12 / CHRONICLESntalents and other qualifications in their children. Only ifnthis process were interfered with, in a systematic manner,ncould both the principles of merit and of equal life chancesnbe achieved. Perhaps a massive system of collectivizednchild-rearing could be instituted. Or perhaps a compulsorynschooling system could be devised so as to even outnhome-inspired developmental advantages and prevent familiesnfrom making any differential investments in humanncapital in their children, either through formal or informalnprocesses. In any case, achieving both merit and equal lifenchances would require a systematic sacrifice in familynautonomy.nImplementing any two of these principles precludes thenthird. While inevitable conflicts might be tolerated bynsystematic theorists in the nonideal world, these conflictsnarise under ideal conditions. Given only moderate scarcitynand strict compliance with the principles chosen, given thatnthere is no aftermath of injustice from the immediate past,nwe are applying these principles in our thought experimentnto the best conditions that could realistically be imaginednfor a modern, developed society.nOf course, liberalism has long been regarded as annamalgam of liberty and equality. And liberals and libertariansnhave long been fearful of the sacrifices in liberty thatnwould be required to achieve equality of result. Equality ofnopportunity, by contrast, has been regarded as a weaklynreformist, tame principle which avoids such disturbingnconflicts. However, even under the best conditions, it raisesnstark conflicts within the one area of liberty which touchesnmost of our lives most directiy. Once we take account of thenfamily, equal opportunity is an extraordinarily radicalnprinciple and achieving it would require sacrifices in libertynwhich most of us would regard as grossly illiberal.nThese principles are not demanding by themselves; theynare demanding in combination. Each of the trilemmanscenarios fully implementing two of the principles leads tondrastic sacrifice of the third. One reasonable, but unsystematicnresponse, is to trade off small increments of eachnwithout full realization of any.nBut this is to live without a systematic solution. Thenaspiration fueling the renewal of liberal theory has been thatnsome single solution in clear focus can be defined for idealnconditions and then policy can be organized so as tonapproach this vision gradually. But if trade-offs are inevitable,neven for ideal theory, then we have ideals withoutnan ideal, conflicting principles without a single unifyingnvision.nMy position is, first, that equal opportunity offers a primencase for this result: second, that despite the lack of ansystematic solution for ideal theory, there are significantnpolicy prescriptions which result without any necessity fornsolving the priority relations among these principles. Wencan live with moral conflict and still offer significantnprescriptions.nHaving sketched the first point, let us turn to the second,nthe issue of policy implications. There are two kinds ofnpolicy implications we can evaluate without having to solventhe problem of priority relations among these three principlesnfor ideal theory—without, in other words, employingnthe model of a single unified and coherent ideal which wenshould, as best we can, approach through partial realiza­nnntion. The first kind of policy implication may be seennwherever we can achieve a major improvement in thenrealization of one of these values without a major loss in anynof the others (or in any other new values which the proposalnimpinges upon). The second kind of policy implicationnapplies to any proposal that would impose a major loss innone of these values without any comparable gain in any ofnthe other values (or in any new values which the proposalnimpinges upon). The first kind of policy implication isnexemplified by all those things we could do, in the secondnterm of the Reagan Administration, to improve equality ofnlife chances, family autonomy, or meritocratic assignment.nWe are far from the possibility frontier in achieving anynof these values, and, in some cases, we have moved furthernaway rather than closer in recent years. We blithely toleratenthe perpetuation of an urban underclass; a whole generationnof urban youth is growing up with blighted life chances andnwith few opportunities to make it into the mainstreamneconomy, and with few policy initiatives now focused onntheir problem. Family autonomy is protected for middleclassnfamilies, but poor families have far greater difficulty innforming and maintaining themselves intact. By neglectingnjob prospects among the poor, this Administration has alsonaffected the incentives for family formation as well as thenability of poor families to provide essential prerequisites fornchild development and socialization. We are also far fromnachieving meritocratic assignment. In my view, discriminationnpersists against blacks, Hispanics, women, and othernminorities. There is no justification for tolerating jobndiscrimination on the basis of arbitrary factors, irrelevant tonthe roles in question. In other words, despite Mr. Reagan’sntalk of protecting “the family” and of creating an “opportunitynsociety,” his policies have promoted middle-classnfamilies and middle-class opportunities at the expense ofnthe disadvantaged.nThe second kind of policy implication is exemplified bynthe major quick-fix for the first set of problemsn—preferential treatment based merely on arbitrary nativencharacteristics. When it is applied in competitive meritocraticncontexts, this policy yields a major sacrifice in one ofnour three values, meritocratic assignment, without a significantngain in any of the others. The difficulty is thatnpreferential treatment, when it is based merely on arbitrarynnative characteristics, is mistargeted as a policy which couldnhave any effect on equality of life chances. It is mistargetednbecause, in competitive meritocratic contexts (e.g., admissionsnto graduate and professional schools), there are strongninstitutional pressures to accept the most qualified applicantsnwith the specified arbitrary native characteristics. Justnas family background provides disproportionate opportunitiesnto develop qualifications among advantaged whitenchildren, it does so among relatively advantaged minoritynchildren. This policy only serves to widen the gap betweennthe urban underclass and the black middle class—despitenthe fact that it is typically justified as special considerationnfor those who are from disadvantaged backgrounds.nMy objection does not apply to policies which applynpreferential treatment to those who are, actually, fromndisadvantaged backgrounds. In that case, meritocratic assignmentnis sacrificed for a gain in equal life chances.nRather, my objection applies to programs which are appliedn