jencks is currently completing a new study Mayer and Christopher Jencks, “Do Of½cial of inequality in America under the auspices of the Poverty Rates Provide Useful Information Russell Sage Foundation. about Trends in Children’s Economic Wel- fare?” Levy Institute, Bard College, June 2001, © 2002 by Christopher Jencks available at
Dædalus Winter 2002 49 Christopher gap between the poorest ½fth of Ameri- he connection between moral obliga- Jencks T on can households and the median house- tions and empirical evidence is most inequality hold also widened between 1979 and obvious in the case of utilitarian morali- 1997, but the trend was far less dramatic. ty, which requires everyone to follow To liberals who feel that economic rules consistent with the greatest good inequality is unjust or socially destruc- of the greatest number. Utilitarian tive, its growth is evidence that America morality tells us, for example, that we has been headed in the wrong direction. should not litter even when there is no To conservatives who feel either that chance of being punished, because the riches are the best way of rewarding cost to others usually exceeds the bene½t those who contribute the most to pros- to ourselves. But a moral obligation to perity or that a generous welfare state follow rules that promote the greatest encourages idleness and folly among the good of the greatest number does not tell poor, the growth of inequality seems us which speci½c rules for distributing either innocuous or desirable. The goods and services produce that result. debate over inequality involves both If humanity lived entirely on manna moral and empirical claims, but because that dropped from heaven, and if each the empirical claims are hard to assess, additional pound of manna yielded a both sides tend to emphasize moral progressively smaller increase in the arguments. But treating inequality as a recipient’s well-being, rulemakers com- moral issue does not make the empirical mitted to the greatest good of the great- questions go away, because the most est number would seek to distribute common moral arguments for and manna equally, at least when recipients against inequality rest on claims about had equal needs. But economic goods its consequences. If these claims cannot and services do not drop from heaven. be supported with evidence, skeptics People have to produce these goods and will ½nd the moral arguments uncon- services in order to sell them to one vincing. If the claims about conse- another. How much people produce quences are actually wrong, the moral depends partly on how generously their arguments are also wrong. efforts are rewarded. Rulemakers there- fore have to make tradeoffs between the needs of consumers, which are relatively
Figure 1 Changes in the percent of Percent of Household Income household income going to Going to the Richest One Percent the richest 1 percent of 15 American households, 13.6 1979–1997. 12 11.4 Source: Congressional Budget 10.5 Office, Historical Effective Tax 9.6 10.3 Rates, 1979–1997, September Percent 10.1 10.0 10.1 9 2001, Table G-1c. 8.1 7.5 6 79 81 83 85 87 89 91 93 95 97 Year
50 Dædalus Winter 2002 equal, and the motives of producers, democracies. One simple way to Does inequality who usually produce more when extra describe income inequality in different matter? effort leads to higher rewards. countries is to compute what is called The most widely discussed alternative the “90/10 ratio.” To calculate this ratio to the utilitarian theory of justice is the we rank households from richest to theory proposed by John Rawls.2 Rawls poorest. Then we divide the income of claimed that when uncertainty is great the household at the ninetieth percentile and downside risks are high, people by the income of the household at the are–or should be–absolutely risk tenth percentile. (Comparing the nineti- averse. This assumption led Rawls to eth percentile to the tenth percentile is believe that if people did not know what better than, say, comparing the ninety- position they would occupy in a society ninth percentile to the ½rst percentile, they would want to organize the society because few countries collect reliable so as to maximize the well-being of the data on the incomes of either the very society’s least advantaged members. If rich or the very poor.) this claim is correct, utilitarian logic also The Luxembourg Income Study (lis), implies that society should maximize the which is the best current source of data well-being of the least advantaged. Even on economic inequality in different if most people are not as risk averse as countries, has calculated 90/10 ratios for Rawls claimed, they may be suf½ciently fourteen rich democracies in the mid- risk averse to feel that maximizing the 1990s. Table 1 shows the results.3 To position of the least advantaged should keep differences between these fourteen be given very high priority in a just soci- countries in perspective I have also ety. included data on two poorer and less But most thoughtful liberals, including democratic countries, Mexico and Rus- Rawls, also recognize that rewarding sia. If we set aside Mexico and Russia, people for producing more goods and the big English-speaking democracies services will often improve the absolute are the most unequal, the Scandinavian well-being of the least advantaged. Iden- democracies are the most equal, and tifying the best strategy for improving Western European democracies fall in the position of the least advantaged the middle. (Italy looks more unequal therefore requires complex empirical than the other continental democracies, calculations that turn out to be rather but the Italian data is somewhat sus- similar to the calculations required to pect.) Within the English-speaking achieve the greatest good of the greatest world the United States is the most un- number. The rest of this article assesses various empirical claims about how eco- nomic inequality affects both the mean 3 lis adjusts household incomes for size-relat- level of well-being and the position of ed differences in households’ economic needs using a scale in which, for example, a house- the least advantaged. hold of four needs twice as much as a house- hold of one, and a household of nine needs Some of the potential costs and bene- three times as much as a household of one. ½ts of inequality emerge when we con- This scale probably underestimates the addi- tional income needed to maintain a constant trast the United States with other rich level of material well-being and probably over- estimates the additional income needed to 2 John Rawls, A Theory of Justice (Cambridge, maintain a constant level of subjective well- Mass.: Harvard University Press, 1971). being when household size rises.
Dædalus Winter 2002 51 Christopher Table 1 many other Latin American countries. Jencks Income inequality and economic output in America’s unusually high level of on various countries during the 1990s inequality inequality is not attributable to its unusually diverse labor force. Years of Country Ratio of GDP Life (and year holdhold per capita expec- schooling are more equally distributed of the income at as a tancy at in the United States than in the Euro- ninetieth to the 90th to percent of birth the tenth 10th per- u.s. level (1995 pean countries for which we have com- percentile) centile a in 1998 b est.) c parable data (Sweden, the Netherlands, and Germany). Adult test scores are more unequally distributed in the Unit- Scandinaviad 2.8 75 77.2 ed States than Europe, partly because Sweden (1995) 2.6 68 78.9 American immigrants score so poorly on Finland (1995) 2.7 68 76.6 tests given in English. But disparities in Norway (1995) 2.8 85 77.8 cognitive skills turn out to play a tiny Denmark (1992) 2.9 79 75.4 role in explaining cross-national differ- Western Europe 3.6 73 77.5 ences in the distribution of earnings. If Nether. (1994) 3.2 75 77.5 one compares American workers with Germany (1994) 3.2 71 76.6 the same test scores and the same Belgium (1996) 3.2 74 76.4 amount of schooling, the Americans’ France (1994) 3.5 66 78.4 wages vary more than the wages of all Switz. (1992) 3.6 84 78.5 Swedish, Dutch, or German workers.4 Italy (1995) 4.8 67 77.6 Almost everyone who studies the caus- Brit. Com. 4.3 73 77.7 es of economic inequality agrees that by Canada (1994) 4.0 78 78.2 far the most important reason for the Australia (1994) 4.3 75 78.0 differences between rich democracies is e U.K. (1995) 4.6 67 77.0 that their governments adopt different U.S. (1997) 5.6 100 75.7 economic policies. There is no agree- ment about which policies are crucial, but Middle-income LIS nations there is a fairly standard list of suspects. Russia (1995) 9.4 21 (?) 65.0 A number of rich countries have central- Mexico (1998) 11.6 25 na ized wage bargaining, which almost always compresses the distribution of a From
52 Dædalus Winter 2002 while others are far less generous. The designed to measure what different cur- Does inequality United States is unusually unequal part- rencies actually buy in the countries matter? ly because it makes little effort to limit where they are used. Column 2 of Table 1 wage inequality: the minimum wage is shows gdp per capita for the fourteen low, and American law makes unioniza- rich democracies on which lis provides tion relatively dif½cult. In addition, the distributional data. At ½rst glance the United States transfers less money to data seem to support the conservative those who are not working than most case, because the most unequal country, other rich democracies. the United States, also has the highest The fact that the American govern- gdp per capita. That fact makes a strong ment makes so little effort to reduce impression on most Americans. But if economic inequality may seem surpris- you compare the other thirteen rich ing in a country where social equality is democracies in Table 1 you will ½nd no so important. American politicians pres- systematic relationship between in- ent themselves to the public as being equality and per capita gdp. Britain and just like everyone else, and once they Italy, for example, rank just below the step outside their of½ces, Americans all United States in terms of inequality, but wear jeans. The way Americans talk and their gdp per capita is lower than any the music they listen to are also affected other country but France. The fact that by egalitarian impulses. But while the egalitarian economic policies have no tenor of American culture may be demo- obvious correlation with per capita gdp cratic, Americans are also far more hos- within Europe or the Commonwealth tile to government than the citizens of makes a strong impression on egalitari- other rich democracies. Since egalitari- ans in those countries. It also suggests an economic policies require govern- that America’s high output per capita mental action, they win far less support may be traceable to something other in the United States than in most other than our tolerance for economic in- rich democracies. equality. Conservatives have argued for cen- Notice, too, that no rich democracy is turies that trying to limit economic as unequal as Mexico or Russia. Some inequality inevitably reduces both the think this is because the combination of incentive to work and the ef½ciency affluence and democracy always leads with which work is organized. As a countries to adopt somewhat egalitarian result, they think egalitarian societies economic policies. Others think the have fewer goods and services to distrib- causal arrow runs the other way, and ute than societies that allow the market that extreme inequality retards econom- to determine household incomes. One ic growth. This debate is unlikely to be simple way to test the claim is to ask settled soon, because it requires histori- whether countries that tolerate a high cal evidence that is hard to ½nd in poor level of inequality really do enjoy a high- countries. er standard of living. If inequality does not account for Measuring a country’s standard of liv- America’s high gdp per capita, what ing is not easy. The most widely used does? A ½rst step toward answering this measure is probably per capita Gross question is to decompose economic out- Domestic Product (gdp), converted to put into two components: the number American dollars using what is known of hours worked in different countries as “purchasing power parity”–a system (“effort”) and the value of the goods and
Dædalus Winter 2002 53 Christopher Table 2 Jencks Estimates of economic inequality, output, effort, & ef½ciency in seven rich democracies for 1998 on inequality U.S. U.K. Australia Canada France Germany Sweden
Inequality (1994–1997) line 1: 90/10 ratio 5.6 4.6 4.3 4.0 3.5 3.2 2.6 Output (1998) line 2: GDP per capita $32,184 $21,673 $24,192 $25,179 $21,132 $23,010 $21,799 Effort (1998) line 3: % of pop. employed 48.6 45.9 45.8 46.6 38.1 43.5 45.1 line 4: Hrs per worker per yr. 1864 1731 1860 1779 1567 1510 1629 Ef½ciency (1998) line 5: GDP per worker $60,106 $44,280 $47,558 $49,007 $55,714 $50,616 $44,000 line 6: GDP per hr. $32.25 $25.58 $25.57 $27.55 $35.55 $33.52 $27.01 line 7: GDP per “available” hr. $30.81 $23.95 $23.51 $25.26 $31.38 $30.38 $24.77
Source by line: Lines 1 and 2: see Table 1. Line 3: see Statistical Abstract 2000, Table 1376. Line 4: see Organiza- tion for Economic Cooperation and Development, OECD Employment Outlook, Statistical Annex, 2001, 225. Line 5 = line 3/line 4. Line 6 = line 5/line 4. Line 7 = line 6 adjusted to include hours available from those not working but seeking work, assuming that they wanted to work the same number of hours as those actually employed.
services that workers produce per hour able” hours–the number of hours actu- (“ef½ciency”). Table 2 shows such statis- ally worked plus the estimated number tics for the United States and six other of hours that those looking for jobs in a rich democracies. Americans are more given week wanted to work. The last row likely to have paid jobs than people in of Table 2 shows the results of this calcu- the other six countries, but except in the lation. After this adjustment is made, the case of France the difference is fairly United States, France, and Germany small. American workers also seem to look about equally ef½cient. If we set the put in more hours per year than workers United States to one side, moreover, elsewhere, although data on hours there is again no obvious correlation worked is not collected in the same way between inequality and ef½ciency in the in all countries, so the numbers must be other six countries. treated gingerly. Still, the estimates of Another objection to the calculations output per hour suggest that while the in Table 2 is that they take no account of United States is considerably more ef½- cross-national differences in the stock of cient than Canada, Australia, Great physical and human capital. This is true, Britain, and Sweden, it is slightly less but since one major rationale for tolerat- ef½cient than France and Germany. ing a high level of inequality is that this One obvious objection to this compar- supposedly encourages capital accumu- ison is that unemployment is higher in lation and investment, holding Ameri- France and Germany than in the United ca’s advantages in these domains con- States. One way to correct for this waste stant would bias the results in favor of of human resources is to divide econom- equality. The calculations in Table 2 also ic output by what Table 2 labels “avail- ignore national differences in natural
54 Dædalus Winter 2002 resources, but such an adjustment er hours and more consumer goods, Does would almost surely make America look they mostly seem to opt for consumer inequality matter? worse, not better. Perhaps the most fun- goods rather than family time or leisure. damental objection of all is that statis- This is a legitimate choice, but it has tics on gdp take little account of differ- nothing to do with economic ef½ciency. ences in the quality of the services in Until fairly recently the United States different countries, since these differ- was so much richer than other countries ences are almost impossible to measure. that even the poor lived better in Ameri- If America’s service sector produces ca than elsewhere, leading conservatives more satis½ed customers than the serv- to argue that laissez-faire policies bene- ice sector in France or Germany, Table 2 ½ted everyone in the long run. Today, may understate the bene½ts of inequali- however, the American poor are no ty. longer the world’s most affluent. Tim If American managers had organized Smeeding, who directs the lis, and Lee the economy in an unusually ef½cient Rainwater, a Harvard sociologist, have way, so that American workers were compared the purchasing power of producing signi½cantly more (or better) households at the tenth percentile of the goods and services per hour than their income distribution in thirteen rich counterparts in other rich democracies, democracies covered by the lis. These it would be fairly easy to argue that they comparisons provide a pretty good indi- deserved their fabulous salaries. Table 2 cation of how the poor fare in different is obviously not the last word on this countries. Table 3, which is based on issue, but it does not suggest that Ameri- their work, shows that the American can workers are producing signi½cantly poor are better off than the poor in more per hour than their counterparts in Britain or Australia but marginally other rich countries. Comparisons that worse off than the poor in Sweden, adjust for the stock of physical and Canada, and Finland, and substantially human capital show the same thing.5 worse off than the poor in Western America’s high standard of living seems Europe. to depend as much on long hours as Conservatives often blame American clever management or clever workers. poverty on the existence of an “under- The fact that Americans spend so class” that rejects mainstream social much time working is rather surprising norms, does little paid work, and has for an affluent nation with a reputation children whom neither parent can sup- for hedonism. Workers in Germany, port. It is certainly true that poor Ameri- France, Japan, and Britain have cut their can households include fewer working hours substantially since 1980. Ameri- adults than affluent American house- cans cut their hours earlier in the twen- holds. This is true in every rich country tieth century but have not done so since for which we have data. But when Lars 1980. Americans tell pollsters that they Osberg, an economist at Dalhousie Uni- would like to work fewer hours, but versity, compared poor households in when they have a choice between short- the United States, Canada, Britain, Swe- den, France, and Germany, he found 5 Robert Hall and Charles Jones, “Why Do that the poor American households Some Countries Produce So Much More Out- worked far more hours per year than put per Worker than Others,” Quarterly Journal their counterparts in the other ½ve of Economics 114 (1999): 83–116.
Dædalus Winter 2002 55 Christopher Table 3 countries.6 This ½nding suggests that Jencks Purchasing power of households at the 10th what distinguishes the United States on and 90th percentiles of each nation’s distribu- inequality tion relative to households at the same per- from the other rich democracies is not centile in the United States in the same year, the idleness of the American poor but 1992–1997 the anger that idleness inspires in more affluent Americans, which helps explain Country Purchasing power as a percent of the stinginess of the American welfare (and year) the U.S. level in the same year state. 1oth 90th Average percentile percentile of all If Rawls is right, disinterested rulemak- percentiles ers in all societies should be trying to maximize the well-being of the least Scandinavia 112 57 77 advantaged. If you accept that claim, Sweden (1995) 103 49 67 Table 3 suggests that Western European Finland (1995) 105 53 73 countries are doing a better job than the Norway (1995) 128 68 88 United States and that Western Euro- Denmark (1995) 110 59 80 pean countries are more just. But if you Western Europe 119 73 88 are a utilitarian whose goal is to maxi- Neth. (1994) 110 64 76 mize the average level of well-being, the Germany (1994) 113 67 82 situation is not so clear. If you want to Belgium (1996) 121 73 80 compare the average level of well-being France (1994) 110 71 84 in countries with different distributions Switz. (1992) 141 89 116 of income, you need some way of com- Commonwealth 94 73 80 paring the value people at different Canada (1994) 105 80 92 points in the income distribution assign U.K. (1995) 85 68 72 to additional after-tax income. Table 3 Australia (1994) 87 71 76 suggests, for example, that poor Canadi- U.S. (1997) 100 100 100 ans have 5 percent more purchasing power than their American counter- parts, while affluent Americans have 25 Source: Columns 1 and 2 are from Timothy Smeeding percent more purchasing power than and Lee Rainwater, “Comparing Living Standards affluent Canadians. If your goal is to Across Countries: Real Incomes at the Top, the Bot- achieve “the greatest good of the great- tom, and the Middle” (paper prepared for a confer- ence on “What Has Happened to the Quality of Life est number,” you need some way of in America and Other Advanced Industrial Nations?” deciding whether the 25 percent advan- Levy Institute, Bard College, Annandale-on-Hudson, tage of affluent Americans over affluent N.Y., June 2001). Local currencies were converted to Canadians should count for more or less dollars using their estimated purchasing power pari- ty. Area averages are unweighted arithmetic means. than the 5 percent advantage of poor Column 3 is calculated from the national means of Canadians over poor Americans. the logarithms of after-tax household income, using When employers want to reward all data provided by Rainwater. members of a hierarchical work group equally, they usually raise every mem- ber’s wage by the same percentage.
6 Lars Osberg, “Labour Supply and Inequality Trends in the U.S.A. and Elsewhere,” available at
56 Dædalus Winter 2002 When social scientists measure econom- land in 1992 than Americans near the Does ic inequality, they too assume that top would have lost. Column 3 of Table 3 inequality matter? inequality has not changed if everyone’s generalizes this logic by comparing income has risen by the same percent- households at every point in each coun- age. Such practices suggest that many try’s income distribution to those at the people think a 1 percent increase in same point in other countries and aver- income is equally valuable to the rich aging the percentage differences.7 Aver- and the poor, even though a 1 percent aging across the entire income distribu- increase represents a much larger tion, Switzerland again does substantial- absolute increase for the rich. In what ly better than the United States in 1992, follows I will refer to the assumption but all the other rich democracies in that a 1 percent gain is equally valuable Table 3 do somewhat worse than the at all income levels as the “One Percent United States. Is Always The Same” rule, or the opiats rule for short. Up to this point I have been focusing The opiats rule implies that if my exclusively on what people can afford to income is $100,000 and I give $20,000 buy. While economic goods and services of it to the poor, my well-being falls by a are obviously important, many people ½fth. If I divide my $20,000 equally believe that inequality also affects between ten people with incomes of human welfare in ways that are inde- $10,000, ten people’s well-being will rise pendent of any given household’s pur- by a ½fth. The gains from this gift will chasing power. Even if my family in- thus exceed the losses by a factor of ten. come remains constant, the distribution The utilitarian case for governmental of income in my neighborhood or my redistribution almost always reflects this nation may influence my children’s edu- logic: taxing the rich won’t do them cational opportunities, my life expectan- much harm, and helping the poor will cy, my chance of being robbed, the prob- do them a lot of good. If you look at the ability that I will vote, and perhaps even actual relationship between income and my overall happiness. The remainder of outcomes like health and happiness, the this article tries to summarize what we opiats rule seldom describes the rela- know about such effects. tionship perfectly, but it comes far closer Educational opportunities: Increases in than a “One Dollar Is Always The Same” economic inequality have raised the rule, which is the only rule under which value of a college degree in the United income inequality does not affect health States. If all else had remained equal, or happiness. making a college degree more valuable If we apply the opiats rule to the should increase both teenagers’ interest tenth and ninetieth percentiles in Table in attending college and their parents’ 3, the percentage gains accruing to those willingness to pay for college. But the at the ninetieth percentile from living in growth of economic inequality in Amer- the United States almost always exceed ica has been accompanied by a change in the percentage gains accruing to those at the tenth percentile from living in West- ern Europe or Canada. Switzerland is a 7 Column 3 is calculated from the differences between national means for the logarithm of notable exception. Americans near the after-tax household income adjusted for house- bottom of the distribution would have hold size. Comparing medians in different gained far more from living in Switzer- countries yields almost the same results.
Dædalus Winter 2002 57 Christopher the way we ½nance public higher educa- Table 4 Jencks tion. Tax subsidies play a smaller role Percent of high-school graduates enrolling in a on 4-year college or some other form of postsec- inequality than they once did, and tuition plays a larger role. Since 1979 tuition at Ameri- ondary education within 20 months of gradua- tion, by income quartile: 1980–1982 and 1992 ca’s public colleges and universities has risen faster than most parents’ income. If American high-school graduates Income Entered a Entered some other quartile 4-year form of post- were as well informed and farsighted as college secondary education economic theory assumes, they would 1980 1992 Change 1980 1992 Change have realized that the monetary value of –82 –82 a college degree was rising even faster than tuition. College attendance would have risen both among children whose Lowest 29 28 –1 28 32 4 parents offered to pay the bills and Second 33 38 5 30 32 2 among children who cover their own Third 39 48 9 33 32 –1 costs, who would either have borrowed Highest 55 66 11 26 24 –2 more or worked longer hours to earn a All 39 45 6 29 30 1 degree. But while some students clearly Source: David Ellwood and Thomas Kane, “Who Is respond to changes in the long-term Getting a College Education? Family Background and payoff of a college degree, many do not. the Growing Gaps in Enrollment,” in Sheldon Indeed, the reason affluent parents offer Danziger and Jane Waldfogel, eds., Securing the Future to pay their children’s college expenses (New York: Russell Sage, 2000). rather than just giving their children cash is that parents fear that if the chil- and Thomas Kane at ucla. It shows dren got the cash they might spend it on changes between 1980–1982 and 1992 in something with more short-term payoff, the fraction of high-school graduates like a flashy car or a trip around the from different economic backgrounds world. If affluent parents are right in entering four-year colleges. Among stu- thinking that their seventeen-year-olds dents from the most affluent families, have short time horizons, the same is the proportion entering a four-year col- probably true for less affluent high- lege rose substantially. Among students school graduates whose parents cannot from middle-income families, whose pay their college expenses. Such students families often help with children’s col- are likely to be far more sensitive to lege expenses but seldom pay the whole changes in tuition than to a change in bill, attendance rose more modestly. Stu- the hypothetical lifetime value of a BA. dents from the poorest quartile were no Tuition is easily observed and has to be more likely to attend a four-year college paid now. The lifetime value of a BA is in 1992 than in 1980–1982. This pattern, always uncertain and cannot be realized in which enrollment rises more at the for a long time. Among students who top than at the bottom, is just what we pay their own bills, higher tuition could would expect if parents respond to easily reduce college attendance even changes in the long-term bene½ts of col- when the long-run returns of a college lege while students respond to changes degree are rising. in short-term costs. It is important to Table 4 is taken from work by two emphasize, however, that the poorest economists, David Ellwood at Harvard quartile’s chances of attending college
58 Dædalus Winter 2002 did not fall appreciably; they just failed to be important, because a school’s abili- Does to rise. The poorest quartile was worse ty to attract effective teachers turns out inequality matter? off only insofar as higher education con- to depend largely on its socioeconomic stitutes a “positional” good, whose value mix. Even when districts with a lot of depends not just on how much you have poor children pay better than nearby but how much others have. That re- districts, as they sometimes do, they sel- mains a contested issue. dom attract teachers who are good at If rising economic inequality ex- raising children’s test scores. Increasing plained the trends in Table 4, the corre- economic segregation is therefore likely lation between parental income and col- to reduce the chances that low-income lege attendance should have grown students will get good teachers. fastest in those states where economic Life expectancy: People live longer in inequality grew fastest. Susan Mayer, a rich countries than in poor countries, sociologist at the University of Chicago, but the relationship flattens out as na- has shown that that is exactly what hap- tional income rises. Indeed, the statistics pened during the 1970s and 1980s.8 in Table 1 show that life expectancy and Overall, growing economic inequality in gdp per capita are not strongly related a state raised college attendance, partly in rich democracies. In particular, life because it was accompanied by in- expectancy is lower in the United States creased spending on all levels of public than in almost any other rich democra- education. The positive effects of grow- cy. ing inequality on college attendance per- Within any given country people with sisted even when Mayer took account of higher incomes also live longer. This re- changes in the payoff of schooling in the lationship flattens out near the top of student’s home state. But in the states the income distribution, but the gap be- where inequality grew the most, the tween richer and poorer families does effect of parental income on educational not seem to narrow when everyone’s attainment also grew. standard of living rises. Despite both ris- Mayer has also shown that the ing incomes and the introduction of Me- increase in economic inequality between dicare and Medicaid, for example, the 1970 and 1990 led to greater economic effects of both income and education on segregation between neighborhoods.9 mortality increased in the United States When the rich got richer they evidently between 1960 and 1986.10 Class differ- moved to affluent suburbs where other ences in mortality also widened in Eng- rich people were also moving. Income land between 1930 and 1960, even disparities within neighborhoods hardly though the overall standard of living changed. Economic segregation is likely rose and the National Health Service
8 Susan E. Mayer, “How Did the Increase in 10 See Harriet Orcutt Duleep, “Measuring Economic Inequality between 1970 and 1990 Socioeconomic Mortality Differentials over Affect Children’s Educational Attainment?” Time,” Demography 26 (May 1989): 345–351, American Journal of Sociology, forthcoming. and G. Pappas, S. Queen, W. Hadden, and G. Fisher, “The Increasing Disparity in Mortality 9 Susan Mayer, “How the Growth in Income between Socioeconomic Groups in the United Inequality Increased Economic Segregation,” States, 1960 and 1986,” New England Journal of Irving Harris Graduate School of Public Policy Medicine 329 (1993): 103–109. Studies, University of Chicago, 2001, available at
Dædalus Winter 2002 59 Christopher equalized access to medical care.11 Such sequent work showed that mortality was Jencks also lower in American states and metro- on facts suggest that the linkage between inequality income and health involves more than politan areas where incomes were more material deprivation. Otherwise, dou- equal. One explanation for this phenom- bling everyone’s purchasing power enon is the opiats rule. A 1 percent would narrow the gap between the top increase in income lowers the odds of and the bottom. dying before the age of sixty-½ve by One reason for the persistent correla- roughly the same amount, regardless of tion between income and health is that what your initial income is. This means poor health lowers people’s earning that adding $1,000 to the income of a power. In addition, big medical bills can million poor families while subtracting deplete a family’s savings, lowering its $1,000 from the incomes of a million ri- unearned income in later years. But cher families should lower overall mor- while poor health clearly affects income, tality. It follows that countries, states, studies that follow the same individuals or cities with the same mean income over time suggest that income, occupa- should have lower death rates when this tional position, and education also affect income is more equally distributed. But people’s health. One reason is that mem- if this were the only way in which in- bers of affluent households are more come inequality affected life expectancy, likely to follow the medical profession’s the difference between the United States advice. Affluent Americans now smoke and Sweden would be quite small. far less than poor Americans, for exam- Wilkinson and his followers believe ple. Affluent Americans also get a bit that inequality also lowers life expectan- more exercise than the poor and are less cy independent of its effect on any given likely to be overweight. But even when household’s income, because it changes we take these differences into account, the social context in which people live. much of the correlation between in- According to Wilkinson, inequality come and life expectancy remains unex- erodes the social bonds that make peo- plained. Experimental studies that mani- ple care about one another and accentu- pulate a monkey’s rank in the hierarchy ates feelings of relative deprivation (the of its troop suggest that rank affects social-science term for what people used health, and the same is pretty clearly to call envy). Other epidemiologists take true for humans. But we do not know what they call a “materialist” position, how much of the association between arguing that inequality kills because it income and health can be explained in affects public policy, altering the distri- this way. bution of education, health care, envi- In 1992 Richard Wilkinson wrote an ronmental protection, and other materi- influential article arguing that a more al resources. Either way, if we compare equal distribution of income improved people with the same income–say life expectancy in rich countries.12 Sub- $50,000 a year–those who live in places where incomes are more unequal should 11 See Elsie Pamuk, “Social Class Inequality in die younger. Mortality from 1921 to 1972 in England and Recent research has raised serious Wales,” Population Studies 39 (1985): 17–31. doubts about such claims. As data on more countries and more time periods 12 Wilkinson summarized his thinking on this have become available, the cross-nation- issue in Unhealthy Societies: The Afflictions of Inequality (London: Routledge, 1996). al correlation between economic in-
60 Dædalus Winter 2002 equality and life expectancy has fallen creased fastest. That ½nding was consis- Does perilously close to zero. If you look at tent with the Wilkinson hypothesis, inequality matter? Table 1 and simply contrast America although the effect could have been due with other rich democracies, the idea to chance. But when we extended our that inequality kills seems to make analysis back to 1970, the relationship sense. But if you compare the other rich was reversed. That relationship could democracies with one another, you ½nd also have been due to chance. no consistent association between in- When Clarkwest and I looked at equality and life expectancy. Incomes changes in economic inequality within are far more unequal in Canada, Aus- the rich democracies that participate in tralia, and Great Britain than in Scandi- the Luxembourg Income Study (lis), we navia, for example, but life expectancy is found that life expectancy had risen about the same in these two groups of everywhere, but it has risen less rapidly countries. in those countries where economic in- Recent work has also raised doubts equality was rising fastest.14 This was about the causal link between inequality consistent with the Wilkinson hypothe- and life expectancy in American states sis, and in this case the relationship was and cities. In America, both economic too large to blame on chance, at least inequality and life expectancy are corre- using conventional statistical standards. lated with the percentage of African Nonetheless, the relationship was weak. Americans in a state or city. Blacks die Economic inequality in the United younger than whites no matter where States rose by about a sixth between they live, so states with large black pop- 1979 and 1997.15 Life expectancy in the ulations have above-average mortality United States rose by three years during rates no matter how their residents’ this period. Had inequality not in- income is distributed. American whites creased, the lis data implied that life also die younger when they live in a state expectancy in the United States would or a metropolitan area with a large Afri- have risen by an additional 0.3 years. To can American population. Once one keep this number in perspective, it helps takes the effects of race into account, the to remember that Americans in the top 5 correlation between economic inequali- percent of the income distribution can ty and mortality tends to disappear.13 expect to live about nine years longer If we want to know whether egalitari- than those in the bottom 10 percent.16 an policies would improve people’s health, however, we need to ask whether 14 See Andrew Clarkwest, “Notes on Cross- changes in economic inequality at the National Analysis of the Relationship between national, state, or local level are associ- Mortality and Income Inequality,” Malcolm Wiener Center for Social Policy, Harvard Uni- ated with changes in life expectancy. The versity, 2000, available at
Dædalus Winter 2002 61 Christopher The apparent effect of even a fairly large Sweden’s would have a big effect on hap- Jencks on change in a nation’s income distribution piness. Just as with health, equalizing inequality pales by comparison. the distribution of income is only likely We also need to bear in mind that the to have large effects on happiness if it cross-national correlation between changes the social context in which peo- changes in economic inequality and ple live. If equality strengthens social changes in life expectancy may not be ties or reduces envy, for example, that causal. Countries that restrained the could reduce unhappiness signi½cantly. growth of economic inequality after Empirical evidence for a correlation 1980 were dominated by political parties between equality and happiness remains that felt either politically or morally thin. Michael Hagerty, a social psycholo- obligated to protect the interests of their gist at the University of California, Da- less affluent citizens. Such countries may vis, has shown that Americans are less have done all sorts of other things that likely to say they are happy when they made people live longer, like reducing live in cities where incomes are more the work week or ensuring that more unequal, but his analysis does not take people got the health care they needed. account of the correlation between eco- Happiness: The relationship between nomic inequality and racial mix. A team income and happiness is much like the of economists at Harvard and the Lon- relationship between income and health, don School of Economics has shown except that it is easier to tell whether that Europeans become less satis½ed someone has died than whether they are with their lot when their country’s in- unhappy. Almost every year since 1972 come distribution becomes more un- the General Social Survey (gss) has equal, but this effect is con½ned to asked national samples of American respondents who identify with the polit- adults the following question: ical Left.17 All this evidence is sugges- tive, but hardly de½nitive. Taken all together, how would you say Crime: Several studies have found that things are these days? Would you say that violent crime is higher in American met- you are very happy, pretty happy, or not ropolitan areas where the distribution of too happy? income is more unequal. But these stud- Those with higher incomes tend to say ies have not looked at whether increases they are happier than those with lower in inequality are associated with increases incomes. This relationship flattens out in crime. For the United States as a near the top of the distribution, but not whole, trends in economic inequality do enough to suggest that making the not match trends in violent crime at all American distribution of income like closely. Inequality hardly changed dur- ing the 1960s, when violent crime rose Angus Deaton, which shows that men in the highest income group have death rates compa- 17 See Michael Hagerty, “Social Comparisons rable to men twelve years younger in the low- of Income in One’s Community: Evidence est income group, and that women in the high- from National Surveys of Income and Happi- est income group have death rates comparable ness,” Journal of Personality and Social Psychology to women six years younger in the lowest 78 (2000): 764–771, and Alberto Alesina, income group. My use of Deaton’s results to Rafael Di Tella, and Robert MacCulloch, infer overall disparities in life expectancy “Inequality and Happiness: Are Europeans and requires several assumptions that are unlikely Americans Different?” Cambridge: National to be exactly correct. Bureau of Economic Research, April 2001.
62 Dædalus Winter 2002 sharply. Inequality rose in the early tion and politicians need to keep all Does inequality 1980s, when violent crime fell. Inequali- income groups happy. When people stop matter? ty rose more slowly in the late 1980s, voting, turnout almost always falls the when violent crime rose again. Inequali- most among the poorest and least edu- ty near the top of the distribution rose in cated. As the income gap between those the 1990s, while violent crime fell. None who vote and the population as a whole of this proves that changes in the distri- widens, politicians have less incentive to bution of income have no effect on push legislation that bene½ts the lower crime, but it does suggest that trends in half of the income distribution. Richard violent crime depend largely on other Freeman, an economist at Harvard, has influences. shown that class disparities in presiden- Political influence: Americans are less tial turnout increased between 1968 and likely to vote today than in the 1960s. 1972 and that the same thing happened The Left sometimes blames this decline between 1984 and 1988.19 I have not seen in turnout on the fact that almost all the any evidence on what has happened bene½ts of economic growth have been since 1988. going to a small minority. Parties of the American political campaigns have Left in most other countries have made also changed in ways that make it riskier sure that the bene½ts of growth were for politicians to upset the rich. Until the more equally distributed. In America, 1960s most political candidates relied the Democrats have barely discussed the largely on volunteers to staff their cam- problem. As a result, voters are said to paign of½ces and contact voters. Now have become convinced that neither they rely largely on paid staff and televi- party cares about their problems. sion advertising. This change reflects Nonetheless, growing economic the fact that politicians can raise more inequality cannot explain the decline in money today than in the past. Political turnout, because this decline occurred in contributions have probably risen be- the early 1970s, well before inequality cause government affects more aspects began to grow. Turnout has hardly of our lives, so both voters and corpora- changed since 1980.18 If growing in- tions are willing to spend more money equality has affected turnout, it must to influence government regulations and have done so by perpetuating a decline spending patterns. Whatever the expla- that occurred for other reasons. nation, people who can contribute The most obvious causal link between money now have more political weight, turnout and equality runs the other way. and people who can contribute time If everyone votes, the electorate is by have less. Politicians also know that the de½nition representative of the popula- 2000, Table 479). If one allows for the fact that citizens constitute a declining fraction of 18 About 62 percent of the voting-age popula- the voting-age population and the fact that tion cast ballots in the three presidential elec- more citizens are disenfranchised because tions conducted during the 1960s. Turnout fell they are–or have been–in prison, turnout to 55 percent in 1972, 54 percent in 1976, and 53 among eligible voters may actually have percent in 1980. Since 1980 presidential turn- increased slightly since 1980. out has averaged 52 percent, with no clear trend. Off-year congressional elections have 19 Richard Freeman, “What, Me Vote?” followed the same trajectory (U.S. Bureau of paper presented at the Workshop on Inequali- the Census, Statistical Abstract of the United ty and Social Policy, Kennedy School of Gov- States, 2000, Government Printing Of½ce, ernment, June 2001.
Dædalus Winter 2002 63 Christopher easiest way to raise the money they need political participation, but declining Jencks on is to court affluent contributors. When political participation among the less inequality the share of income going to the top 1 affluent may help explain why American percent rises, politicians have more in- politicians remained so passive when centive to raise money from this group. inequality began to grow after 1980. If politicians had to rely exclusively on My bottom line is that the social con- contributions of less than $100, they sequences of economic inequality are would also have to rely more on volun- sometimes negative, sometimes neutral, teers to do a lot of their campaign work. but seldom–as far as I can discover – positive.20 The case for inequality seems I began this inquiry by arguing that to rest entirely on the claim that it pro- America does less than almost any other motes ef½ciency, and the evidence for rich democracy to limit economic that claim is thin. All these judgments inequality. As a result, the rich can buy a are very tentative, however, and they are lot more in America than in other afflu- likely to change as more work is done. ent democracies, while the poor can buy Still, it is worthwhile to ask what they a little less. If you evaluate this situation would imply about the wisdom of trying by Rawlsian standards, America’s poli- to limit economic inequality if they cies are clearly inferior to those of most were, in fact, correct. rich European countries. If you evaluate Readers’ answers to that question the same situation using a utilitarian cal- should, I think, depend on four value culus, you are likely to conclude that judgments. First, readers need to decide most American consumers do better how much weight they assign to improv- than their counterparts in other large ing the lot of the least advantaged com- democracies. Much of this advantage is pared with improving the average level due to the fact that Americans spend of well-being. Second, they need to more time working than Europeans do, decide how much weight they assign to but that may not be the whole story. increasing material well-being compared I also looked at evidence on whether with increasing “family time” or economic inequality affects people’s “leisure.” Third, they need to decide lives independent of its effects on their how much weight they assign to equaliz- material standard of living. At least in ing opportunities for the young as the United States, the growth of inequal- against maximizing the welfare of ity appears to have made more people adults. Fourth, they need to decide how attend college but also made educational much value they assign to admitting opportunities more unequal. Growing more people from poor countries such as inequality may also have lowered life Mexico to the United States, since this expectancy, but the evidence for such an almost inevitably makes the distribution effect is weak and the effect, if there was of income more unequal. one, was probably small. There is some If you are a hard-core Rawlsian who evidence that changes in inequality af- thinks that society’s sole economic goal fect happiness in Europe, but not much evidence that this is the case in the Unit- 20 Mayer’s ½nding that inequality raises edu- ed States. If inequality affects violent cational attainment among the affluent is a crime, these effects are swamped by partial exception, since the increase among the affluent is larger than the decline among the other factors. There is no evidence that poor, making the net effect on educational changes in economic inequality affect attainment positive.
64 Dædalus Winter 2002 should be to improve the position of the why most rich societies are deeply divid- Does inequality least advantaged, European experience ed about the issue. Yet given the central- matter? suggests that limiting inequality can ity of redistribution in modern politics, bene½t the poor. If you are a hard-core it is remarkable how little effort rich utilitarian, European experience sug- societies have made to assemble the gests–though it certainly does not kinds of evidence they would need to prove–that limiting inequality lowers assess the costs and bene½ts of limiting consumption. But European experience inequality. Even societies that redistrib- also suggests that lowering inequality ute a far larger fraction of their gdp reduces consumption partly by encour- than the United States spend almost aging people to work fewer hours, which nothing on answering questions of this many Europeans see as a good thing. If kind. Answering such questions would you care more about equal opportunity require collecting better evidence, for children than about consumption which costs real money. It would also among adults, limiting economic require politicians to run the risk of inequality among parents probably being proven wrong. Nonetheless, moral reduces disparities in the opportunities sentiments uninformed by evidence open to their children. have done incalculable damage over the All things considered, the case for lim- past few centuries, and their malign iting inequality seems to me strong but influence shows no sign of abating. Rich not overwhelming. That is one reason democracies can do better if they try.
Dædalus Winter 2002 65 James F. Crow
Unequal by nature: a geneticist’s perspective on human differences
In February of 2001, Craig Venter, presi- Still, we only have to look around to dent of Celera Genomics, commenting see an astonishing variety of individual on the near-completion of the human differences in sizes, shapes, and facial genome project, said that “we are all features. Equally clear are individual dif- essentially identical twins.” A news ferences in susceptibility to disease–and headline at the time made a similar in athletic, mathematical, and musical point: Are We All One Race? Modern Sci- abilities. Individual differences extend to ence Says So. In the article that followed, differences between group averages. the author quoted geneticist Kenneth Most of these average differences are Kidd: “Race is not biologically de½nable, inconspicuous, but some–such as skin we are far too similar.” color–stand out. Venter and Kidd are eminent scien- Why this curious discrepancy between tists, so these statements must be rea- the evidence of dna and what we can sonable. Based on an examination of our clearly see? If not dna, what are the dna, any two human beings are 99.9 causes of the differences we perceive percent identical. The genetic differ- between individuals and between groups ences between different groups of of human beings? human beings are similarly minute. Dna is a very long molecule, com- posed of two strands twisted around James F. Crow, a Fellow of the American Acade- each other to produce the famous double my of Arts and Sciences since 1966, is professor helix. There are forty-six such dna mol- emeritus of medical genetics at the University of ecules in a human cell, each (along with Wisconsin. Over a career that has spanned more some proteins) forming a chromosome. than ½fty years, he and his collaborators have The dna in a human chromosome, if studied a variety of traits in Drosophila, dissected stretched out, would be an inch or more the genetics of ddt resistance, measured the in length. How this is compacted into a effects of minor mutations on the overall ½tness of microscopic blob some 1/1000 inch long populations, described the behavior of mutations without getting hopelessly tangled is an that do not play the game by Mendel’s rules, stud- engineering marvel that is still a puzzle. ied the effects of nonrandom mating, and consid- The “business” part of the dna, the ered the question “What good is sex?” part that carries genetic information, is the sequence of nucleotides, or bases, in
Dædalus Winter 2002 81 James F. the molecule. There are four of these, that time. As a result, we share roughly Crow on commonly designated as A, G, T, and C. 90 percent of our dna with mice, dogs, inequality (I could tell you what these letters stand cattle, and elephants. for, but you wouldn’t understand this Coming closer to home, the dna of essay any better if I did, so I won’t.) human beings and chimpanzees is 98 to In the double helix, there are four 99 percent identical. The differences kinds of base pairs: AT, GC, TA, and CG. between us that we (and presumably the The speci½c pairing rules–A with T and chimps) regard as signi½cant depend on G with C–are dictated by the three- only 1 or 2 percent of our dna. dimensional structure of the bases. Much of human dna is very similar to In a chromosome, the base pairs are in even more remote ancestors: reptiles, in- a precise sequence, and the orderly vertebrates, and even plants. All living process of cell division assures the repro- things share many functions (e.g., respi- duction of this sequence with remark- ration) going back to a very distant past. ably few errors. Chromosomes occur in Most of our dna determines that we are pairs, one member of each pair from human, rather than determining how we each parent, and the dna sites in the are different from any other person. So it two corresponding chromosomes match is not so surprising that the dna of any up. We have twenty-three pairs of chro- two human beings is 99.9 percent identi- mosomes, or a total of forty-six, as previ- cal. ously mentioned, in each cell. These What produces variability between forty-six chromosomes contain about individual organisms–and makes possi- six billion base pairs. If we randomly ble evolutionary change–is errors in the choose a pair of bases from correspon- dna copying process. Sometimes, ding sites in two persons, 99.9 percent of because of this, one base is changed to the time they will be the same. This per- another–it mutates. Among the six bil- centage depends only slightly on lion base pairs each of us inherits from whether the two people are from the our parents, a substantial number–a same or from different continents, from hundred or more–are new mutations. the same or from different population How can we reconcile this large num- groups. ber with the extremely slow rate of evo- In order to make sense of how the lutionary change? The explanation is dna of human beings can be so similar, that only a tiny fraction of mutations despite all the important visible and persist over time. Some mutations sur- physiological differences among individ- vive as a matter of either luck or–if the uals and groups, it is helpful to recount mutation confers a biological advan- our evolutionary history. tage–natural selection. Even if advanta- All mammals, including ourselves, are geous, an individual mutation has little descended from an ancestral species that chance of surviving a long evolutionary lived about one hundred million years trip. The slow rate of evolutionary ago. In our mammalian ancestry an aver- change explains why we mammals are so age base has changed, say from an A to a similar in our dna. T, at the almost unbelievably slow rate of Molecular studies of dna have been about one change per billion years. This extremely fruitful in working out the means that only a small fraction of the evolutionary history of life. Much of bases, one hundred million divided by what we know about human ancestry one billion, or 1/10, have changed during comes from dna studies, supplemented
82 Dædalus Winter 2002 by a rather spotty fossil record. The dna very tiny proportion of our dna, we dif- Unequal evidence strongly supports the idea that fer by a large number of dna bases. by nature the human species originated in Africa, Some noteworthy evolutionary and that European and Asiatic popula- changes in human beings have occurred tions–indeed, all non-Africans–are relatively rapidly, despite the slow over- descended from a small number of all rate of change at the dna level. The migrants from Africa. The strongest evi- difference between the skin color of dence for this is that Africans are more Africans and Europeans probably variable in their dna than are other pop- evolved in less than ½fty thousand years, ulations. an adaptation to differences in climate. Analysis of dna allows us to measure Still more rapid were changes in genes with some precision the genetic distance that confer resistance to malaria in between different populations of human Africa and Mediterranean regions; it beings. By this criterion, Caucasians and only took between four and eight thou- Asians are relatively similar, whereas sand years for the new genes to evolve. Asians and Africans are somewhat more What genetic analysis reveals is that different. The differences between the some of the genetic changes that seem so groups are small–but they are real. signi½cant to us depended on a very tiny dna analysis has provided exciting fraction of our dna. new answers to old questions. But its But, as I said, this tiny fraction is still a ½ndings can also be misleading. Take the very large number of bases. No two case of men and women and sex chro- human beings are alike in the traits they mosomes. Females have two X chromo- possess. Some are tall, others are short; somes, while males have an X and a Y. some are stocky, others thin; some are The Y chromosome makes up perhaps 1 gifted musically, others tone deaf; some percent of the dna. But there is very lit- are athletic, others awkward; some are tle correspondence between the Y and outgoing, others introverted; some are the other chromosomes, including the X. intelligent, others stupid; some can In other words, the dna of a human write great poetry or music, most can- male differs as much from that of a not. And so on. female as either does from a chimpanzee To understand our differences, we of the same sex. What does this mean? need to consider not just dna, but its Simply that dna analysis, which has cellular products as well. This area of given us a revolutionary new under- study is new, but it is progressing rapid- standing of genetics and evolution, ly. The emphasis is changing from dna doesn’t give sensible answers to some sequences to genes. A gene is a stretch of contemporary questions that society is dna, usually several thousand base pairs interested in. long. The function of most genes is to produce proteins. The genome sequenc- Most of the differences that we notice ing project has revealed that we humans are caused by a very tiny fraction of our have thirty to forty thousand genes. But dna. Given six billion base pairs per since a gene often produces more than cell, a tiny fraction–1/1000 of six billion one kind of protein, sometimes produc- base-pairs–is still six million different ing different kinds for different body base pairs per cell. So there is plenty of parts, the number of kinds of protein is room for genetic differences among us. more like one hundred thousand. Although we differ from each other in a We share a number of genes with
Dædalus Winter 2002 83 James F. chimpanzees, genes that make us pri- and Europeans. But, for the most part, Crow on mates rather than elephants or worms. group differences are small and largely inequality Evolutionary scientists believe that overshadowed by individual differences. many of the differences that we observe Biologists think of races of animals as between ourselves and chimpanzees groups that started as one, but later split involve changes in the amount rather and became separated, usually by a geo- than in the nature of gene products. graphical barrier. As the two groups Human beings and chimpanzees share evolve independently, they gradually proteins that produce body hair and diverge genetically. The divergences will brains, but in chimpanzees these pro- occur more quickly if the separate envi- teins produce more hair and less brains. ronments differ, but they will occur in Why this should be so is still far from any case since different mutations will being fully understood. But this is a inevitably occur in the two populations, research area that is advancing very rap- and some of them will persist. This is idly, and there are good genetic leads to most apparent in island populations, be followed up. where each island is separate and there is Of course, not every human difference no migration between them. Each one has a genetic cause. Many are environ- has its own characteristic types. In much mental, or are the result of interactions of the animal world, however, and also between genes and environment. Even in the human species, complete isolation genetically identical twins develop into is very rare. The genetic uniformity of distinct individuals. geographical groups is constantly being The ability to learn a language is large- destroyed by migration between them. ly innate, built into the nervous system In particular, the major geographical of all normal people, as demonstrated so groups–African, European, and beautifully in the effortless way in which Asian–are mixed, and this is especially young children learn to speak. But the true in the United States, which is some- particular language any individual learns thing of a melting pot. obviously depends on the social setting. Because of this mixing, many anthro- Mozart was a great composer partly pologists argue, quite reasonably, that because of his genes and partly because there is no scienti½c justi½cation for of his training. Ramanujan had a great applying the word “race” to populations talent for mathematics, but without his of human beings. But the concept itself being exposed to a textbook–not a very is unambiguous, and I believe that the good one, by the way–he could never word has a clear meaning to most peo- have made his astounding discoveries. ple. The dif½culty is not with the con- Michael Jordan has a talent for basket- cept, but with the realization that major ball, but it would never have developed human races are not pure races. Unlike had he grown up among the Inuits. those anthropologists who deny the use- fulness of the term, I believe that the Just as there are great differences among word “race” can be meaningfully applied individuals, there are average differ- to groups that are partially mixed. ences, usually much smaller, between Different diseases are demonstrably groups. Italians and Swedes differ in hair characteristic of different racial and eth- color. Sometimes the differences are nic groups. Sickle cell anemia, for exam- more conspicuous, such as the contrast- ple, is far more prevalent among people ing skin color and hair shape of Africans of African descent than among Euro-
84 Dædalus Winter 2002 peans. Obesity is especially common in Asian Americans represent about 12 Unequal Pima Indians, the result of the sudden percent of the California population, yet by nature acquisition of a high-calorie diet to they represent 45 percent of the student which Europeans have had enough time body at the University of California at to adjust. Tay-Sachs disease is much Berkeley. Asians have only slightly high- more common in the Jewish population. er average sat scores than Caucasians, There are other examples, and new ones but the university’s policy of admitting are being discovered constantly. students with the highest sat scores has The evidence indicating that some dis- yielded a much larger proportion from eases disproportionately afflict speci½c the group with the higher mean. ethnic and racial groups does not ordi- Two populations may have a large narily provoke controversy. Far more overlap and differ only slightly in their contentious is the evidence that some means. Still, the most outstanding indi- skills and behavioral properties are dif- viduals will tend to come from the popu- ferentially distributed among different lation with the higher mean. The impli- racial groups. There is strong evidence cation, I think, is clear: whenever an that such racial differences are partly institution or society singles out individ- genetic, but the evidence is more indi- uals who are exceptional or outstanding rect and has not been convincing to in some way, racial differences will everyone. become more apparent. That fact may To any sports observer it is obvious be uncomfortable, but there is no way that among Olympic jumpers and around it. sprinters, African Americans are far The fact that racial differences exist more numerous than their frequency in does not, of course, explain their origin. the population would predict. The dis- The cause of the observed differences proportion is enormous. Yet we also may be genetic. But it may also be envi- know that there are many white people ronmental, the result of diet, or family who are better runners and jumpers structure, or schooling, or any number than the average black person. How can of other possible biological and social we explain this seeming inconsistency? factors. There is actually a simple explanation My conclusion, to repeat, is that that is well known to geneticists and whenever a society singles out individu- statisticians, but not widely understood als who are outstanding or unusual in by the general public or, for that matter, any way, the statistical contrast between by political leaders. Consider a quantita- means and extremes comes to the fore. I tive trait that is distributed according to think that recognizing this can eventual- the normal, bell-shaped curve. iq can ly only help politicians and social policy- serve as an example. About one person makers. in 750 has an iq of 148 or higher. In a population with an average of about 108 These are times of very rapid change in rather than 100, hardly a noticeable dif- our understanding of biological process- ference, about 5 times as many will be in es. The genome project is but one exam- this high range. In a population averag- ple. At the same time, we are getting ing 8 points lower, there will be about 6 much closer to a deep understanding of times fewer. A small difference of 8 the nervous system and of human be- points in the mean translates to several- havior. Medical knowledge improves, as fold differences in the extremes. does data collection and computer
Dædalus Winter 2002 85 James F. analysis. All of these tell us more about that in the past has depended on obser- Crow on individual and group differences. What vation and statistical analysis of often inequality will be the impact of this new knowledge vaguely de½ned traits. We shall be able, on societal issues? What are the political as individuals, to know a great deal implications of modern biology? about our own genetic makeup. We have seen that the dna sequence The magni½cent advances in molecu- similarities revealed by the genome proj- lar biology will bring new depths of ect, valuable as these are for answering understanding of human differences, many interesting and important ques- normal and pathological, and the extent tions, are misleading in regard to impor- to which these are genetic or environ- tant human differences. But this situa- mental–or, as usually will be the case, tion is rapidly changing. The current both. Whether society will accept this emphasis goes beyond simple dna knowledge willingly and use it wisely I sequences to identifying the individual don’t know. My hope is that gradual genes, their products, and their complex progress, starting with small beginnings, interactions. At the same time, not only can lead to rational individual behavior the kinds of gene products (usually pro- and thoughtful, humanitarian social teins) but their relative amounts are policies. being investigated by much sharper new tools. Genes differ greatly in their pro- It is important for society to do a better ductivity, including differences in activi- job than it now does in accepting differ- ty in different parts of the body. ences as a fact of life. New forms of sci- In the near future, biologists will be enti½c knowledge will point out more able to tell us much more than we now and more ways in which we are diverse. I know about the genetic and environ- hope that differences will be welcomed, mental causes of human differences. The rather than accepted grudgingly. Who most obvious and immediate human wants a world of identical people, even if bene½ts will be in medicine. We can they are Mozarts or Jordans? foresee the time when many–we can A good society ought to provide the hope most–of our individual suscepti- best kind of environment for each per- bilities to disease will be understood, so son and each population. We already do that the disease can be predicted in this in part. We give lessons to musically advance, allowing doctors to anticipate gifted children. We encourage athletes and tailor treatments for the particular and give them special training (and person. Small steps in this direction have sometimes dubious drugs). Students already been made. New treatments are elect courses according to their abilities under development. As a result of our and interests. We have special classes for genetic understanding, we also now bet- those with disabilities, and such classes ter understand how to manipulate the are becoming more speci½c as the causes environment in order to help prevent of the disabilities are understood. disease. We cannot, of course, tailor-make a At the same time, the study of gene special environment for every individ- products and their regulation is being ual, but we can continue to move in this extended to normal traits. We can expect direction. Finding a genetic basis for a that the molecular biology of the future, trait doesn’t mean that environment is perhaps the quite near future, will pro- unimportant. Indeed, more environ- vide precisely the kind of information mental influences on the human organ-
86 Dædalus Winter 2002 ism are constantly being discovered, What about physicians? There may Unequal often through genetic studies. well be social considerations, perhaps by nature A test of our democratic institutions temporary ones in our society, that will be the degree to which people can would make race more important than accept all our differences and ½nd ways test scores in selecting students for med- to ½t them into a smooth-working, hu- ical schools. manitarian society. And I argue that we To achieve political and social equality should strive not only for maximum per- it is not necessary to maintain a ½ction sonal satisfaction but for maximum con- that important human differences do tribution; each of us owes society the not exist. The great evolutionist Theo- fruits of our special gifts. I believe dosius Dobzhansky said it well: “People strongly that research into the genetic need not be identical twins to be equal and environmental causes of human dif- before God, before the law, and in their ferences should continue and be sup- rights to equality of opportunity.” ported. The newer procedures brought I have emphasized that people differ, about by molecular advances and com- and differ greatly. They differ not only in puters will greatly accelerate discover- shapes and sizes, but also in abilities and ies. talents. They also differ in tastes and I believe that knowledge, even preferences. As Shaw said, “Do not do unpleasant knowledge, is far preferable unto others as you would that they to ignorance. I hope that American soci- should do unto you. Their tastes may ety can be less fearful of learning the not be the same.” Society’s business, I truth about biological inequalities and think, is not to minimize individual dif- more courageous in using discoveries in ferences. We shouldn’t try to ½t people ways that are humanitarian and pro- into one mold. mote human welfare. While I expect that science will con- The question of equal opportunity tinue to provide us with further evi- versus equal outcomes becomes particu- dence of human variability, and while I larly vexing in those occupations and welcome such variability as a source of professions for which only a small frac- social enrichment, there are some kinds tion of a population can qualify. I have of human variability that we could well already mentioned the gross overrepre- do without. I refer to serious, painful, sentation of African Americans among debilitating diseases. Many of these are Olympic runners. This is closer to a true the result of an unlucky throw of the meritocracy than anything else I can genetic dice. Already there are ways of think of: a stopwatch is color-blind. In discovering, preventing, and treating this case, there seems to be no social some of them. More treatments are sure purpose in demanding equal racial rep- to come. I hope they will be accepted resentation. willingly and used responsibly. I for one In some important professions, such would be content if the genes for Tay- as physics and engineering, Asian Amer- Sachs disease and Duchenne muscular icans are overrepresented and African dystrophy were to become extinct, along Americans underrepresented. We pre- with the malaria parasite and aids sumably get better research because of virus. I hope the great humanitarian this. This may or may not outweigh the bene½ts that could come from genetic inequity of unequal group representa- research will not be held up by fears of tion. That is a social decision. possible future misuse.
Dædalus Winter 2002 87 James F. Let me leave the last word for Jim Crow on Watson, co-discoverer of the double inequality helix and a major ½gure in the genome project: If the next century witnesses failure, let it be because our science is not yet up to the job, not because we don’t have the cour- age to make less random the sometimes most unfair courses of human evolution.
88 Dædalus Winter 2002 Martha C. Nussbaum
Sex, laws, and inequality: what India can teach the United States
In every house there is fear. night. Because her sleeping-car reserva- Let’s do away with that fear. tion had not yet been con½rmed, she Let’s build a women’s organization. contacted the train ticket examiner, who asked her to wait in the ladies’ waiting –“Mahila Samiti” (“A Women’s Organi- room. At around 5 p.m., two railway zation”), song sung all over India in of½cials came to con½rm her sleeping women’s groups berth; they also offered to show her to the station’s restaurant, where she could Hanuffa Khatoon, a citizen of get dinner before the departure. Ms. Bangladesh and also an elected of½cial of Khatoon followed a station-boy to the that nation’s Union Board, arrived at restaurant and ordered some food, but Howrah Station in Calcutta, India, on immediately began to vomit. She the afternoon of February 26, 1998, plan- returned to the ladies’ waiting room, ning to catch the Jodhpur Express that quite ill. The railway of½cials then offered to take her to the of½cial station hotel managed by the Railways Board. Martha C. Nussbaum, Ernst Freund Distin- She insisted on checking their creden- guished Service Professor of Law and Ethics at the tials ½rst, but when the of½cial on duty University of Chicago, is appointed in the philoso- at the ladies’ waiting room told her that phy department, Law School, and Divinity their credentials were in order, she School. She is an associate in the classics depart- agreed to go. In the hotel room she was ment, an af½liate of the Committee on Southern brutally gang raped for several hours by Asian Studies, and a member of the board of the a group of four station employees. Final- Center for Gender Studies. She has been a Fellow ly she escaped and returned to the plat- of the American Academy since 1988. Former form, bleeding and in a state of shock. president of the American Philosophical Associa- There she found another railway of½cial tion (Central Division), Nussbaum chaired the who pretended to assist her. He said he association’s Committee on the Status of Women would take her to his wife, who would as well as its Committee on International Cooper- take care of her until she could get ation. Her most recent books are “Women and another train in the morning. At the Human Development: The Capabilities wife’s alleged residence she was brutally Approach” (2000) and “Upheavals of Thought: gang raped again, and two of the em- The Intelligence of Emotions” (2001). ployees tried to suffocate her. Hearing
Dædalus Winter 2002 95 Martha C. her cries, the landlord called the police, voting slips, with her number on them, Nussbaum who ½nally rescued her.1 to be given to voters on election day.2 A on inequality What is signi½cant–and speci½cally gentle, soft-spoken woman, Poonam Indian–about this story, however, is not Devi has for two years been president of the sad fact of gang rape, familiar a woman’s collective, where she has throughout recorded history in all helped to arrange loans for all ½fteen nations. What is signi½cant is its members of her group. dénouement. What is most astonishing about Two years later, in an unprecedented Poonam Devi’s campaign, however, is judgment, Ms. Khatoon won a large not the fact of her candidacy–but the damage award from the Railways Board. fact that she is running against her hus- It was a landmark case in which the band, who is af½liated with the bjp Supreme Court of India declared rape to (Bharatiya Janata Party, the currently be a violation of the fundamental right dominant party nationwide, with a to live with human dignity, under both Hindu fundamentalist program). Origi- the Indian Constitution and the Uni- nally it was thought that this constituen- versal Declaration of Human Rights. cy would be among those reserved for “Rape,” wrote the Court, “is a crime not women in the current election, so only against the person of a woman, it is Poonam Devi’s husband groomed her a crime against the entire society. It for candidacy, assuming that he would destroys the entire psychology of a be unable to run. But when the electoral woman and pushes her into deep emo- plan was announced, the constituency tional crisis. Rape is therefore the most was not reserved for women, and the hated crime. It is a crime against basic husband could run. But Poonam Devi human rights and is violative of the vic- decided to run anyway, with support tim’s most cherished right, namely, right from her parents and brothers. Her hus- to life which includes right to live with band asked her to withdraw, but she human dignity. . . .” refused. He is angry. After all, he says, she is a weak and insigni½cant candidate It is a mid-April evening in Bihar, in next to him. He is educated, he owns northeastern India. A woman is sitting some land, he has been a teacher–and, with her brother in the backyard of her he points out, he is even unemployed, so mud hut in a poor area of this state, one he has lots of time for the council. A of the most corrupt and anarchic in the reporter from the national news media nation. Women have traditionally had asks Poonam Devi, “Why are you ½ght- little political power in Bihar, where, in ing against your husband?” She ques- some regions, the sex ratio is as low as tions right back: “Why can’t I ½ght the 75 women to 100 men–a ½gure indica- elections, husband or no husband? Why tive of the differential nutrition and can’t a woman and a man be candidates health care of girls, sex-selective abor- from the same family?” Her platform tion, and, probably, outright infanticide. focuses on unemployment, the old-age But Poonam Devi, mother of two girls, is pension, and the insecure economic a candidate for election to her panchayat, or local council, and she is arranging the 2 In Indian elections, voters receive slips with the symbol of each candidate, and they then deposit the slip of their choice in the box–a 1 Chairman, Railway Board v. Mrs. Chandrima procedure designed to make voting easy for Das air 2000, SC 988. illiterate voters.
96 Dædalus Winter 2002 position of single women and widows.3 boyfriend, husband, or date.4) Sex, laws, and The outcome of Poonam Devi’s candi- Nor is lack of political power a distant inequality dacy remains unclear. What is clear, dif½culty. Women in the United States however, is that the Seventy-Second and hold only 13.8 percent of its national leg- Seventy-Third Amendments to India’s islative seats–one of the lowest ½gures constitution, which establish a bold pro- among the developed nations, according gram of af½rmative action for women in to the Human Development Report 2001. the local panchayats, are bringing large And in no nation does the ½gure come numbers of women into politics all over very close to equality: Sweden and Den- India, with clear results for the salience mark take the lead, with 42.7 percent and of issues pertaining to the welfare of 37.4 percent, respectively; outside the women and children. Nordic countries, the highest ½gures are for the Netherlands at 32.9 percent and Inequality on the basis of sex is a stag- Germany at 30.4 percent; highest in the gering problem worldwide. India is developing world is South Africa at 27.9 hardly unique in this regard. Women in percent. all nations–including the United But women are also contesting age-old States–still suffer serious inequalities forms of subordination with increasing in at least some central areas of human success, creating innovative proposals life. for change in both custom and law. And Gang rape is hardly a problem indige- sometimes nations that are widely per- nous to Calcutta: it is the regular fare of ceived as lagging behind the “advanced u.s. courts. (A recent showing of Law democracies” of the United States and and Order reruns managed to ½ll an Europe can actually take the lead, with entire evening with programs on this bold measures like those that altered the one theme, most of them based on real lives of Hanuffa Khatoon and Poonam stories.) And it is just one especially ter- Devi. rible aspect of the general worldwide In this essay I shall look at the problem problem of violence against women, a of women’s inequality through the lens problem that seems to be particularly of today’s India, a nation with both grave in the United States. (According to enormous gender problems and rich a report recently published in the Journal political creativity. I shall begin by offer- of the American Medical Association, one- ing a thumbnail sketch of the situation ½fth of the Massachusetts high-school of women in India and of the Indian girls studied have suffered some type of constitutional tradition, which has been violence from a date, either assault or remarkably woman-friendly, and discuss sexual violence. A recent national study conceptions of equality and the role of concludes that 25 percent of adult law that offer rich resources for those women have experienced violence from seeking to advance women’s position in a romantic partner. The Justice Depart- society. I shall then return to the cases ment estimates that more than 1.5 mil- with which I began, showing how a rea- lion u.s. women experience physical or sonable conception of af½rmative action sexual violence each year from a and a reasonable openness to the norms
3 See Mukul Sharma, “Bihar: Making of a Pan- 4 Erica Goode, “Study Says 20% of Girls chayat Election,” Economic and Political Weekly, Reported Abuse by a Date,” New York Times, 1 12 May 2001. August 2000, A10.
Dædalus Winter 2002 97 Martha C. of the international community (both deny to any person “equality before the Nussbaum on rather lacking in current u.s. politics) law or the equal protection of the laws.” inequality have enabled India to progress. Article 15 prohibits state discrimination “on grounds only of religion, race, caste, It is extraordinarily dif½cult to sum up sex, place of birth or any of them.” Oth- succinctly the situation of women in er rights that are highly relevant to sex India, since there is probably no nation equality include Article 13 (invalidating in the world with greater internal diver- all laws inconsistent with the Funda- sity and plurality. In what follows I shall mental Rights); Article 16 (equality of be mentioning some of those differences opportunity in public employment); (of caste, religion, regional background, Article 19 (protecting freedom of speech wealth and class, and still others). All and expression, freedom of association, generalizations cover multiple differ- freedom of travel, freedom of residence, ences. and freedom to form labor unions); India celebrated the ½ftieth anniver- Article 21 (stating that no citizen shall be sary of its independence from Britain on deprived of life or liberty “except ac- August 15, 1997. It is the world’s largest cording to procedure established by democracy, with a population of 846.3 law”); Article 23 (prohibition of traf½c million. It is a constitutional parliamen- in human beings and forced labor); and tary democracy, with a written account Article 25 (freedom of conscience and of Fundamental Rights containing the religion). (Article 17 abolishes untoucha- abolition of untouchability and an elab- bility: “its practice in any form is forbid- orate set of equality and nondiscrimina- den.”) tion provisions. Its legal system is in The understanding of equality in the some respects similar to (and modeled Constitution is explicitly aimed at secur- on) that of the United States, combining ing substantive equality for previously a basically common-law tradition with subordinated groups. The framers care- the constraints of a written constitution fully distanced their conception from including the extensive list of Funda- the idea, already familiar in those days, mental Rights. Its Supreme Court, like that equality requires treating everyone ours, is the ultimate interpreter of these the same and not using race or sex as rights. grounds for any type of differential treat- India’s Constitution is in some ways ment–an understanding that has been very attuned to issues of sex equality, used in the United States to subvert which were prominently debated when af½rmative action. In India, by contrast, the Constitution was adopted in 1950. the Constitution’s so-called Directive The framers of the Constitution were Principles of State Policy (a nonenforce- very conscious of deeply entrenched able section of the Constitution) devotes inequalities, both those based on caste a great deal of attention to promoting and those based on sex, and they made economic equality, and the Fundamental the removal of them one of their central Rights are themselves speci½ed in a way goals. The text of the Constitution is in that makes room for af½rmative-action many ways exemplary in its treatment of programs designed to advance the mate- issues of gender and sex, particularly in rial situation of women and the lower the section dealing with Fundamental castes. Rights. Thus, Article 15 states that “Nothing in Article 14 says that the state shall not this article shall prevent the State from
98 Dædalus Winter 2002 making any special provision for women de½ned by laws passed in Parliament, Sex, laws, and and children,” and that “Nothing in this but they assign to religious bodies con- inequality article. . . shall prevent the State from siderable power in the areas of marriage, making any special provision for the divorce, child custody, and property. advancement of any socially and educa- There are some individual secular laws tionally backward classes of citizens or of property, marriage, and divorce, but for the Scheduled Castes and the Sched- they do not form a system, and, because uled Tribes.” Similar clauses appear in one is typically classi½ed into a religious Article 16 (equality of opportunity in system at birth, it is not so easy for indi- public employment) and in Article 19 viduals to disengage themselves, partic- (various other rights and liberties). Even ularly when property is jointly owned in before independence, quotas and other family consortia (as it often is) from af½rmative-action measures for de- which individuals may not extricate prived groups were an accepted part of their shares. These systems of personal the Indian scene, and they became even law have made it uniquely dif½cult to more salient at independence. In short, end discrimination based on caste and the framers understood the goal of sex.6 To explore these dif½culties, how- equality in terms of an end to systematic ever, would take us rather far from our hierarchy and discrimination based on primary topic. both caste and sex. Unlike the United States, India is an In light of this tradition it is not sur- extremely poor nation. It ranks 115th out prising that India has long been a center of the 162 nations of the world on the of thought and planning about sex equal- Human Development Index of the 2001 ity, or that, when the United Nations Human Development Report. The average Development Programme needed a life expectancy at birth is 62.9 (as op- major report on gender and governance, posed to 80.8 in Japan, 76.8 in the United it turned the writing of this report over States, and somewhere between these to its New Delhi of½ce.5 two numbers in Canada and most of There is one great structural differ- Europe7), and infant mortality is high, ence between the Indian legal system at 70 for 1,000 live births (although this and the Anglo-American systems to represents a great decline from 165 in which it is related: India has no uniform 1960). code of civil law (even within each Women do even worse than men in region). Criminal law is uniform for the basic nutrition and health. If equal nation as a whole and is administered by nutrition and health were present, it is the state. But with the exception of com- mercial law, which was uniformly codi- 6 See my discussion in Women and Human Development: The Capabilities Approach (New ½ed for the nation as a whole by the York: Cambridge University Press, 2000), British and has remained so, civil law chap. 3. remains the province of the various reli- gious systems of law–Hindu, Muslim, 7 Ireland and Denmark are the only nations in Western Europe to have lower life expectancy Parsi, and Christian. These systems are than the United States, although most of the nations of Eastern Europe and the former Sovi- 5 This report will be published shortly; its et Union also have lower expectancy. Also authors include citizens of India, Sri Lanka, and ahead of the United States are Australia, New the United States. (I wrote the introduction and Zealand, Israel, Hong Kong, Cyprus, Singapore, the discussion of issues of sex equality within and Malta; Costa Rica and Barbados are close the family.) (76.2 and 76.6, respectively).
Dædalus Winter 2002 99 Martha C. estimated that the sex ratio would be tion compulsory, these laws have little Nussbaum on approximately 103 women to 100 men. relation to reality. Many regions utterly inequality India’s sex ratio has not been even 1:1 at lack schools of any kind, just as they fre- any time since measurements began in quently lack reliable electricity, medical the early twentieth century. From a high services, water, and decent roads; many of 97 women to 100 men in 1901, the local functionaries are corrupt, and so ratio dropped steadily, reaching a low of teachers in many regions take pay with- around 93:100 in 1971; after a slight rise, out ever even showing up in the region it declined again even further, reaching where they are supposed to be teaching. 92.7:100 in 1991. These are of½cial ½g- In some rural areas, female literacy is as ures. Things are probably much worse, low as 5 percent. The national govern- at least in some regions. A house-to- ment, though well-intentioned, has done house count by a good ngo in rural little to ½ll these gaps, although some Bihar arrived at a ratio of 75:100, and a adult education programs have been similar count in a region of Karnataka established in some of the poorer states, found 65:100. Some of these differences and many nongovernmental organiza- should be attributed to the differential tions run both adult education programs nutrition of boys and girls and to un- and after-work programs for working equal health care, but sex-selective abor- girls. tion and active infanticide are playing an Still, this does not seem to be a neces- increasing role. A recent study by the sary or unbreakable pattern, since some Indian Association of Women’s Studies otherwise poor regions have done estimates that 10,000 female fetuses are extremely well. Kerala has adult literacy aborted every year. Some regions tell a of 90 percent and near-universal literacy very different story: Kerala, for example, among adolescent boys and girls. This has more women than men. (This situa- remarkable record is the outcome of tion results from a combination of rela- more than a hundred years of concerted tively female-friendly traditions and public action. Recently a constitutional gender-friendly state governance.) But amendment was introduced that would clearly, on the whole, women face spe- make the right to education a justiciable cial obstacles in India. fundamental right in India.8 It may be In education, the male-female gap is hoped that the passage of this amend- even more striking: the adult literacy ment will goad government into acting rate for women is 44.5 percent, as against more aggressively on its good intentions. 67.8 percent for men. (In China, the ½g- ures are 75.5 percent for women and 91.2 Among the greatest obstacles to fully percent for men.) Such statistics are equal citizenship that women face, in all hard to interpret, since local govern- nations, is their unequal exposure to sex- ments tend to be boastful and since it is based violence.9 In India the problem of hard to establish a clear measure of liter- acy. Yet what is unambiguously clear is 8 Amendment 83, to be inserted in the Funda- that, despite the fact that education is a mental Rights section of the Constitution as Article 21a. See the full text of the amendment state responsibility, India has done very in From the Lawyers Collective 13 (April 1998): 10. badly in basic education across the board, and even worse in basic educa- 9 For data on the United States, and the failure of law to deal adequately with these problems, tion for women. Although all Indian see chap. 5 of my Sex and Social Justice (New states have laws making primary educa- York: Oxford University Press, 1999).
100 Dædalus Winter 2002 violence against women is compounded, women’s groups also address domestic Sex, laws, and often, by the low age of marriage and the violence directly, and politicians such as inequality lack of economic options for a woman Poonam Devi ½ght to make life a bit fair- with little or no education. The marriage er for widows and single women, two of girls as young as four or six, although groups that suffer greatly from discrimi- long since illegal, is a common reality, nation and vulnerability to violence. especially in some regions where it is Rape, however–in India as in so many traditional. Laws against it are not other nations–has been badly dealt with enforced, and it shapes a girl’s life from under the law for many years, and the birth, often discouraging her family number of rapes appears to be on the from educating her. rise. It is easy to ½nd cases in which Within marriage, at all ages, domestic acquittal was secured on the grounds violence is so pervasive that three states that the woman was of low caste, or have adopted alcohol prohibition laws in “immodest,” even when there is ample response to women’s lobbying in an evidence of forcible rape in the particu- effort to reduce such violence. Police do lar instance. Rape is also used as a not aggressively investigate domestic weapon against women crusading for abuse, and virtually no women’s shel- political change. In 1993 Bhanwari Devi, ters exist. Rape within marriage is not a member of the state of Rajasthan’s even illegal. Thus, women who wish to Sathin movement for women’s welfare, protect themselves against marital vio- was campaigning against child marriage lence have few options. If they are not when she was gang-raped by men from a equipped for employment outside the community that supports the practice of home, they have virtually no exit op- child marriage. Because the men were tions; many women endure lives of influential community leaders, police abuse because they know that prostitu- refused to register the case until it was tion is their only alternative. too late to perform the necessary med- The problem of domestic violence is ical examination; a lower court in Jaipur being addressed, above all, through edu- acquitted all the accused. Although cation, credit, and economic options. Bhanwari appealed this judgment and Hundreds of nongovernmental organi- the Rajasthan High Court agreed in 1996 zations, from the large Self-Employed to hear her appeal, arguments in the case Women’s Association (sewa), with over have not yet been heard. ½fty thousand members, to the small vil- In general, delays in the criminal jus- lage-based women’s collective led by tice system often create a lapse of ten Poonam Devi, have been educating girls years between rape and court date, mak- and women outside the formal state ing it very dif½cult for women to pursue structure, lending them money, and their cases, even when they want to. teaching them employment-related Often they don’t want to, because a skills so that they can do something on woman’s sexual history is still admitted their own if they decide to leave a bad as evidence, and assumptions about the marriage. Education, credit, and the woman’s behavior and dress continue to reform of antiquated property laws to influence the resolution of rape trials. give women land rights in their own Defendants can usually win a continu- names are probably the three most sig- ance on the flimsiest of pretexts, and ni½cant strategies against domestic vio- their strategy typically is to delay and lence. At the same time, most local delay until the woman gives up the pros-
Dædalus Winter 2002 101 Martha C. ecution. A friend of mine who is a pro- Recently, however, the Supreme Court, Nussbaum on fessor of philosophy and women’s stud- at least, has shown greater sensitivity to inequality ies at University of Lucknow urged a for- the issue of sexual violence. Hanuffa mer student to pursue her rape com- Khatoon’s case shows a determination plaint and promised to join her in court to confront the problem head-on, using whenever the case surfaced–until, after the resources of the constitutional tradi- ½ve years, the woman had remarried, tion, which has already held that the and didn’t want to think about her rape right to life guaranteed in Article 21 any longer. includes a right to life with human digni- One case that spurred awareness of ty. (The landmark case was one defend- women’s grievances in this area was the ing the rights of the homeless.) In an 1979 case of Mathura, a sixteen-year-old earlier case not centrally dealing with tribal woman who was raped by two rape, the Court had already opined that policemen within a police compound. rape is a constitutional issue, and they The lower court acquitted the policemen quoted from that case at the outset of on the grounds that Mathura had eloped their opinion in Hanuffa Khatoon’s case, with her boyfriend and hence was declaring that rape is a “crime against “habituated to sexual intercourse”; they the entire society” because it “destroys thus reasoned that she could not be an the entire psychology of a woman.” It is unconsenting victim–therefore she was therefore a “crime against basic human not, technically, raped. The High Court rights” and a violation of the right to life overturned the decision, holding that with dignity guaranteed under Article 21. mere passive surrender under threat can- This judgment the justices then applied not be counted as consent to inter- to Ms. Khatoon’s gang rape by the rail- course. The Supreme Court, however, way employees. The justices argued, reinstated the lower court decision. moreover, that the fundamental right to This judgment triggered widespread life with dignity belongs not only to citi- public protest and publicity; rape and zens of India, but to all “persons” (like rape law were discussed widely and the Bangladeshi visitor Ms. Khatoon) openly for the ½rst time. Four Delhi Uni- within the territory of India. versity law professors wrote a petition to Then, in a most interesting discussion, the Supreme Court calling for a rehear- the courts pointed out that the Funda- ing of the case. The petition, unfortu- mental Rights are closely modeled on nately, was dismissed. It did, however, the list of rights in the un’s Universal energize the women’s movement to Declaration of Human Rights. They demand legal change. More important, a mention particularly the declaration’s law commission was set up by the gov- emphasis on equal human dignity (Arti- ernment to consider changes in rape law. cle 1); the right to life, liberty, and secu- One signi½cant result was a shift in the rity of person (Article 3); the prohibition burden of proof in custodial rape cases, of “cruel, inhuman or degrading treat- as well as a set of mandatory minimum ment” (Article 5); the guarantee of sentences for rape. Other feminist de- nondiscrimination and the equal protec- mands, such as the demand that a tion of the laws (Article 7); and the pro- woman’s prior sexual history should not hibition of arbitrary detention (Article be deemed relevant evidence, were not 9). They argue that the purpose of the included in the version of the new legis- section on Fundamental Rights in the lation that was passed in 1982. Indian Constitution was to enact the
102 Dædalus Winter 2002 Universal Declaration and “to safeguard India. In another signi½cant judgment Sex, laws, and the basic human rights from the vicissi- concerning sexual harassment, the inequality tudes of political controversy. . . .” This Supreme Court ruled that the guidelines being so, the meaning of the word “life” on harassment in the Convention on the in the Indian Constitution can be fur- Elimination of All Forms of Discrimina- ther interpreted with reference to the tion Against Women (cedaw) are bind- declaration. They note that earlier ing on the nation through its rati½cation Supreme Court decisions have already of that treaty.12 In this way the universal given “life” a broad construction, human rights guaranteed in treaties may including the idea of life with human enter a nation without violation of its dignity. Since gang rape is obviously democratic sovereignty and after due inconsistent with human dignity, and deliberation by the body that has been the rape was committed by government entrusted with the interpretation of fun- employees, the judgment of the Calcutta damental rights. High Court awarding Ms. Khatoon dam- In short, when a nation understands ages from the Railways Board was itself to be a member of the world com- upheld. munity, committed to taking its treaty This creative judgment shows how a obligations seriously, creative legal legal tradition can be fruitfully mined to change may ensue. Unfortunately, the give women redress against violence. United States is currently reverting to Thus far, it has a function similar to that old isolationist habits, giving the of the u.s. Violence Against Women impression that it does not need to con- Act, passed by Congress in 1994, which sult with any other nation and that it is offered victims of sex crimes a federal powerful enough to show disdain for the avenue of redress, given the evident world community. unevenness and unreliability of the criminal justice system in the states.10 The surprising candidacy of Poonam (Of course, our Supreme Court, moving Devi also is the fruit of creative constitu- in the opposite direction from its Indian tional thinking. At the time of India’s counterpart, has declared the 1994 Vio- founding, in keeping with the generally lence Against Women Act unconstitu- substantive understanding of equality in tional on the grounds that it allegedly its Constitution, various schemes of exceeds the power of Congress.11) But af½rmative action on behalf of tradi- the Indian Supreme Court’s judgment tionally subordinated groups were con- shows something more: it shows that a templated. The Constitution created a national legal tradition may deepen and system of representation meant to strengthen its fundamental rights reflect the proportion of every caste and through incorporation of the rights tribe in the total population of each guaranteed in the international docu- state. The system works by a complex ments it has rati½ed. scheme of rotations: in successive elec- This move has been made before in tions, only members of certain groups may run for of½ce, although all citizens 10 On this issue see Stephen J. Schulhofer, Unwanted Sex: The Culture of Intimidation and the 12 Vishaka v. State of Rajasthan, AIR 1997, 6 SCC Failure of Law (Cambridge, Mass.: Harvard Uni- 241. Compare the invocation of CEDAW as versity Press, 1998). binding on the nation in a nationality case in Botswana: Attorney General v. Unity Dow, 1992 11 U.S. v. Morrison, 120 S. Ct. 1740 (2000). LRC (Cons) 623 (July 2, 1992).
Dædalus Winter 2002 103 Martha C. may vote. Despite controversy and many arguing that they would compromise the Nussbaum on complaints, the system seems to have struggle for women’s full equality. At inequality worked reasonably well, effectively independence, accordingly, reservations enfranchising a variety of previously dis- for women were rejected, although, as advantaged groups and promoting their noted above, af½rmative action on the economic and social well-being. There is basis of sex won general support in the little doubt that it would have been dif½- Constitution. In 1971, the government cult to achieve progress against the appointed a Committee on the Status of deeply entrenched realities of caste Women in India to study the progress without such af½rmative legal measures. that had been made by women since On the other hand, legitimate objec- independence. In its famous 1974 report tions can be made to the system. First of Towards Equality, the committee deliv- all, no reserved seats have ever been cre- ered a scathing critique of the political ated or even seriously championed for process, arguing that the political posi- Muslims, arguably as vulnerable a tion of women in India had, if anything, minority as the Scheduled Castes and worsened since 1950, and that women Tribes. Second, the practice of reserva- were neither able to claim their legal tion has led over time to a situation in rights nor, in many cases, even aware of which castes at the very bottom of the them. The majority of the committee social ladder do considerably better than continued to oppose reserved seats as a those just above them. Thus, more remedy, but a minority report signed by recently, as a result of the 1980 report of some especially prominent feminist the Mandal Commission, reservations leaders argued that this remedy was nec- for obcs (Other Backward Castes) were essary for the resumption of social and added to the list. (Estimates of the pro- political progress for women. portion of India’s population that A generation later, the representation belongs to obc groups range from 25 of women in central and state govern- percent to 37 percent, and many of these ment continues to be very low: 6–7 per- people are economically advantaged. cent in the Lok Sabha (the analogue of Thus it can now be argued that the sys- the House of Commons), one of the low- tem of reservations no longer protects est parliamentary ½gures in the world. the most vulnerable and otherwise Political parties have talked about unrepresented groups.) As a result of the reserving a certain proportion of their system of representation, a politics of own candidacies for women, but have caste has to some extent displaced a poli- done nothing about it. At the same time, tics of national issues, and recent gov- women’s voter turnout has signi½cantly ernmental instability at both regional increased and is now at 55 percent, only and national levels can be partially slightly less than the national average. In attributed to the proliferation of caste- this situation, it is not surprising that the based parties. Despite these problems, idea of reserved legislative seats for however, the quota system seems to women has attracted new political and most Indians to be a source of more constitutional attention, in connection good than harm, and there is no serious with a push for greater local self-rule. demand for its abolition. Arguing, like John Stuart Mill, that Reserved seats for women have been participation in local politics teaches cit- discussed since before independence. izens how to appreciate the common Early feminists opposed reservations, good, national legislators successfully
104 Dædalus Winter 2002 amended the Constitution in 1992 to been possible without the Amendments. Sex, laws, and give formal legal status to the system of In addition, the system has increased inequality panchayats, or local village councils, an demands for female education: mothers aspect of governance central to Gandhi’s can now urge their daughters to go to vision of India but never fully imple- school in order to prepare themselves for mented. The Amendments established a role in politics. They report that this a 33 percent quota for women in the gives them more power in the family to panchayats and set up a system of rota- decide which children shall go to tion that is similar to that by which reser- school.13 vations for lower castes have already More recently, proposals to introduce been implemented at the national level. reservations for women at the national Initially, advocates for women were level have encountered tremendous split about the merits of this system. opposition–largely from lower-caste Many feared that the women who would parties, who fear that the new quotas be selected would simply be tools of would result in fewer lower-caste legisla- male interests. But nearly ten years of tors, since they believe that educated experience with the plan has shown that, women will be the most likely to be on balance, its merits outweigh its draw- elected. They propose a subquota in the backs. Certainly in some cases women general women’s quota for lower-caste do initially function as proxies for the women, but so far proponents of the powerful men in their families. Poonam Amendment have rejected this proposal. Devi was initially groomed for of½ce by Certainly such a quota for lower-caste her husband, who believed that he women would exacerbate some of the would be unable to run for the seat. But problems already produced by caste- even such women learn political skills in based reserved seats at the national the process. Poonam Devi became so level. A possible outcome of the current interested in politics that she is now run- debate is that parties will agree to ning for of½ce against the wishes of her reserve a certain proportion of their tick- husband. Whether she wins or loses, she ets for female candidates (as they do in is gaining valuable experience; if she France and quite a few other countries). loses, in due course she will be able to run for a reserved seat. It is ironic–and telling–that similarly Moreover, the new system’s extension creative proposals are non-starters in the of political power to poor and illiterate United States. Even though the United women has been dramatic. Studies show States has one of the lowest proportions that a majority of women who serve in of women in the national legislature the panchayats are illiterate or barely lit- within the developed world, we are not erate. Moreover, approximately 40 per- looking around with genuine curiosity to cent of female representatives come see what other nations have done about from families with income below the poverty line. Women report many obsta- 13 On all these matters, see Nirmala Buch, cles to their effective participation, From Oppression to Assertion: A Study of including harassment and the threat of Panchayats and Women in Madhya Pradesh, violence. Nonetheless, a number of Rajasthan and Uttar Pradesh (New Delhi: Centre for Women’s Development Studies, 1999); I am women are evidently learning political also grateful to Zoya Hasan and to Niraja Gopal skills and participating in decision- Jayal for allowing me to read unpublished making in a way that would not have works on this topic, and for discussion.
Dædalus Winter 2002 105 Martha C. this problem. ple voting and proportional representa- Nussbaum on This is not to say that the Indian solu- tion, which have been used successfully inequality tion ½ts the u.s. situation. Very likely by some municipalities for years to it does not. Quotas for women in the enfranchise underrepresented groups, panchayats are a solution well adapted to should also be considered. In general, we the situation of the rural poor in India, should attend to the issue, debate it where illiteracy and lack of employment without phobic reactions (such as the outside the home pose daunting obsta- term “af½rmative action” so often cles to women’s political participation. evokes), and learn from other nations. In the United States, by contrast, many Both of these issues show us one large more women already work–and no real fact: the world is moving on, with or equivalent of India’s panchayats exits. without u.s. participation, to ½nd cre- Still, Indian politicians and jurists are ative solutions to pressing problems of thinking–as ours have too often refused human inequality. Usually u.s. citizens to think–creatively. We should more don’t know anything much about these vigorously confront the problem of vio- developments, and some of our politi- lence against women and the problem of cians encourage disdain for what is hap- the underrepresentation of women in pening elsewhere. We need to learn new politics by considering a wide range of habits of curiosity and respect if we are remedies–½rst on the list being cam- to be productive members of an increas- paign ½nance reform, which has at least ingly interdependent global community. received a hearing. But systems of multi-
106 Dædalus Winter 2002 Note by All too often, we ignore goals, genres, or Frank Wilczek values, or we assume that they are so apparent that we do not bother to high- light them. Yet judgments about whether an exercise–a paper, a project, an essay Frank Wilczek response on an examination–has been done intelligently or stupidly are often on the dif½cult for students to fathom. And since these evaluations are not well world’s understood, few if any lessons can be numerical drawn from them. Laying out the criteria recipe by which judgments of quality are made may not suf½ce in itself to improve qual- ity, but in the absence of such clari½- cation, we have little reason to expect our students to go about their work intelligently.
Twentieth-century physics began around 600 b.c. when Pythagoras of Samos pro- claimed an awesome vision. By studying the notes sounded by plucked strings, Pythagoras discovered that the human perception of harmony is connected to numerical ratios. He examined strings made of the same material, having the same thickness, and under the same tension, but of different lengths. Under these conditions, he found that the notes sound harmonious precisely when the ratio of the lengths of string can be expressed in small whole numbers. For example, the length ratio
Frank Wilczek, Herman Feshbach Professor of Physics at MIT, is known, among other things, for the discovery of asymptotic freedom, the develop- ment of quantum chromodynamics, the invention of axions, and the discovery and exploitation of new forms of quantum statistics (anyons). When only twenty-one years old and a graduate student at Princeton University, he and David Gross de½ned the properties of gluons, which hold atom- ic nuclei together. He has been a Fellow of the American Academy since 1993.
142 Dædalus Winter 2002 2:1 sounds a musical octave, 3:2 a musi- only to unravel completely. The world’s numerical cal ½fth, and 4:3 a musical fourth. Students today still learn about recipe The vision inspired by this discovery Kepler’s three laws of planetary motion. is summed up in the maxim “All Things But before formulating these celebrated are Number.” This became the credo of laws, this great speculative thinker had the Pythagorean Brotherhood, a mixed- announced another law–we can call it sex society that combined elements of Kepler’s zeroth law–of which we hear an archaic religious cult and a modern much less, for the very good reason that scienti½c academy. it is entirely wrong. Yet it was his discov- The Brotherhood was responsible for ery of the zeroth law that ½red Kepler’s many ½ne discoveries, all of which it enthusiasm for planetary astronomy, in attributed to Pythagoras. Perhaps the particular for the Copernican system, most celebrated and profound is the and launched his extraordinary career. Pythagorean Theorem. This theorem Kepler’s zeroth law concerns the relative remains a staple of introductory geome- size of the orbits of different planets. To try courses. It is also the point of depar- formulate it, we must imagine that the ture for the Riemann-Einstein theories planets are carried about on concentric of curved space and gravity. spheres around the Sun. His law states Unfortunately, this very theorem that the successive planetary spheres are undermined the Brotherhood’s credo. of such proportions that they can be Using the Pythagorean Theorem, it is inscribed within and circumscribed not hard to prove that the ratio of the about the ½ve Platonic solids. These ½ve hypotenuse of an isosceles right triangle remarkable solids–tetrahedron, cube, to either of its two shorter sides cannot octahedron, dodecahedron, icosahe- be expressed in whole numbers. A mem- dron–have faces that are congruent ber of the Brotherhood who revealed equilateral polygons. The Pythagoreans this dreadful secret drowned shortly studied them, Plato employed them afterwards, in suspicious circumstances. in the speculative cosmology of the Today, when we say Ö2 is irrational, our Timaeus, and Euclid climaxed his Ele- language still reflects these ancient anxi- ments with the ½rst known proof that eties. only ½ve such regular polyhedra exist. Still, the Pythagorean vision, broadly Kepler was enraptured by his discov- understood–and stripped of cultic, if ery. He imagined that the spheres emit- not entirely of mystical, trappings– ted music as they rotated, and he even remained for centuries a touchstone for speculated on the tunes. (This is the pioneers of mathematical science. Those source of the phrase “music of the working within this tradition did not spheres.”) It was a beautiful realization insist on whole numbers, but continued of the Pythagorean ideal. Purely concep- to postulate that the deep structure of tual, yet sensually appealing, the zeroth the physical world could be captured in law seemed a production worthy of a purely conceptual constructions. Con- mathematically sophisticated Creator. siderations of symmetry and abstract To his great credit as an honest man geometry were allowed to supplement and–though the concept is anachronis- simple numerics. tic–as a scientist, Kepler did not wallow In the work of the German astronomer in mystic rapture, but actively strove to Johannes Kepler (1570–1630), this pro- see whether his law accurately matched gram reached a remarkable apotheosis– reality. He discovered that it does not. In
Dædalus Winter 2002 143 Note by wrestling with the precise observations -22 Frank Sun, and the size is a factor 10 small- Wilczek of Tycho Brahe, Kepler was forced to er; but the leitmotif of Bohr’s model is give up circular in favor of elliptical unmistakably “Things are Number.” orbits. He couldn’t salvage the ideas that Through Bohr’s model, Kepler’s idea ½rst inspired him. that the orbits that occur in nature are After this, the Pythagorean vision precisely those that embody a conceptu- went into a long, deep eclipse. In New- al ideal emerged from its embers, reborn ton’s classical synthesis of motion and like a phoenix, after three hundred gravitation, there is no sense in which years’ quiescence. If anything, Bohr’s structure is governed by numerical or model conforms more closely to the conceptual constructs. All is dynamics. Pythagorean ideal than Kepler’s, since Newton’s laws inform us, given the posi- its preferred orbits are de½ned by whole tions, velocities, and masses of a system numbers rather than geometric con- of gravitating bodies at one time, how structions. Einstein responded with they will move in the future. They do not great empathy and enthusiasm, referring ½x a unique size or structure for the solar to Bohr’s work as “the highest form of system. Indeed, recent discoveries of musicality in the sphere of thought.” planetary systems around distant stars Later work by Heisenberg and have revealed quite different patterns. Schrödinger, which de½ned modern The great developments of nineteenth- quantum mechanics, superseded Bohr’s century physics, epitomized in Max- model. This account of subatomic mat- well’s equations of electrodynamics, ter is less tangible than Bohr’s, but ulti- brought many new phenomena with the mately much richer. In the Heisenberg- scope of physics, but they did not alter Schrödinger theory, electrons are no this situation essentially. There is noth- longer particles moving in space, ele- ing in the equations of classical physics ments of reality that at a given time are that can ½x a de½nite scale of size, “just there and not anywhere else.” whether for planetary systems, atoms, Rather, they de½ne oscillatory, space- or anything else. The world-system of ½lling wave patterns always “here, there, classical physics is divided between ini- and everywhere.” Electron waves are tial conditions that can be assigned arbi- attracted to a positively charged nucleus trarily, and dynamical equations. In and can form localized standing wave those equations, neither whole numbers patterns around it. The mathematics nor any other purely conceptual ele- describing the vibratory patterns that ments play a distinguished role. de½ne the states of atoms in quantum Quantum mechanics changed every- mechanics is identical to that which thing. describes the resonance of musical Emblematic of the new physics, and instruments. The stable states of atoms decisive historically, was Niels Bohr’s correspond to pure tones. I think it’s fair atomic model of 1913. Though it applies to say that the musicality Einstein in a vastly different domain, Bohr’s praised in Bohr’s model is, if anything, model of the hydrogen atom bears an heightened in its progeny (though Ein- uncanny resemblance to Kepler’s system stein himself, notoriously, withheld his of planetary spheres. The binding force approval from the new quantum is electrical rather than gravitational, the mechanics). players are electrons orbiting around The big difference between nature’s protons rather than planets orbiting the instruments and those of human con-
144 Dædalus Winter 2002 struction is that her designs depend not masses or their other properties; these The world’s numerical on craftsmanship re½ned by experience, were simply taken from experiment. recipe but rather on the ruthlessly precise ap- That pragmatic approach was extremely plication of simple rules. Now if you fruitful and to this day provides the browse through a textbook on atomic working basis for practical applications quantum mechanics, or look at atomic of physics in chemistry, materials sci- vibration patterns using modern visuali- ence, and biology. But it failed to pro- zation tools, “simple” might not be the vide a theory that was in our sense sim- word that leaps to mind. But it has a pre- ple, and so it left the ultimate ambitions cise, objective meaning in this context. of a Pythagorean physics unful½lled. A theory is simpler the fewer noncon- Starting in the early 1930s, with elec- ceptual elements, which must be taken trons under control, the frontier of fun- from observation, enter into its con- damental physics moved inward, to the struction. In this sense, Kepler’s zeroth nuclei. This is not the occasion to re- law provided a simpler (as it turns out, count the complex history of the heroic too simple) theory of the solar system constructions and ingenious deductions than Newton’s, because in Newton’s that at last, after ½fty years of strenuous theory the relative sizes of planetary international effort, fully exposed the orbits must be taken from observation, secrets of this inaccessible domain. For- whereas in Kepler’s they are determined tunately, the answer is easier to describe, conceptually. and it advances and consummates our From this perspective, modern atomic theme. theory is extraordinarily simple. The The theory that governs atomic nuclei Schrödinger equation, which governs is quantum chromodynamics, or qcd. electrons in atoms, contains just two As its name hints, qcd is ½rmly based nonconceptual quantities. These are the on quantum mechanics. Its mathemati- mass of the electron and the so-called cal basis is a direct generalization of ½ne-structure constant, denoted a, that qed, incorporating a more intricate speci½es the overall strength of the elec- structure supporting enhanced symme- tromagnetic interaction. By solving this try. Metaphorically, qcd stands to qed one equation, ½nding the vibrations it as an icosahedron stands to a triangle. supports, we make a concept-world that The basic players in qcd are quarks and reproduces a tremendous wealth of real- gluons. For constructing an accurate world data, notably the accurately meas- model of ordinary matter just two kinds ured spectral lines of atoms that encode of quarks, called up and down or simply their inner structure. The marvelous u and d, need to be considered. (There theory of electrons and their interac- are four other kinds, at least, but they tions with light is called quantum elec- are highly unstable and not important trodynamics, or qed. for ordinary matter.) Protons, neutrons, In the initial modeling of atoms, the p mesons, and a vast zoo of very short- focus was on their accessible, outlying lived particles called resonances are con- parts, the electron clouds. The nuclei of structed from these building blocks. The atoms, which contain most of their mass particles and resonances observed in and all of their positive charge, were the real word match the resonant wave treated as so many tiny (but very heavy!) patterns of quarks and gluons in the black boxes, buried in the core. There concept-world of qcd, much as states was no theory for the values of nuclear of atoms match the resonant wave pat-
Dædalus Winter 2002 145 Note by terns of electrons. You can predict their and qcd cook up a concept-world of Frank Wilczek masses and properties directly by solv- mathematical objects whose behavior ing the equations. matches, with remarkable accuracy, the A peculiar feature of qcd, and a major behavior of real-world matter. These reason why it was hard to discover, is objects are vibratory wave patterns. Sta- that the quarks and gluons are never ble elements of reality–protons, atomic found in isolation, but always in com- nuclei, atoms–correspond, not just plex associations. qcd actually predicts metaphorically but with mathematical this “con½nement” property, but that’s precision, to pure tones. Kepler would be not easy to prove. pleased. Considering how much it accounts for, This tale continues in several direc- qcd is an amazingly simple theory, in tions. Given two more ingredients, New- our objective sense. Its equations con- ton’s constant gn and Fermi’s constant tain just three nonconceptual ingredi- gf, which parametrize the strength of ents: the masses of the u and d quarks gravity and of the weak interaction, a and the strong coupling constant s, respectively, we can expand our concept- analogous to the ½ne structure constant world beyond ordinary matter to de- of qed, which speci½es how powerfully scribe virtually all of astrophysics. There quarks couple to gluons. The gluons are is a brilliant series of ideas involving uni- automatically massless. ½ed ½eld theories and supersymmetry Actually even three is an overestimate. that might allow us to get by with just The quark-gluon coupling varies with ½ve ingredients. (Once you’re down to distance, so we can trade it in for a unit so few, each further reduction marks an of distance. In other words, mutant epoch.) These ideas will be tested deci- a qcds with different values of s gener- sively in coming years, especially as the ate concept-worlds that behave identi- Large Hadron Collider (lhc) at cern, cally, but use different-sized metersticks. near Geneva, swings into operation Also, the masses of the u and d quarks around 2007. turn out not to be very important, quan- On the other hand, if we attempt to do titatively. Most of the mass of strongly justice to the properties of many exotic, interacting particles is due to the pure short-lived particles discovered at high- energy of the moving quarks and gluons energy accelerators, things get much they contain, according to the converse more complicated and unsatisfactory. of Einstein’s equation, m = E/c2. The We have to add pinches of many new in- masses of the u and d quarks are much gredients to our recipe, until it may seem smaller than the masses of the protons that rather than deriving a wealth of and other particles that contain them. insight from a small investment of facts, Putting all this together, we arrive at a we are doing just the opposite. That’s the most remarkable conclusion. To the state of our knowledge of fundamental extent that we are willing to use the pro- physics today–simultaneously tri- ton itself as a meterstick, and ignore the umphant, exciting, and a mess. small corrections due to the u and d The last word I leave to Einstein: quark masses, qcd becomes a theory I would like to state a theorem which at with no nonconceptual elements whatsoever. present can not be based upon anything Let me summarize. Starting with pre- more than upon a faith in the simplicity, cisely four numerical ingredients, which i.e., intelligibility, of nature: there are no must be taken from experiment, qed
146 Dædalus Winter 2002 arbitrary constants . . . that is to say, nature is so constituted that it is possible logical- ly to lay down such strongly determined laws that within these laws only rationally completely determined constants occur David G. Nathan (not constants, therefore, whose numeri- cal value could be changed without on clinical destroying the theory). research & the future of medicine
Biomedical inquiry as it is practiced in America today is an amalgam of three different kinds of research: basic research, population research, and clini- cal research. While all three are of criti- cal importance, it is clinical research that underpins our national medical efforts. Only clinical researchers are able to apply the knowledge of the cell and organ systems developed by basic researchers, and the population data gathered by epidemiologists and biosta- tisticians, to patients, making this knowledge and data relevant to medical practice by “translating” it into novel
David G. Nathan has been a Fellow of the Ameri- can Academy since 1983. The Robert A. Strana- han Distinguished Professor of Pediatrics at Har- vard Medical School and president emeritus of the Dana-Farber Cancer Institute, he was also physician in chief at the Children’s Hospital in Boston from 1985 to 1995. A recipient of the National Medal of Science for his research on blood disorders, he is also well known for his men- torship of young physicians during the formative years of their research careers.
Dædalus Winter 2002 147