COLUMN ONE : Where to Draw the Lines? : Changes in American society make it trickier to define the beneficiaries and goals of affirmative action. Thirty years ago, determining who was a minority was less complex.

Who is a minority?

Today, in an era of affirmative action, when the answer to that question can have an impact on college admissions, jobs, promotions and government contracts, just who should be counted has become an emotion-charged issue.

In this vast and polychromatic nation, the simple question of who is a minority has never had a simple answer. Definitions of what groups make up the minority--and who holds power as part of the majority--have shifted over time.

When affirmative action programs began in the mid-1960s, immigration was low, the population was nearly all white and, in most parts of the country, blacks were the only minority group--most of them with much lower incomes than the average white.

But now, the nation has changed dramatically: There are high levels of immigration, several major racial groups and an expansive minority middle class.

Those changes have heightened and complicated the affirmative action debate. The population of Asian Americans has gone from roughly 1 million in the 1960s to at least 8.5 million. The Latino population has grown from 3.5 million to roughly 23 million. Racial and ethnic groups officially counted as “minorities” make up roughly a third of the U.S. population--up from just over 10% three decades ago.

Since women are included in most such programs, “when you add it all up, about two-thirds of the American population is eligible” for one form of affirmative action or another, said Sen. Daniel Patrick Moynihan (D-N.Y.).

“That’s a lot of minority. And I think what began as an effort to redress the legitimate concerns and needs of African Americans has expanded to other things entirely unexpected.”

Even the black population has changed. Once, almost every black American was descended from American slaves. But in New York City, for example, roughly 25% of blacks are immigrants or the children of immigrants--mostly from Africa and the Caribbean, says Harvard sociology professor Mary Waters.

How did the current definitions of “minority” arise?

In the beginning, it wasn’t much of an issue and received little debate, according to historians and officials involved in the early policy-making. The focus was on the plight of black Americans. But aware of longstanding discrimination against Puerto Ricans in northeastern cities, Mexican Americans in the Southwest and Chinese and Japanese--primarily in California--lawmakers also sought to protect those groups.

In the early 1960s, federal officials established four categories for equal opportunity programs--”Negroes, Spanish-surnamed, Oriental and Indian.” With adjustments to reflect changing fashions in language, those groups remain the basic categories for most affirmative action programs.

The programs vary widely. Set-aside programs, for example, place a share of government contracts in a pool for minority- and female-owned firms or allow those firms to win contracts even when they do not submit the lowest bids. This embodies what most people would consider true “preferences.”

Other programs, such as those in most workplaces, do not require that special preference be given to specific groups. They direct companies to count employees by race and sex and to be prepared to justify why numbers differ markedly from percentages in the labor force as a whole.

Defining Questions

Critics insist this also ends up providing preferences. “The burden of explanation is so high that people say: ‘Oh, screw it, I’ll just hire by the numbers,’ ” says Harvard Law School’s Charles Fried, the Ronald Reagan Administration’s chief advocate at the Supreme Court and an adviser to Senate Judiciary Committee Chairman Orrin G. Hatch (R-Utah).

Supporters, by contrast, argue that without affirmative action, equal opportunity for minorities and women would remain an empty slogan. The people who make most decisions on hiring, firing and promotion are mostly still white men, as the Labor Department’s recent “Glass Ceiling” report showed. Unless someone keeps track of the numbers, those men will, perhaps unconsciously, continue to give preference to people who look like them, program backers argue.

Affirmative action programs, under that theory, resemble contact lenses that help correct managers’ sometimes faulty vision.

Now, amid increased controversy, policy-makers are grappling with questions of definition:

* How should the nation define the proper beneficiaries of affirmative action programs?

* Should such programs concentrate on groups that have faced discrimination in the past or look primarily at current barriers?

* How much relevance does past discrimination have when considering groups who have only recently arrived? For example, what impact does denial of basic civil rights to Chinese residents of California in the early 20th Century have on Vietnamese immigrants today?

* Does the strong evidence of continued job bias justify keeping affirmative action for those who have already moved into the middle class? Or should programs be limited to minorities who are poor? Or should the country go further and move to a new system that defines beneficiaries in terms of class, not race?

Those questions would be hard enough even if the definition of a minority were simple. But it is not. For example, the law seeks to protect Asians, but where does Asia end? Over the years, federal officials have more or less set the boundary at the western border of Pakistan, much to the dismay of groups representing immigrants from a place such as Iran.

The Role of Politics

Politics has played a role. During the Richard Nixon Administration, officials made a point of specifying that Cubans--a group that is relatively prosperous, but also heavily Republican--should be included as “Hispanics.”

States have often added to--or subtracted from--the basic federal definitions. Massachusetts decided to include its large Portuguese immigrant population. Louisiana counts “French-Acadians.”

Some groups have lobbied for inclusion. Arab American organizations have argued that the government should recognize “Middle Eastern” as a category. As usually proposed, that group would include Arabs, not Jews. But some Jewish groups, particularly the ultra-Orthodox Hasidim, have argued that they face pervasive discrimination.

At least one program run by the Department of Housing and Urban Development includes Hasidim in its eligibility definitions.

To make matters more complicated, Americans have demonstrated tremendous flexibility in defining themselves.

The number of Americans who say they are of Irish descent has exploded over the years--not because of high birth rates, but because Americans have come to see “Irish” as a desirable heritage, Waters says. As the descendants of Irish immigrants married into other groups more Americans have been able to claim Irish heritage.

In the past two decades, the census has found a huge increase in the Native American population, from just over 500,000 in the early 1960s to an estimated 2.2 million, in large part because people who had once said they were part of other groups now call themselves Native Americans.

Taken together, similar shifts can lead to redefinition of the whole concept of “minority.” In 1910, “you would have said: ‘America will be a nation of minority races by 1980,’ ” Waters notes. At the time, immigrants from places such as Italy and Eastern Europe were considered part of separate--and discriminated against--minority races, she says.

Over time, intermarriage and other forms of assimilation broke down those definitions.

In the future, the differences among the Asian, Latino and non-Latino white populations of the United States--among whom intermarriage rates are on the rise--are apt to “become like the differences among white ethnic groups now,” leaving blacks and perhaps some dark-skinned Latinos on the disadvantaged side of the color line, Waters predicts.

Other specialists doubt America can truly blend away its current racial and ethnic barriers, noting that divisions between whites and Asians or whites and Latinos run far deeper than those that separated European ethnic groups three generations ago.

Until 1952, the skeptics point out, federal law forbade Asian immigrants to apply for citizenship--a barrier far more extreme than European groups faced. And residential segregation now separates population groups more--in part because suburbs allow more distance than did the crowded urban melting pots of the 1920s and 1930s.

Past discrimination, of course, is one of the three main rationales that supporters have offered for existing affirmative action. Compensation for past discrimination, correction of current discrimination and diversification as an end worth pursuing for its own sake are often lumped together, but in fact have very different implications.

Jews, for example, or Italian Americans can easily point to historic discrimination but would have difficulty proving widespread current disadvantage. Others, such as Koreans, may suffer from bias now, but would have trouble demonstrating historic discrimination given that few arrived until recent years.

Diversity can be even more complex, says Yale Law School professor Paul Gewirtz, because “it’s not just groups, but sub-groups that can make claims.” Some colleges subdivide Latinos--counting Puerto Ricans separately from Chicanos.

Compensation was the first rationale for affirmative action and still underlies set-aside programs for contracts. The theory is that set-asides make up for past bias that has limited the number of minority-owned firms and made them less competitive now.

In the workplace, by contrast, most affirmative action rests on the other two rationales.

The distinction is important in part because many white Americans, particularly men, tell pollsters they do not believe minorities still face widespread discrimination. That feeling helps bolster the image many opponents of affirmative action focus on: Middle-class minorities benefiting from race-based programs, a doctor’s son getting a boost on his college applications, for example. The desire to eliminate that image has helped build support for the idea of converting the basis of affirmative action from racial or ethnic to economic.

Evidence of Bias

Supporters insist those images are misleading and that all minorities, not just the poor, still face discrimination.

As state Assembly Speaker Willie Brown puts it: “Racism doesn’t necessarily track with poverty. Racism is racism.”

Several types of evidence suggest that bias remains a serious issue not only for blacks, but for Latinos and Asians.

One type of evidence comes from experiments using pairs of people--one white, one non-white--who apply for jobs simultaneously. The pairs are given matching resumes and coached to eliminate as many differences as possible.

Such studies continue to show prejudice against blacks and Latinos and little “reverse” bias against whites.

A second type of evidence comes from the courts. Each year, in thousands of cases plaintiffs demonstrate they have suffered discriminated. Cases in which whites prove “reverse” discrimination remain rare.

The most powerful evidence comes from data comparing earnings across racial lines. An extensive, recently published analysis of census data examined 37 variables that might affect earnings--where people live, what type of work they do, how much education they have, their race, immigration status, and so on.

The conclusion: Black, Latino and Asian men suffer a substantial earnings gap when compared to similarly educated and experienced white men. Among men with graduate degrees, for example, being black costs a person $150 a week, every week of the year. Black men with a college education earn, all other things being equal, $3,500 a year less than similarly educated and employed white men.

A Different Pattern

Similar earnings gaps exist for Latino and Asian men. The finding for Asians might seem to contradict other figures that show many Asian American groups having among the highest per capita incomes. But Asian Americans are far more likely than most Americans to have graduated from college. That raises their average incomes. But when Asian American men are compared with similarly educated white men, a substantial earnings gap appears.

The evidence “would suggest that discrimination is very much persistently with us, and it is with us even for groups, like Asians, who we think of as doing well,” says Roderick J. Harrison of the Census Bureau. Worse yet, he adds, at least for black Americans, the earnings gap failed to narrow throughout the 1980s--reversing a four-decade pattern of slowly increasing parity.

The pattern for women is significantly different. Asian, Latina and white women have similar earnings--although all women experience an earnings gap when compared with similarly educated and experienced men.

Unlike the racial gaps, the gap between men and women has continued to narrow, but it remains significant--on the order of 15 cents on the dollar--research by Francine Blau and Lawrence Kahn of Cornell University shows.

Discrimination does not necessarily cause the entire gap. “We try to be careful and honest,” says Harrison. “There are always other factors that could account for some of the differentials.”

But just as a doctor looks for characteristic symptoms and makes a tentative diagnosis, he says, people looking at minorities and women should examine the different types of evidence as symptoms. Taking them together, “there’s probably a real illness out there.”

Where does all that leave affirmative action policy?

“There’s a temptation to say, ‘Well, if we’ve gotten to this point, maybe we’ve gone far enough,’ ” Blau says. “But we don’t know if we changed the policies that we would stay at this point. We could slip back.”

The argument against changing existing policies is strongest when applied to blacks, given that the earnings gap has not narrowed in 10 years--even for black men with college educations and middle-class jobs.

For Latinos and Asian Americans, the situation is considerably more complex because of immigration. Immigrants who have been here for at least a decade earn considerably more than those who arrived more recently, lowering average incomes for both groups. The evidence indicates those wage gaps will narrow as immigrants gain command of English and familiarity with American customs.

In addition, both categories include groups with markedly different experiences. Immigrants from Southeast Asia, for example, on average have been poorer and less well-educated than those from Japan, Korea and Taiwan. As a result, policies that address “Asians” across the board may miss the mark.

In the end, however, none of the debates can be resolved without considering the most basic question: Just what is it that the country wants its affirmative action policies to accomplish?

“Unless you clarify what your rationale is, you can’t really narrow the debate and try to find some grounds for agreement,” notes Gewirtz. “What groups you should include depends on the goal.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

About This Series

In this series The Times examines affirmative action, a policy that has left its imprint on the American workplace and college campuses over the past 30 years. With some now questioning whether the practice of giving preferences to minorities has been fair to all, this series, which will appear periodically throughout 1995, will measure its impact on a wide sector of American life--its institutions, ideas and attitudes.

* Previously: Why affirmative action became an issue in 1995, its legal underpinnings and its impact on presidential politics.

* Today: The question of who is a minority heightens and complicates the affirmative action debate, with implications for college admissions, jobs, promotions and government contracts.

* Wednesday: Beneficiaries of affirmative action say it has transformed their lives, but for many it is an opportunity that is alloyed with fear and uncertainty.

* Thursday: A Times Poll measures attitudes about affirmative action.