THE RIGHTWARD shift of the U.S. Supreme Court -- likely to result in the reversal of earlier decisions upholding affirmative action -- makes it imperative for Americans to find a new road to racial equality.

I propose that we move in two directions -- toward the reconstruction of civil society in minority communities and the promotion of broad, nonracial policies for economic opportunity and security.

By civil society, I mean the "intermediate" institutions between the state and the individual -- community associations, schools, local media, independent social agencies -- that have always been critical to community economic and political development. Because of the legacy of slavery and discrimination, black institutions have never had the capital resources of comparable, predominantly white institutions.

A renewal of minority institutions, backed by fresh capital support, would have a potentially large "multiplier" effect on social development over generations because it would represent a lasting contribution to the stock of community wealth, not just to an individual's advantage.

Yet such institution-building would have limited impact unless complementary efforts addressed wider economic and social problems. Scarce employment opportunities, eroding incomes in lower brackets and lack of access to health care hit minority groups with particular force. But rather than rely on racially targeted efforts, a post-affirmative-action strategy should emphasize broad-based, nonracial policies that make sense on their own -- for example, national health reform, financing for training and higher education, family leave and other policies benefiting families with young children.

As William Julius Wilson and others have argued, broad-based policies that promote the interests of lower- and middle-income Americans -- and that deliver substantial benefits to minorities on the basis of their economic condition -- will do more to reduce minority poverty than narrowly based, and hence poorly funded, measures for minority groups or the poor alone. These efforts can also be designed to dovetail with strengthened intermediate institutions and thereby to contribute to the overall process of civil reconstruction and renewal.

Neither minority institution-building nor nonracial social policy implies any diminished commitment to fight racial discrimination. Given the judicial realities, however, discrimination will generally have to be fought with means provided by civil rights law other than preferential hiring and similar practices adopted under the rubric of affirmative action.

Originally, affirmative action meant outreach for minority candidates for college admissions, jobs and contracts. But it has also come to mean use of racial preferences to remedy a historic pattern of discrimination either in a particular institution or in society at large, or to achieve balance or diversity in line with demographic patterns. The constitutional and political problems concern affirmative action in this broadened use.

For many, affirmative action in the extended sense has become a litmus test of loyalty to civil rights and minority interests. Yet it has never been a good test. Affirmative action -- that is, racial preference, above and beyond antidiscrimination enforcement -- has had only a modest impact on the structure of opportunity and little, if any, effect on the ghetto poor.

And while it has improved minority representation in the professions and at some corporations, universities and public agencies, these gains have not come without cost. Affirmative-action policies have insidiously helped to perpetuate racism, as whites are inclined to believe that blacks have gotten positions despite having lower scores and lesser abilities -- and accordingly expect blacks to perform less ably than whites. There is no case currently on the court calendar likely to result in a decision overturning precedents for affirmative action. The departure of Justices William Brennan and Thurgood Marshall and the addition of Clarence Thomas, however, make likely a reversal of earlier decisions upheld by narrow majorities. In the wake of such a ruling, many liberals will want Congress to undo the court's action, as recent civil rights legislation reverses several court decisions regarding job discrimination.

But should the court overturn Bakke (the 1978 decision upholding affirmative action in public institutions), it may virtually preclude any congressional reversal by ruling that racial preferences are constitutionally impermissible, except as a narrowly tailored remedy in specific cases of discrimination. On the other hand, in restricting or overturning United Steelworkers v. Weber (a key 1979 decision approving private affirmative-action plans), the court would be revising its interpretation of a civil rights statute. Congress, to be sure, could respond by giving explicit approval to private affirmative-action plans. But, with opinion polls showing overwhelming majorities against affirmative action, this seems highly unlikely.

Nonetheless, the curtailment of affirmative action may mean less than is generally expected. Much of what is conventionally attributed to affirmative action will be sustained under antidiscrimination law, including the 1991 civil rights legislation (although this may be complicated if the court unleashes a wave of reverse-discrimination lawsuits). Furthermore, affirmative action is now a firmly established practice of many university, corporate and political decision makers. Just as discrimination did not end with formal rulings against it, neither will affirmative action end with formal rulings.

Some institutions may adopt more general, nonracially defined policies promoting diversity. In university admissions, there is already a move afoot to adopt affirmative action for the socio-economically disadvantaged of all races. Blacks could benefit because they form a higher proportion of the disadvantaged. But because low-income Asians and whites now outperform low-income blacks academically, such a policy would help blacks only when they make greater educational progress at earlier ages. That is surely where the chief emphasis should be -- along with a broad, progressive program of financial support for post-secondary education and new capital for historically black institutions.

Here, as elsewhere, we need complementary strategies -- nonracial social policy and community self-development.

Strengthening communal self-development in minority communities could appeal to conservatives, because it emphasizes nongovernmental means; to minority communities, because it conveys respect and provides support for their autonomous and indigenous institutions; and to the public at large, because it favors the kinds of institutions that promise to bring order and stability to violence-torn ghettoes.

Some may worry that support for community self-development is tantamount to an acceptance of separatism. But other ethnic and religious groups have strong communal institutions -- Catholic, Protestant and Jewish philanthropies, social agencies, hospitals, newspapers and magazines -- without anyone raising the specter of separatism. Those institutions have facilitated the integration of immigrant minorities into American society. Even if some black institutions, such as Afrocentric schools, do have a separatist philosophy, the history of the black community and inescapable economic realities strongly suggest that separatism will fail to command general support from black Americans. The more practical question is this: Does white America owe black America resources for community self-development? I think it does. Slavery and its aftermath deprived blacks as a community of opportunities to accumulate wealth. Today, a wide economic gulf separates even the black middle class from the white middle class. Overall, according to a study by Melvin L. Oliver and Thomas M. Schapiro based on 1984 data, black households have only one quarter the net worth and 11 percent of the net financial assets of white households. Astonishingly, white households with annual incomes between $7,500 and $15,000 in 1984 had higher mean net worth and net financial assets than black households making $45,000 to $60,000. Not just individuals, but the black community as a whole stands to inherit relatively little wealth from one generation to the next.

Still, I am under no illusion that American voters are ready to accept -- let us use the correct term -- reparations. But private philanthropy could recognize such obligations and lay the groundwork for a new National Endowment for Black America. This fund would receive capital contributions from industry and charitable foundations and individuals and would support a variety of social and cultural organizations in the black community. (It could also aid the formation of minority businesses with programs aimed at fostering entrepreneurship.) Once in operation, the endowment might receive public as well as private funds, which it could disburse in the form of "challenge" grants, requiring additional fund-raising to leverage its own resources and to reinforce practices of communal savings and investment.

My use of the word "black" rather than "minority" here is not accidental. While strengthening intermediate institutions among Hispanics and other minorities would also be worthwhile, I believe the obligations of the United States to black Americans are historically singular. And because the problems of social isolation and decline of civil society are more acute in the black ghettoes, civil reconstruction in black America is especially urgent.

Would a National Endowment for Black America be subject to the same criticism as affirmative action? The crucial difference is that this approach provides resources to institutions rather than individuals and thereby sidesteps charges that compensatory programs undermine merit-based standards. Unlike affirmative action, it does not affect decisions about individual careers and thereby avoids generating the sense of personal grievance that affirmative action has produced among some whites. The National Endowment for Black America would be a private institution -- the private, race-specific branch of a strategy whose public branches are race-neutral.

Yet community development efforts, even with greater capital support, will fail unless combined with broader social and economic policies benefiting low- and middle-income Americans. Capital endowments provide institutions a margin of independence, often vital to developing their own long-term mission. But without strong, complementary systems of public financing for education, social services and child care -- to mention three of the most critical areas -- a strong independent sector in minority communities will not emerge.

The shift of the Supreme Court to the right is, I believe, an ominous development. But on affirmative action, it could be a blessing in disguise, forcing Americans to come up with answers to racial inequality that are more effective in liberating minorities from poverty and more successful in generating the broad coalitions needed to carry out a wider progressive reform. When the Supreme Court overturns racial preferences, those committed to racial justice should seize the occasion to mark a new beginning in the struggle.

Paul Starr, a Princeton University sociologist, is co-editor of the American Prospect, where a longer version of this article originally appeared.