Report: Racism

New Perspectives on Race and Ethnicity

Part 2

Table Of Contents

PART FOUR — EDUCATION

Desegregation and Resegregation in the Public Schools

The Racial Gap in Academic Achievement

Schools That Work for Minority Students

Preferential Admissions in Higher Education

==============================

Hoover Institution

http://www-hoover.stanford.edu/publications/books/colorline.html

Beyond the Color Line: New Perspectives on Race and Ethnicity in America

Edited by Abigail Thernstrom and Stephan Thernstrom

Hoover Institution Press Publication No. 479

2002

==============================

Abstract

From color-blind to color-consciousness—a counterproductive approach to racial equality?

The American racial and ethnic landscape has been radically transformed over the past three decades. A generation ago, blacks had much less education, much poorer jobs and were more likely to live in solidly black neighborhoods than they are today. Yet the old notion of “two societies, one black, one white—separate and unequal” still persists thirty years after it first appeared in the misguided diagnosis of the Kerner Commission report.

America’s changing racial and ethnic scene is the central theme of Beyond the Color Line. In essays covering a range of areas including education, law, religion, immigration, family structure, crime, economics, politics, and more, this volume examines where we’ve been, where we are, and where we’re going. Along the way, the authors attempt to illuminate how we have moved from Dr. Martin Luther King’s dream of all Americans being judged solely by the “content of their character, not the color of their skin” to today’s vaguely Orwellian civil rights orthodoxy—that it is necessary to treat some persons differently in order to treat them “equally.”

The product of the Citizens’ Initiative on Race and Ethnicity—formed in 1998 as an alternative to the one-sided official “dialogue” on questions of color—many of these twenty-five brief essays offer either explicit or implicit public policy recommendations. A common theme unites them—new realities require new thinking, and old civil rights strategies will not solve today’s problems. Beyond the Color Line takes the first steps toward a new civil rights agenda.

Abigail Thernstrom, a senior fellow at the Manhattan Institute and a member of the Massachusetts State Board of Education since 1995, and Stephan Thernstrom, the Winthrop Professor of History at Harvard University and Manhattan Institute senior fellow, are coauthors of America in Black and White: One Nation Indivisible (Simon & Schuster, 1997) and write frequently for a variety of journals and newspapers, including The New Republic, the Wall Street Journal, and the UCLA Law Review.

Contributors: David J. Armor, Michael Barone, Douglas J. Besharov, Clint Bolick, David Brady, Linda Chavez, William A.V. Clark, Ward Connerly, John J. DiIulio Jr., Tamar Jacoby, Everett C. Ladd, George La Noue, William J. Lawrence, Nelson Lund, Christine H. Rossell, Sally Satel, Thomas Sowell, Shelby Steele, Abigail Thernstrom, Stephan Thernstrom, Martin Trow, Reed Ueda, Eugene Volokh, Finis Welch, James Q. Wilson, C. Robert Zelnick

==============================

Foreword ix

John Raisian and Larry Mone

Contributors xi

Introduction 1

PART ONE THE BIG PICTURE

The Demography of Racial and Ethnic Groups 13

Stephan Thernstrom

Immigration and Group Relations 37

Reed Ueda

What Americans Think About Race and Ethnicity 53

Everett C. Ladd

Wrestling with Stigma 69

Shelby Steele

PART TWO PRIVATE LIVES AND PUBLIC POLICIES

Residential Segregation Trends 83

William A. V. Clark

African American Marriage Patterns 95

Douglas J. Besharov and Andrew West

Crime 115

James Q. Wilson

Health and Medical Care 127

Sally Satel

Supporting Black Churches 153

John J. DiIulio Jr.

PART THREE ECONOMICS

Discrimination, Economics, and Culture 167

Thomas Sowell

Half Full or Half Empty? The Changing Economic Status of African Americans, 1967–1996 181

Finis Welch

Discrimination in Public Contracting 201

George R. La Noue

PART FOUR EDUCATION

Desegregation and Resegregation in the Public Schools 219

David J. Armor and Christine H. Rossell

The Racial Gap in Academic Achievement 259

Abigail Thernstrom

Schools That Work for Minority Students 277

Clint Bolick

Preferential Admissions in Higher Education 293

Martin Trow

PART FIVE LAW

Racial and Ethnic Classifications in American Law 309

Eugene Volokh

Illusions of Antidiscrimination Law 319

Nelson Lund

PART SIX POLITICS

Race, Ethnicity, and Politics in American History 343

Michael Barone

The Politics of Racial Preferences 359

David Brady

From Protest to Politics: Still an Issue for Black Leadership 369

Tamar Jacoby

PART SEVEN ONE NATION, INDIVISIBLE

The New Politics of Hispanic Assimilation 383

Linda Chavez

In Defense of Indian Rights 391

William J. Lawrence

The Battle for Color-Blind Public Policy 405

C. Robert Zelnick

One Nation, Indivisible 415

Ward Connerly

Index 425

==============================

PART FOUR — EDUCATION

Desegregation and Resegregation in the Public Schools

DAVID J. ARMOR and CHRISTINE H. ROSSELL

When the Supreme Court declared the end of official (de jure) segregation in Brown v. Board of Education in 1954, the public schools became the center stage for the struggle to promote racial integration and equity in America. Most of us born by the beginning of World War II will never forget the graphic images of black children in Little Rock, Arkansas, being escorted into school buildings by soldiers, surrounded by crowds of jeering white adults. About a decade later, we saw similar crowds of white adults shouting epithets, throwing stones, and burning buses when school desegregation moved to the North in such cities as Pontiac, Michigan, and Boston, Massachusetts. Unlike other social policies, vehement public protests did little to deter the school desegregation movement because it was being advanced and enforced by the (almost) politically immune federal courts.

From the mid-1960s to the late 1970s a vast transformation took place in American public schools as federal courts and government agencies demanded race-conscious policies in every facet of school operations. The most controversial aspect of school desegregation during this period involved the rules for assigning students to schools, when racial balance quotas were adopted instead of neighborhood or other geographic rules. In larger school districts these racial quotas required mandatory busing, whereby students were transported long distances from their former school to different schools across a city or county in order to attain racial balance.

But school desegregation court orders went far beyond student assignment, with requirements for racial quotas in hiring, racial balance in the assignment of faculty and staff, and racial equity in facilities (resources), transportation, and extracurricular activities. These six desegregation plan components—student assignment, faculty, staff, facilities, transportation, and extracurricular activities—became known as the Green factors. All school systems under court order had to show they had complied with each of them before they could be declared unitary (nondiscriminating) systems and released from court orders.1

There has been much debate about whether school desegregation should be judged a success or a failure, not just in the attainment of school racial balance but also with respect to other social and educational goals such as improved race relations and academic performance of minority children. So far as racial balance is concerned, initially the most important objective of desegregation plans, there is general agreement that substantial improvement occurred during the early 1970s. But some critics, especially Gary Orfield and his colleagues on the Harvard Project on School Desegregation, have asserted that resegregation began occurring in the late 1980s and worsened in the early 1990s, particularly as federal courts began declaring school districts unitary and ending court supervision.2

Although early Supreme Court school decisions did not address social and educational outcomes, there is little question that educators and civil rights activists viewed racial balance as merely a means to an end. According to these views, the ultimate goal of school desegregation was to reduce racial prejudice and improve the academic achievement of African American children; schools were to be the pathway to full economic and social parity with whites.

Assessing the extent to which school desegregation has achieved these broader goals, sometimes called extra-Green factors by the courts, is much more complicated than assessing compliance with the six Green factors. In particular, we have to assess a myriad of social and educational effects of desegregation, and to be complete we have to compare these effects with the costs of desegregation—monetary expenditures, political controversy, white “flight,” and loss of local control. It is by no means obvious to the average citizen that school integration, and especially the more intrusive practice of mandatory busing, has any benefits at all, much less benefits that justify the costs. Indeed, many Americans believe that mandatory reassignment or “forced busing” has reduced the quality of education in school districts where it has been implemented.3

This essay will summarize the successes and failures of school desegregation with regard to these issues. First, we assess the impact of desegregation policies on actual racial balance in the public schools. The evidence indicates that school desegregation has created substantial racial balance in our public education systems. Second, we address the issue of resegregation to determine whether the racial balance established in the 1970s by school desegregation plans has been reversed by the unitary status findings of the 1980s and 1990s. We shall show that at least as late as 1995 racial balance trends are not reversing and that the changes in racial and ethnic isolation discussed by Orfield are in fact caused by long-term demographic trends of declining white and increasing minority enrollment, not the dismantling of desegregation plans.

Indeed, as James Coleman first found, the mandatory reassignment plans of the 1970s exacerbated these long-term demographic trends by accelerating the decline in white enrollment, thereby limiting the extent of actual integration in the school districts in which they were implemented.4 This effect was greatest in our largest school districts. Finally, we evaluate evidence on the social and educational effects of desegregation, and especially academic achievement. We argue that in this area more than any other, school desegregation has failed to deliver on its promises, in spite of the early optimism of many social scientists and civil rights activists.

Racial Balance

Prior to Brown, most public schools in the South were onerace schools, either white or black. Ten years after Brown, one study estimated that 99 percent of black children in the South were in one-race schools. The first nationwide study of school segregation was ordered by the Civil Rights Act of 1964 and carried out by James Coleman and his colleagues during the 1965–66 school year.5 The Coleman report estimated that nationally 65 percent of all black students attended schools that were over 90 percent black, while 80 percent of all white students attended schools that were over 90 percent white.

PRE-SWANN PROGRESS

The extent of racial isolation in the South was far greater than in the North, mainly because of de jure segregation in the South. Table 1 shows the percentage of elementary black and white students in schools over 90 percent black or white, respectively, for twenty-two of the largest Southern school districts in 1965 or 1968. Although it is fair to say that some racial integration had taken place in these Southern cities ten years after Brown, it was clearly nominal for black students, with the notable exceptions of Kansas City, Nashville, and Dallas. Indeed, it was precisely this token progress that led to the Green decision, which called for the elimination of segregated schools “root and branch.”6

Table 1 Percentage of Black and White Elementary Students in Schools over 90 Percent Black or White, Southern Cities in 1965–1966 (except as noted)

*School district*	*Blacks in black schools*	*Whites in white schools*
Birmingham, Ala.a	99	99
Mobile, Ala.	100	100
Little Rock, Ark.	96	97
Miami-Dade, Fla.	91	95
Jacksonville-Duval, Fla.a	92	92
Tampa-Hillsboro, Fla.a	91	91
Atlanta, Ga.	97	95
East Baton Rouge Parish, La.a	95	95
New Orleans Parish, La.	96	84
Kansas City, Mo.	69	65
St. Louis, Mo.	91	66
Jackson, Miss.a	99	100
Charlotte-Mecklenberg, N.C.	96	95
Oklahoma City, Okla.	90	96
Tulsa, Okla.	91	99
Charleston County, S.C.a	99	99
Memphis, Tenn.	95	94
Nashville, Tenn.	82	91
Dallas, Tex.	83	90
Houston, Tex.	93	97
Norfolk, Va.a	90	90
Richmond, Va.	98	95

Source: U.S. Commission on Civil Rights, Racial Isolation in the Public Schools (Washington, D.C.: U.S. Government Printing Office, 1967), pp. 4–5, except as noted.

a Computed by the authors from the 1968 Office for Civil Rights enrollment data.

Racial imbalance also existed in Northern cities during the mid-1960s, but racial isolation was not nearly so extensive. Most Northern school segregation at that time was thought to be de facto, that is, brought about by the private decisions of citizens to live in different geographic areas. The highest levels of racial isolation existed in Chicago, Cleveland, Detroit, Indianapolis, Milwaukee, and Philadelphia where the black population swelled from post–WorldWar II migration of Southern blacks looking for jobs in the large urban centers of the North. This migration overwhelmed the capacity of white neighborhoods to absorb blacks and still remain integrated, although none but Chicago approached school racial isolation rates of 90 percent. Other large cities with sizable, but in some cases smaller, black enrollments such as Boston, Cincinnati, Columbus (Ohio), Los Angeles, Newark, New York, and San Francisco had no more than half of their black students in predominately black schools.

POST-SWANN PROGRESS

The situation in the South changed dramatically in the 1970s. In the 1971 Swann decision for Charlotte-Mecklenburg, North Carolina, the Supreme Court endorsed strict racial balance quotas for all schools in a system and approved cross-district mandatory busing to attain complete racial balance.7 In effect, the Supreme Court abandoned geographic school assignment (i.e., being assigned to the closest school) for Southern school systems unless it resulted in racially balanced schools, which was impossible in most larger school districts because of segregated housing patterns.

Court-ordered school desegregation moved to the North only two years later with the Supreme Court’s Keyes decision for Denver, Colorado.8 We shall not go into the complicated legal basis for Northern desegregation orders, almost none of which involved state-enforced segregation. Suffice it to say that despite de facto segregation in the North, systemwide racial balance remedies and mandatory busing plans were ordered for many Northern districts after Keyes. In addition, the Department of Health, Education, and Welfare (HEW) was active in pressuring school systems to implement desegregation plans, and several school systems adopted mandatory reassignment plans under this pressure.

As a result of local civil rights pressure, many other school systems adopted voluntary transfer plans that involved M to M (majority to minority) transfers where any student could transfer from a school in which his or her race was in a majority to a school in which his or her race was in a minority. Another common local initiative was to close some predominantly minority schools and reassign the students to predominantly white schools; this was carried out in such cities as Riverside, California, and Evanston, Illinois. Although these local measures did not involve mandatory busing of whites, they nonetheless accomplished some degree of school integration.

Table 2 Prevalence of Formal School Desegregation Plans, 1990

Size of district	Percent with past or present desegregation plans	Percentage share of black/Hispanic students	Total districts in sample
Very large (N>27,000)	72	53	145
Large (N>10,000>27,000)	39	22	421
Medium (N>5,000>10,000)	34	13	770
Small (N>5,000)	11	12	4,012

Source: David J. Armor, Forced Justice: School Desegregation and the Law (New York: Oxford University Press, 1995), table 4.1.

The first reliable data for assessing the impact of school desegregation on racial balance was collected in 1968 by the Office for Civil Rights in HEW (OCR).9 The survey consisted of enrollment data by individual schools and by five racial-ethnic categories (white, black, Hispanic, Asian, and American Indian) in a sample of school districts. There was no survey in 1969, but from 1970 through 1974 OCR collected data annually. Beginning in 1974, the survey was conducted every other year and included all districts with court-ordered desegregation plans. The sampling scheme used by OCR varied from year to year, and thus after 1974 the OCR data do not constitute a representative sample of school districts. In fact, in some years important school districts are simply missing.

In 1990 the authors participated in a national survey of school desegregation and magnet schools sponsored by the U.S. Department of Education, for which a statistically representative sample of 600 school districts was drawn. The original sample included all 150 largest school districts in the country, those with enrollments over 27,000 students, and smaller percentages of the large, medium, and small districts in the U.S., selected randomly from their size category.10 In this essay we use this national sample to assess racial balance trends, relying on OCR enrollment data from Fall 1968 through Fall 1987 and Common Core of Data (CCD) enrollments from Fall 1989 through Fall 1995.

The 1990 survey gathered information about whether school districts had adopted formal desegregation plans and what kinds of desegregation techniques were used. Table 2 shows the percentage of school districts that had past or current desegregation plans by size of district. The survey estimated from this random sample that nationally nearly 1,000 school districts had some type of formal desegregation plan, and the prevalence of formal plans increased with the size of the school district. Of the 145 “very large” school districts that responded to the survey, 102—more than 70 percent—had a formal desegregation plan at some point in time. We note that the largest districts enroll about half of all black and Hispanic students, while the smallest districts enroll only about one-tenth of these minority groups. Thus, black and Hispanic students are more likely to be found in larger districts, which are also more likely to have school desegregation plans.

Of the 28 percent of very large districts that did not adopt a formal desegregation plan, most were predominately white districts in the early 1970s, and indeed many remained predominately white until at least the early 1990s. Examples include Anoka County, Minnesota; Fairfax County, Virginia; Gwinnett County, Florida; and Spokane,Washington. These districts were over 90 percent white in 1972 and remained over 70 percent white until at least 1991. Obviously, there is less need for a formal desegregation plan when there are few minority students.

How effective have these plans been in achieving racial balance? To answer this question, we shall use an index that summarizes the degree of racial balance in a school system. Racial balance is defined as the degree to which each school’s racial composition matches the districtwide racial composition for a given race. The index of racial imbalance, also called the dissimilarity index, ranges from 0 to 1, where 1 indicates total segregation (all schools are one race), and 0 means perfect racial balance (every school has exactly the same racial composition as the total district).11 Intermediate values represent the proportion of students of one race who would have to be reassigned, if no students of another race were reassigned, in order to attain perfect racial balance.12 The index can be computed for any two racial or ethnic groups; for example, whites and nonwhites, blacks and whites, or Hispanics and whites.

Fig. 1. Trends in black-white school imbalance in medium, large, and very large districts.

BLACK TRENDS

Figure 1 shows the trends in black-white racial imbalance from 1968 to 1995, separated by size of school district.13 All three size categories show significant declines in black-white imbalance (or increased desegregation), with the sharpest drops from 1968 to 1972 corresponding to the widespread implementation of desegregation plans in the South. For very large districts, racial imbalance continues to decline until 1982 and remains stable thereafter; there is a slight upturn of just 1 point between 1991 and 1995.

Medium-sized districts show slight improvements in racial balance until 1991. For the nation as a whole, then, contrary to Orfield’s claims, there is no evidence of significant resegregation in terms of increasing racial imbalance as late as 1995.

Fig. 2. Trends in black-white imbalance by plan status, very large districts.

There was, however, substantial variation by region in the timing and scope of desegregation plans, with Southern districts being the first to desegregate with comprehensive desegregation plans involving white reassignments. After Swann, most Southern districts that still had substantial racial imbalance were immediately back in court and typically ordered to adopt busing remedies along the lines of the Charlotte-Mecklenburg plan. As a result, comprehensive mandatory desegregation plans were in place for most Southern districts by 1972.

Northern mandatory busing remedies did not become commonplace until after the 1973 Keyes decision in Denver, Colorado. Therefore, implementation of comprehensive desegregation plans in the North tended to be distributed more evenly throughout the 1970s. In addition, while most Southern districts had to adopt formal desegregation plans, there were fewer formal plans in Northern districts primarily because there was no history of de jure segregation and ongoing litigation associated with dismantling it. Thus, fewer lawsuits were filed in the North, and occasionally a lawsuit was dismissed because the courts found only de facto segregation (e.g., Cincinnati). These different histories, not surprisingly, produced different patterns of racial balance trends.

Figure 2 shows the trends in black-white racial balance for very large districts—those that have received most of the attention by the courts and by school desegregation analysts—separated according to region and formal plan status. We do not show the small number of Southern districts that said they did not have formal plans.14 The trends confirm the differences in desegregation timing in the North and South. As expected, the most dramatic improvements in racial balance occurred for Southern districts as they rapidly implemented mandatory reassignment plans in the early 1970s. For Southern districts with plans, the index of racial imbalance dropped by 40 points in the four years 1968–1972, from an average of 0.81 to an average of 0.41.

The fact that the average index for Southern school districts was 0.81 in 1968 means that the dismantling of the de jure system of segregation had started before 1968; had it not, the index would have been 1.0. Freedom-of-choice plans became popular following enactment of the U.S. Civil Rights Act of 1964, and they were the primary means of desegregation for many Southern school systems until the policy was ruled insufficient by the U.S. Supreme Court in the Green decision. Freedom-of-choice plans increased the number of black students attending former white schools, but not vice versa, and they are the major reason that the index was 0.80 in 1968 rather than 1.0.

The pattern for Northern school systems with plans is quite different from that of Southern systems with plans, in that the decline in imbalance is less rapid and is spread evenly throughout the 1970s. The index dropped 36 points during the decade, from 0.70 in 1968 to 0.44 in 1980, making them nearly as balanced as Southern districts. Interestingly, Northern districts without formal plans also experienced desegregation during this period, with the imbalance index declining from 0.62 to 0.47. This trend is explained by the fact that many school systems adopted desegregation practices during this time, such as closing older imbalanced schools, building new schools in easier-to-integrate locations, and paying closer attention to attendance zone changes, but did not adopt a formal plan.

After 1980, racial balance trends level off for Southern districts but continue to improve slightly for Northern districts, and in 1991 all three groups show the same degree of racial imbalance (0.42). Most important, none of the categories of school districts shows any dramatic worsening of racial balance between 1991 and 1995, with gains of only 1 or 2 points in the imbalance index for Southern and Northern districts with plans, respectively. Once again, these data from the largest 145 school systems in the nation contradict Orfield’s argument that school districts are resegregating as a result of the dismantling of desegregation plans.

HISPANIC TRENDS

In recent years the Orfield reports have expressed concern about the increasing segregation and isolation of Hispanic students. This problem did not exist in the era of de jure segregation because most early court decisions in the South did not identify Hispanic students as a minority group that was a victim of discrimination. Indeed, most Southern school districts classified students as black and nonblack, and Hispanic students were included in the nonblack category along with white students.

After the 1973 Keyes decision in Denver, however, where both black and Hispanic students were found to be victims of discrimination, it became commonplace to treat all minority (nonwhite) students as a group in student assignments to achieve racial balance. Interestingly, in a few school districts, such as Yonkers, New York, the federal court found that only black and Hispanic students were victims of discrimination, thereby combining Asian students with white students for the purposes of desegregation. The court agreed with plaintiffs’ argument that Asian students were not a disadvantaged minority group, and therefore the Asian minority could be used to desegregate either black or Hispanic minority students! This same definition is being used in a recent desegregation plan in Rockford, Illinois. In the vast majority of school districts, however, Asians are classified as racial minorities and assigned accordingly.

Fig. 3. Trends in Hispanic-white school imbalance.

Figure 3 shows the racial balance trends for Hispanic enrollment in relation to white enrollment.15 Before 1970, Hispanics were clearly much less segregated than black students in all size categories, which may explain in part why Hispanics were not treated as victims of discrimination in early court decisions. In 1968 the index is only 0.55 for Hispanics in very large districts, compared with a value of 0.71 for black students. In large school districts the Hispanic-white imbalance is 10 points less than the blackwhite imbalance. The lower imbalance rates are explained by the greater residential integration of Hispanics and whites during the 1960s, which is due in part to their smaller population size.

After 1972, Hispanic-white imbalance declined steadily until trends begin to level off in 1987, although there are still small declines as late as 1995. In very large districts, where most black and Hispanic students are found, Hispanic imbalance is less than black imbalance during the entire time period, and by 1995 Hispanic imbalance had fallen to 0.39, compared with 0.43 for blacks. In large districts Hispanic students were more imbalanced than blacks between 1972 and 1980 but caught up with them by 1995. Hispanic imbalance has remained somewhat higher in medium-sized districts during the 1990s, but we note that index levels of about 0.30 reflect a relatively high degree of balance overall. That is, an index of 0.30 means that in a district with half Hispanic and half white students only 15 percent of Hispanics and 15 percent of whites would have to be reassigned to attain perfect balance.

These data show quite clearly that Hispanic-white imbalance has not been increasing from the mid-1980s to the present time; in fact, it has continued to decrease slightly during the 1990s and remains lower than black imbalance in very large districts. Thus, contrary to the claims of Orfield, in our national representative sample of school systems there is no evidence of increasing segregation of Hispanic students in terms of racial imbalance, the primary yardstick over the last four and a half decades for measuring the success of desegregation plans.

Segregation, Desegregation, and Resegregation

Having said that there is no evidence of resegregation in terms of racial balance, we must acknowledge that there are different ways of looking at this issue and different ways of defining these terms. The original legal meaning of segregation, as defined by the Supreme Court in 1954 in Brown, was the separation of the races by official state action (de jure segregation). At that time, desegregation was simply the abolition of state laws and government practices that enforced these laws. But the elimination of state laws requiring the separation of the races did not change segregated residential patterns, nor did it prevent a variety of other strategies adopted by some Southern states and school districts for avoiding meaningful integration.

The Supreme Court decision in Green (1968) put Southern school districts on notice that they must not merely stop discriminating but also must actually achieve desegregated schools. It was not until the Swann decision (1971), however, that a desegregated school was defined as one whose racial composition is roughly the same as the racial composition of the entire school system; that is, desegregation equals racial balance. This definition quickly became the standard throughout the nation for a desegregated school system, although the amount that a school could deviate from perfect balance varied from case to case.

Using this racial balance definition, there is no evidence of significant resegregation in our nationally representative sample of American public school systems. But the problem with a racial balance standard is that it ignores the total number or proportion of white students in a school system. A racial balance standard cares only that each school mirrors the school district’s racial composition, not what the actual racial composition might be. A school system that is 90 percent black and 10 percent white would be perfectly balanced if every school had 10 percent white enrollment. Yet such a school system would not be considered desegregated by most interested parties. Many courts have defined a 90 percent black school as racially isolated, regardless of the systemwide composition.

Moreover, a desegregation plan can cause white flight and thus be the cause of a school system’s being only 10 percent white. If a district starts out with a student enrollment that is half black and half white but becomes only 10 percent white several years later, most observers would not call the plan successful even if each school was highly balanced at or near 10 percent. Thus, a racial balance standard by itself gives us only a partial view of the amount of school desegregation that exists in a school system.

In order to overcome the limitation of racial balance definitions, James S. Coleman created a second definition of desegregation that takes into account the absolute proportion of white students in schools attended by black students (or any other minority group).16 This definition is measured by a second summary statistic called the index of interracial exposure— that is, the percentage of white in the average black (or minority) child’s school.17 In this case a value of 0 means total black-white segregation, or no whites in schools attended by black students. The maximum value of this index is the proportion of students who are white in the school system. If every school’s percentage of white exactly matches the school system’s percentage of white, the index will be the same as the proportion (or percentage) of white in the school system.

For example, if a system is 30 percent white, then an index of 0.30 would mean that every school was 30 percent white (and therefore perfectly balanced), and an index of 0.25 would indicate substantial degree of desegregation relative to the available whites. But in absolute terms an index of 0.25 means that the average school is 75 percent black, which indicates a relatively low level of desegregation in absolute terms. Thus, the interracial exposure index reflects both the extent of racial balance in the school system and the absolute level of contact, that is, the percentage of white in the average black (or minority) child’s school. When the exposure index is examined over time and is compared with the percentage of white in a system, one can obtain a more comprehensive picture of both the relative and absolute levels of desegregation.

Orfield’s studies have used recent declines in the exposure index, rather than the imbalance index, to argue that resegregation is occurring, a phenomenon he attributes to the dismantling of desegregation plans. But there are two major causes for a declining exposure index, one of which is simply a decrease in the percentage of white in a school system. The decline in the percentage of white enrollment can be due to nondesegregation-related demographic changes that have nothing to do with racial balance, such as high black or Hispanic in-migration, declining white birthrates, or normal middle-class white suburbanization. It can also be caused by “white flight,” where whites leave a school system to avoid mandatory busing. The other major cause of a declining exposure index, the one that concerns Orfield, is a reduction in the number of racially balanced schools, as might occur when a school district dismantles a desegregation plan and returns to neighborhood schools.

If the exposure index has been changing, is it due to declines in the percentage of white or is it due to the dismantling of desegregation plans? This question can only be answered by comparing the exposure index to trends in racial imbalance and also to trends in the percentage of white. We have already demonstrated in Figures 1–3 that racial imbalance has not changed significantly; it remains to examine trends in the exposure index and in the percentage of white.

Fig. 4. Trends in racial composition, very large districts.

Figure 4 shows the national trends in racial composition for very large school systems, including the average percentage of white that is the maximum for the exposure index. In 1968 very large public school districts averaged 71 percent white, 22 percent black, 6 percent Hispanic, and 1 percent Asian. For the next twenty-five years the percentage of white declined steadily, while the percentage of black, Hispanic, and Asian rose correspondingly, and by 1995 the percentage of white enrollment had fallen to only 48 percent—less than half for the first time. The decline in the percentage of white was somewhat steeper during the mid-1970s, undoubtedly influenced by white flight from desegregation. In the meantime, the percentage of black, Hispanic, and Asian in the public schools increased to 31, 16, and 5, respectively, as a result of demographic forces, including both in-migration and birthrates. Any interpretation of trends in school desegregation indexes, in particular trends in interracial exposure, must take into account these basic demographic patterns because they limit the amount of interracial exposure that can be achieved.

Fig. 5. Black-white exposure in very large districts with formal desegregation plans.

Figure 5 shows why the declining percentage of white cannot be ignored in an analysis of interracial exposure. It shows the trends in black-white exposure by region for very large districts with formal desegregation plans, juxtaposed with the trend for percent of white enrollment. For Southern districts, the largest changes in interracial exposure (solid squares) occurred between 1968 and 1972, when the index rose from 0.20 to 0.53, an improvement of 33 points in the average percentage of white students in schools attended by blacks. After 1972, however, the index began a long, steady decline, falling to 0.42 by 1995. This reduction in the absolute exposure index is not caused by increases in racial imbalance, which remained nearly flat for the next twenty years (see Fig. 2), but rather the steady decline in the percentage of white enrollment (open squares), which dropped from 68 percent in 1972 to 53 percent in 1995. In fact, the exposure index for Southern districts fell at a slower rate than the percentage of white, indicating that interracial exposure was actually increasing slightly relative to the available white enrollment. The exposure index is 15 points from its maximum in 1972 (68 minus 53), and it is only 11 points from its maximum in 1995 (53 minus 42).

A similar pattern exists for very large Northern districts with plans. As was the case with racial balance, the exposure index (solid diamonds) increased over a longer period, improving from 0.32 in 1968 to 0.41 in 1980, a gain of 9 points in the average percentage of white in schools attended by blacks. The index does not even reach 0.50, however, before it begins a long, inexorable decline over the next fifteen years, falling to only 0.29 by 1995. As in the South, this decline is the result not of increasing racial imbalance, which is nearly flat during this time, but rather of a steeply falling percentage of white enrollment: this fell from 50 in 1980 to only 35 in 1995. Again, relative to the available whites, interracial exposure actually improves slightly; it is 9 percentage points from its maximum in 1980 and is only 6 points from its maximum in 1995.

Fig. 6. Hispanic-white exposure for very large districts.

Finally, Figure 6 shows the trends in the exposure index and the percentage of white for Hispanic students in very large districts with Hispanic enrollment greater than 1 percent.18 As Orfield has noted, the exposure of Hispanics to white students has declined substantially from its high in 1972 (0.59) to its low in 1995 (0.45). But the decline is not due to increasing imbalance, because Figure 3 shows that Hispanic-white imbalance declined during this time for very large districts. Again, the decline is caused by the declining percent of white enrollment in these districts, which has fallen from 64 in 1972 to only 47 in 1995. Relative to the available whites, then, Hispanic-white exposure has actually increased, and in 1995 it is only 2 points from its maximum value. In other words, Hispanic students in very large districts are actually more desegregated than black students, both in absolute terms and relative to the available white enrollment.

By comparing the trends in racial composition with the trends in racial balance and interracial exposure, a much clearer picture emerges about how and why desegregation levels have changed in recent years. For the fifteen-year period between 1980 and 1995 (1972–1995 for Hispanics), three patterns emerge: the percentage of white has declined, racial balance has remained relatively constant, and interracial exposure has declined. Thus, the cause of declining interracial exposure is the overall decline in the percentage of white, rather than the dismantling of desegregation plans.

It may be appropriate to say that resegregation is occurring for both black and Hispanic students, in that they find themselves attending schools with a dwindling number of white students, particularly in larger Northern districts. But we must be clear about the cause of that resegregation: if ending desegregation plans were a significant factor, we would see significant increases in racial imbalance. Because this is not happening, and because the percentage of white enrollment is decreasing,we must conclude that the resegregation is due primarily to demographic change in the form of falling white enrollments and increasing minority enrollments.

Social and Educational Results

The existence of long-term national school enrollment data by race makes it a good deal easier to determine the effect of school desegregation plans on school racial balance and interracial exposure than to determine their effect on social and academic outcomes. Although there is a great amount of data on academic achievement, either from national efforts like the National Assessment of Educational Progress (NAEP) or from local school districts that administer standardized achievement tests, it is very difficult to sort out the impact of desegregation from the effects of many other factors that influence academic achievement, especially family socioeconomic characteristics, because there are so few data on these other factors. This has led to substantial disagreement among researchers about the academic effects of desegregation.

It is even more difficult to assess the effect of school desegregation on social and psychological outcomes such as race relations, racial prejudice, and self-esteem because we lack standardized measures for these variables. Consequently, with the exception of opinion polls on racial issues, there are few national data on social outcomes, and there is no way to relate that data to desegregated schools. Accordingly, we must rely on case studies carried out in individual school districts to assess the effect of desegregation on such outcomes as student interracial attitudes and behaviors. Fortunately, there are reviews of this research literature that are helpful.

ACADEMIC ACHIEVEMENT

The debate over desegregation and achievement has continued unabated since the early 1970s, when the first evidence appeared that desegregation was not improving black achievement.19 In spite of the claims and expectations of many supporters of desegregation during the 1960s, and in spite of the existence of comprehensive and well-funded desegregation plans in many school districts throughout the nation, there is not a single example in the published literature of a comprehensive racial balance plan that has improved black achievement or that has reduced the black-white achievement gap significantly.

Significantly, the most recent social science study of school desegregation— Orfield and Eaton’s attack on the courts for “dismantling desegregation”— cited only a single comprehensive desegregation plan that has led to significant minority achievement gains or a reduction of the achievement gap.20 Indeed, out of seven major desegregation plans analyzed as case studies in this book, the authors did not even discuss minority achievement in three widely discussed cases—Charlotte-Mecklenburg, Detroit, and Little Rock.

Further, they were largely skeptical of local school staff reports claiming significant achievement gains in three other case studies—Kansas City, Missouri, and two Maryland counties, Montgomery and Prince Georges— but they did not conduct any original analyses to support their argument. Finally, the only case they cite where desegregation allegedly improved black achievement was Norfolk, Virginia, where the data they used were from a study by David Armor that concluded just the opposite!21 We shall remedy these omissions by presenting new and independent analysis of achievement results for Charlotte, Kansas City, and Prince Georges County.

There are a number of major reviews of the research literature on this question, and though they do not agree precisely on what the research says, a general picture does emerge. If there is any effect of desegregation on academic achievement, it is highly variable (sometimes it occurs and sometimes it does not), the effects are modest at best, and positive effects occur for reading but not for math. These are the conclusions of three major reviews of research, including one by Nancy St. John in 1975, one by Thomas Cook et al. in 1984, and one by Janet Schofield in 1994.22

The Cook review, synthesizing the separate reviews of a panel of six experts convened by the National Institute of Education, captured the essence of this equivocal state of affairs: “On the average, desegregation did not cause an increase in achievement in mathematics. Desegregation increased [black] mean reading levels . . . between two and six weeks. . . . Little confidence should be placed in any of the mean results presented earlier . . . [because] I find the variability in effect sizes more striking and less well understood than any measure of central tendency.”23

Moreover, we now have massive amounts of national achievement data from the NAEP project, which also includes information on student background and school characteristics.How do the NAEP data inform us about the performance of minority students in desegregated versus segregated schools? What have we learned about the academic achievement of minority students who have spent most of their education in desegregated schools? Interestingly, officialNAEPreports do not address these important questions. Indeed, it is remarkable, given all the national data, the controversy, and local experience, that there are few published studies that have tried to answer these questions using NAEP or case study data.

At the outset of this discussion, it is important to distinguish two different types of processes that would cause desegregation plans to have an effect on achievement. The first, and the one assumed by most studies of desegregation and achievement, is that the major causal mechanism is the change in the racial composition of schools; that is, by the improved racial balance that occurs with desegregation. Under this causal assumption, racial isolation is harmful because it deprives minority students of contact with more middle-class and (usually) higher-achieving white students, who help set the pace of study and the standards of achievement. Racial isolation might also lead to a concentration of less effective teachers in minority schools, if more effective teachers gravitate to more integrated or predominately white schools.

The second possible causal mechanism by which school desegregation might have an effect on achievement is when a school board makes significant changes in the types and distribution of programs and resources among schools as part of the desegregation plan. These effects might be most pronounced during the early years of desegregation, if a district maintained inferior programs and resources in predominantly black schools at the time of desegregation (a circumstance not generally found by the 1965 Coleman study of equal opportunity).24 In this case improvement in minority achievement should take place regardless of racial balance.

A corollary to this second effect is a condition where desegregation improves programs and resources for all schools, in which case we might see improvement for all students regardless of race or racial balance. If the first effect is true, minority achievement should be higher or the gap narrower whenever racial composition improves, regardless of programs and resources, and there should be no achievement gains for segregated minority students or for white students. If the second effect is true, minority achievement could improve in desegregated or segregated schools, and white achievement might also improve. Of course, factors other than desegregation and program changes can improve academic achievement; the leading nonschool factor would be improved socioeconomic status. The impact of socioeconomic status on academic achievement is well documented in social science research, and therefore this possibility must be considered whenever achievement differences or achievement gains are studied.

NAEP STUDIES

The NAEP program has documented a significant improvement in black achievement (but not white achievement) and therefore a closing of the achievement gap between 1970 and 1990. In 1970 the gap in reading was just over one standard deviation for three different age groups, and by 1990 it had declined to between 0.7 and 0.8 standard deviations.25 Similar patterns were observed for math achievement. Because this improvement corresponds to a period of extensive desegregation, some have suggested that desegregation was the primary cause.

In 1998 a Rand team carried out one of the few studies to use NAEP data to examine the possible effect of desegregation on black achievement gains. They concluded that desegregation might explain part of the improvement. 26 Unfortunately, their complicated methodology utilized aggregate data on several regions of the country, and they did not analyze achievement trends for black students who actually attended segregated or desegregated schools. Therefore, from their methodology it is impossible to determine whether potential effects are due to improved racial balance or to changes in programs that affected all minority students, regardless of their racial balance status.

The only published study to date that compares NAEP achievement trends for black students according to school racial composition found that black achievement gains were approximately the same in majority nonwhite and majority white schools.27 Although this finding could reflect program improvements due to desegregation and implemented in all schools, the study suggested that improvement in parents’ education (as documented in the NAEP data) was a major contributor.

A more direct way to determine the relationship between academic achievement and desegregation is to show how students are performing on NAEP tests according to their school racial composition. Because students who attend predominately white schools are more likely to have higher socioeconomic (SES) levels than those who attend predominately minority schools, such as higher parent education, more two-parent families, and more educational materials at home, we have to adjust the test scores of students to take these SES characteristics into account. This is done using a statistical technique (multiple regression) that removes the effect of SES on student achievement.28 This statistical analysis enables us to adjust students’ test scores for SES and in that way to make clear comparisons of students of the same social class in schools of varying racial composition.

Fig. 7. School percent white and reading achievement, 1992 NAEP, age 13, adjusted for individual SES.

Figure 7 shows the results of an original analysis of the national reading scores for thirteen-year-olds in 1992, adjusted for individual SES characteristics. Across the first four categories of racial composition, ranging from predominately black to 80 percent white, there are no significant differences or trends in black achievement or in the black-white gap. These schools enroll more than 97 percent of the national black sample. There is a significant improvement for a small group of black students in predominately white schools (only 45 out of 1,329 in the black sample), most of whom are in 90 percent–plus white schools. Even assuming this is a reliable result, it is not likely due to racial composition itself, given the lack of a trend in the other categories; more likely, it is a self-selection effect or an effect of unmeasured family characteristics that cannot be evaluated.

Fig. 8. School percent white and math achievement, 1992 NAEP, age 13, adjusted for individual SES.

Figure 8 shows an analysis of math achievement for the same group of students. The pattern here is quite different; both black and white students score higher in schools that are over 40 percent white, and therefore the achievement gap remains relatively constant across the first four categories of school (it narrows significantly only for the small group of black students who are in highly white schools). Because the improvement occurs for both groups, it is probably not due to racial balance; rather, it is more likely to be due to programmatic differences among the schools or unmeasured SES factors. Whatever the cause, attending a racially integrated school does not reduce the math achievement gap between black and white students in this national sample.

Similar results were obtained for the seventeen-year-old age groups as well as for Hispanic students. Therefore, the NAEP data do not support the thesis that desegregated schools significantly benefit black or Hispanic reading achievement, nor do they reduce the black-white or Hispanicwhite achievement gap in mathematics.

CASE STUDIES

The NAEP data have the advantage of being national in scope, but they have some serious drawbacks as well. For example, the NAEP data do not include information on the existence, scope, or duration of desegregation plans within a district. Although the data include the racial composition of the student’s school, they do not explain why the racial composition exists or how long it has been that way. For this reason it is useful to look at a number of case studies of the effects of comprehensive desegregation plans on achievement.

Several recently published case studies include Pasadena, California; Norfolk, Virginia; and Charleston, South Carolina.29 In 1970, three years before the Keyes decision, Pasadena became one of the first Northern school districts to implement a districtwide racial balance plan, having been ordered to do so by a federal district court. From 1970 to 1973 there was no improvement in the reading scores of a first-grade cohort of black students in fully desegregated schools.

Norfolk, Virginia, was fully desegregated and racially balanced by 1970. Fourth-grade test scores were available from 1965 to 1982. Both black and white test scores actually fell dramatically after desegregation. The blackwhite gap narrowed somewhat between 1970 and 1973 but only because white achievement fell more than black achievement. Between 1973 and 1980 achievement scores improved for both races, but they did not reach their predesegregation levels until 1978; the gap remained constant during this time. A special compensatory program that was initiated in 1978 may have produced the gains in 1979 and 1980. But in 1981, when a new test form was introduced, scores fell again. There was also some evidence that improper coaching and teaching before the test may have accounted for some of these increases. Even so, one can certainly conclude that racial balance had no positive impact on black or white fourth-grade achievement between 1970 and 1978.

Similar results were found in Charleston, South Carolina, where the court-approved desegregation plan did not produce racial balance in every school. Charleston was a countywide district with subdistricts that governed student and teacher assignment, hence racial balance was required only within subdistricts. The case study compared reading achievement for black third- and fourth-graders in schools with varying degrees of racial composition across subdistricts and found no significant difference in test scores, or change in test scores, among predominately black schools, predominately white schools, and racially balanced schools.

Fig. 9. Trends in New Castle County, Delaware, 6th-grade reading.

Some additional case studies (unpublished to date) are of special interest because of the nature of their desegregation plans. Three of these case studies present original data analyses from school districts discussed by Orfield and Eaton: Kansas City, Missouri; Prince Georges County, Maryland; and Charlotte-MecklenburgCounty,North Carolina. Another is from the Wilmington–New Castle County district, which had a unique courtordered desegregation plan.

In 1978 a federal court ordered the merger of predominately black Wilmington, Delaware, with ten predominately white New Castle County districts to form a single metropolitan school district, and a countywide desegregation plan was implemented. It is one of the few metropolitan consolidation and busing plans to be ordered by a court and ultimately adopted. The case is of special interest because, unlike many central city plans, black students attend most of their school years in majority white suburban schools whose student bodies have remained relatively middle class (in some cases, affluent) since the start of the plan. As late as 1993 the New Castle County districts (the single district was divided into four districts in 1981) were among the most racially balanced districts in the country. Not only did the vast majority of schools range from 65 to 75 percent white, but also every school was at least majority white.

Figure 9 shows trends in sixth-grade reading scores for students in all four New Castle County districts.30 In 1985, when sixth-graders would have attended racially balanced schools since kindergarten, we see a black-white gap of 15 points, or about three-fourths of a national standard deviation. Moreover, there is no improvement in black sixth-grade test scores, nor is there a reduction in the achievement gap over the nine years shown. In 1993 the black-white gap is 17 points (on a different test), or about eighttenths of a standard deviation, similar to the 1992 achievement gaps for the NAEP reading tests. In spite of these stable, desegregated, and majority white schools located in middle-class neighborhoods, the presumed ideal environments for raising black achievement, there are neither significant black gains nor reductions in the black-white gap. Clearly, the Wilmington case does not support the thesis that racial balance will reduce the achievement gap.

Fig. 10. Trends in Kansas City, Missouri, 5th-grade achievement, ITBS composite scores.

Nor is it a question of money. In 1986 a federal court ordered Kansas City, Missouri, to implement what may well be the most expensive remedial plan in history. Kansas City had been operating a desegregation plan since the late 1970s. But in 1986 the court ordered an expanded plan involving extensive construction, renovation, and the addition of magnet programs to most of the elementary and all the secondary schools, whose purpose was to attract suburban whites into this 70 percent minority school system. With a unique court-ordered tax levy and court-ordered funding from the state, total school expenditures reached $10,000 per pupil by 1990, with total funding exceeding $1.5 billion over approximately an eight-year period. Unfortunately, not enough white students came to the city from the suburbs to lower the minority percentage significantly.

As the achievement trends in Figure 10 reveal, this extraordinary degree of expenditures for one of the largest magnet programs in the nation apparently did nothing to raise the achievement levels of black students, which remained substantially below white achievement from 1988 to 1995. In fact, when a new test form was implemented in 1995, scores of all students fell significantly, raising the possibility that some coaching or teaching of test content may have taken place during preceding years. Although the black-white achievement gap is somewhat narrower than in Wilmington or in the NAEP, this appears to be due to low white scores rather than to high black scores (Kansas City has a relatively high proportion of white students in the free lunch program). Thus, spending an extraordinary amount of money on a school desegregation plan and on magnet schools does not seem to improve minority achievement significantly or decrease the minority-white achievement gap.

Fig. 11. Prince Georges County, Maryland, black test scores by school type, 1996, grade 3.

The Prince Georges County case is noteworthy because of a different approach taken in its second school desegregation plan. A predominantly white suburban county outsideWashington, D.C., Prince Georges adopted a comprehensive pairing and busing plan in 1973. This was followed by extensive white flight and black in-migrationfrom the District of Columbia, and the county schools became majority black in the early 1980s. Recognizing that not all schools could achieve meaningful racial balance, a modified desegregation plan was adopted in 1985 that provided for several types of schools: desegregated magnet schools, desegregated regular schools, and “Milliken II” schools, which were predominantly black and received additional resources and funding. Later, as white enrollment losses continued, some of the magnet schools and some of the regular schools became predominately black; they also added a category of “Model” schools, which were predominately black and received extra resources but not as much as Milliken schools.

These different types of programs and desegregation levels offer a unique opportunity to sort out the potential effects of racial balance from the effects of special compensatory programs and extra funding. Figure 11 shows third-grade test scores (from a Maryland statewide test) for black students in six types of schools, after adjusting scores for student background characteristics. Though the scores vary somewhat across the different types of schools, there is no apparent benefit from attending a desegregated school, most of which are approximately half black and half white. After controlling for SES and initial ability, black students attending desegregated magnet and desegregated regular schools score no higher than black students at predominately black regular schools (which average around 90 percent black), and they score somewhat lower than black students in the enriched but predominately black Milliken and Model schools (which exceed 90 percent black). It is not clear why black students score lowest at the predominately black magnet schools and highest at the predominately black Model schools, but whatever the reason it is not because of their racial composition, which is identical.

Fig. 12. Achievement trends in Charlotte-Mecklenburg, North Carolina, grades 3 and 6.

The final case study to be presented is Charlotte-Mecklenburg, North Carolina, which is especially important historically because of the famous U.S. Supreme Court decision in this case. The Swann decision upheld a lower court order to adopt a comprehensive, countywide mandatory busing program to attain racial balance in all schools, thereby creating a standard for desegregation that was applied throughout the nation for many years. Because Charlotte was a large county district with a relatively low black enrollment in 1970, it was able to sustain very extensive racial balance for the next twenty years while at the same time preserving a majority white school system. It has been described in the Orfield and Eaton book as one of the most successful desegregation plans in the country, with unprecedented local support from educators, civil rights leaders, and the business community.

Figure 12 shows the long-term achievement trends for third and sixth graders from 1978 to 1997. In 1978 black elementary students had been in well-desegregated schools from kindergarten on, yet they scored only at the 20th percentile in both reading and math compared with national norms. Test scores began increasing dramatically for both black and white students, reaching maximums in about 1983, when black students scored at the national norm of 50. The reason for the increase is not clear, but it is not likely due to desegregation, which had been in place for eight years before 1978. Some of the increases may have been due to coaching or teaching test content because scores for both races dropped significantly when a new test form was adopted in 1986. Though the black-white gap had diminished somewhat by 1983, it returned to its original magnitude in 1986 (about one standard deviation) and remained relatively constant until the last CAT test was administered in 1992. A sizable test score gap also existed between 1994 and 1997, but it cannot be compared directly with the CAT gap because it comes from a different statewide test that uses a different metric. Thus, one of the most successful desegregation plans in the country did not reduce the black-white achievement gap significantly.

These case studies bolster the NAEP results shown in Figures 7 and 8, that racial composition by itself has little effect on raising the achievement of minority students or on reducing the minority-white achievement gap. Some studies show that there is no relationship at all between black achievement and racial composition (controlling for student SES), and other studies show that there is no relationship between the black-white achievement gap and racial composition. In either case, though there is some evidence here that achievement can be affected by programmatic changes, there is no evidence that it responds to improved racial balance by itself.

RACE RELATIONS AND ATTITUDES

Studies of the effect of desegregation on racial attitudes and race relations were fairly common during the 1970s, when the desegregation movement was at its peak. There seems to be less interest in these issues now, perhaps because the early studies offered little support for the notion that desegregation would bolster self-esteem, lower racial prejudice, and improve race relations.

The first major review of the effect of desegregation on race relations and attitudes was by St. John, whose study was published in 1975.31 She evaluated 35 studies conducted between 1966 and 1973 of changes in white attitudes or behaviors after desegregation. She found 11 studies in which white prejudice worsened, 11 where it improved, and 13 where there was no change or mixed results. She also evaluated 28 studies of changes in black prejudice after desegregation, finding 10 where black prejudice worsened, 6 where it improved, and 12 where there was no change or results were mixed. Her review also evaluated the effect of desegregation on the self-esteem of black children, one of the key concerns of the Supreme Court in Brown. Out of 35 studies, she found 14 where black self-esteem was higher in segregated than desegregated schools, 5 where it was higher in desegregated schools, and 16 where there was no significant difference.

Walter G. Stephan conducted a similar type of review about ten years later.32 Although he reviewed a smaller number of studies, his 28 studies covered a longer time span (1963–1981) and included only those with a reasonably sound methodology. His findings were remarkably similar to those of St. John. Out of 24 studies of change in white attitudes or behavior, he found 11 where desegregation increased white prejudice, 4 where it decreased, and 9 where there was no change. Of the 17 studies of black attitudes and behaviors, he found 5 showing that desegregation worsened black prejudice, 4 where it improved, and 8 where there was no difference. In his review of black self-esteem in 28 studies, he found 7 where desegregation lowered black self-esteem, 1 where it was improved, and 19 where there was no change. It is interesting that the majority of studies during this period found that black self-esteem was generally higher than white self-esteem, thus calling into question the Supreme Court’s psychological harm thesis.33

The evidence on the benefit of school desegregation for race relations is probably the weakest of all. Indeed, there are more studies showing harmful effects than studies showing positive effects. This led to another and more recent reviewer of the race relations literature to conclude, somewhat generously: “In general, the reviews of desegregation and intergroup relations were unable to come to any conclusion about what the probable effects of desegregation were. . . . Virtually all of the reviewers determined that few, if any, firm conclusions about the impact of desegregation on intergroup relations could be drawn.”34

The reluctance of reviewers to draw conclusions about the benefits of school desegregation for race relations or self-esteem only reinforces our conclusion that the psychological harm theory of de facto segregation and the social benefit theory of desegregation are clearly wrong, at least when applied to desegregation as a racial balance policy. Of course, the original psychological harm theory presumed official segregation sanctioned by laws, which was pretty much gone by the mid-1960s and before the time when most of the studies above were conducted. But supporters of the harm and benefit thesis have applied it far more broadly, to cover any type of racial imbalance or racial disparity arising from any cause, and even today it forms the ideological basis for desegregation and affirmative action policies. It is this broader formulation that fails under scrutiny of the research evidence.

Conclusions

What can we say about the success and failures of the school desegregation movement? If we evaluate school desegregation policy within the constitutional framework established by the Supreme Court on the matter of race in Brown, which is much narrower than the goals of many civil rights advocates, there are indeed successes. The failures come, for the most part, from attempts to extend this original framework to see school desegregation as a means of resolving a whole host of racial issues never contemplated by Brown.

First, we must not diminish the fact that Brown forever changed the fundamental way in which America deals with race, not only in schools but also throughout society. In particular, it forbade all official actions that segregated students or staff by race, or that sanctioned discriminatory distribution of resources, facilities, and activities. Second, after the Court allowed consideration of race in fashioning desegregation plans, what followed was an unprecedented change in the way students and staff were assigned to schools, eventually creating extensive racial balance in schools to an extent never attained in any other sector of society.

That racial balance has largely continued to the present time, with or without court orders. Although the Supreme Court has returned to its original conception of race, permitting school systems to be declared “unitary,” after which race cannot be used for student assignment (whether for racial balance or imbalance), there has been no rush to dismantle desegregation plans and return to strict neighborhood school policies. The reason is that racial balance and racial diversity have become desired goals, especially within the educational establishment. True, a number of unitary school systems, like Norfolk, Oklahoma City, Cleveland, and Prince Georges County, now have some neighborhood schools, but most of these districts have retained such politically palatable desegregation policies as magnet programs, voluntary transfers, and minimal integrative geographic zoning. Indeed, that is why the racial balance indexes have remained so stable for the past fifteen years.

The biggest threat to desegregation is not the dismantling of plans but rather the inexorable demographic changes that have left the majority of larger school systems predominately minority. No type of permissible plan will result in substantial desegregation in New York, Los Angeles, Chicago, Cleveland, Detroit, Atlanta, Dallas, and dozens of other large cities.35 The declining measures of interracial exposure for both black and Hispanic students are caused not by dismantling desegregation plans but rather by the falling percentage of white students in all but the smallest public school districts. Though these demographic trends do effect the success of desegregation plans, they do not reflect failures of the courts or of public policy in general.

It is only when we turn to educational and social benefits that we find the most significant failures of desegregation policy. Of course, these are not failures of Supreme Court doctrines or federal law, which never contemplated such benefits; rather, they are a failing of the benefit theory as embraced by many civil rights advocates, some lower courts, and even a few Supreme Court Justices. Desegregation and racial balance have not brought the gains in minority achievement and a closing of the achievement gap envisioned by so many believers in this thesis because the thesis is invalid.

There is some evidence that special programs have benefited achievement slightly—even when not accompanied by racial balance—but the benefit has not been sufficient to close the black-white or the Hispanicwhite gaps. In all the case studies discussed here, and in many others not discussed, the achievement gaps in well-desegregated school systems remain very close to the national achievement gap, and this gap is only slightly smaller than the gap that existed when major desegregation plans began in 1970. If desegregation is responsible for this modest reduction in the gap, it is most likely due to programmatic and resource changes, not racial balance policies.

There is even less evidence that school desegregation has reduced racial prejudice, improved race relations, or benefited self-esteem. Here the expectations of benefits were based either on inadequate social theory or on misinterpretations of the widely accepted contact theory of Gordon Allport. 36 Given the very large differences in academic achievement between black and white students, and the fact that desegregation did not eliminate these differences, it may have been naı¨ve to expect that integrated schools would improve black self-esteem, when blacks were plunged into an academic environment where the competition was far more rigorous than in segregated schools. The unequal achievement and resulting unequal grades in desegregated schools also violated one of the contact theory’s requirements for positive race relations—that there would be equal status contact. This is not to say there are no social benefits of school desegregation.

First, school desegregation has brought each race greater personal knowledge about other races living in their community, albeit without guaranteeing positive race relations in all cases. Second, knowledge, proper instruction, and tolerant staff can help schools become places where children learn how to treat and respect persons of other races, which helps prepare them for the later realities of living in a multiracial society. Finally, many individual students do find meaningful interracial experiences and friendships, which they value and want to have, even though this outcome may not apply to the majority of students. This is one of the reasons we have always supported voluntary desegregation plans, which let individual students and families decide whether or not to attend an integrated school, for whatever reasons are important to them.

It is this last set of social benefits that leads us to support desegregation policies for unitary school systems, though these are more restricted than the original types of plans. Desegregation should be strictly voluntary, as with magnet schools or open enrollment options, and it should be available across district lines as part of choice programs that include private schools. Race cannot be used explicitly as a criteria for school assignment in a unitary system, even in the case of voluntary options. Poverty (i.e., free lunch status) is, however, a permissible criteria for government decisions. By offering voluntary transfer options to all students, with some financial incentive (such as transportation) for free lunch students, we can expand the opportunities for desegregation for those families that desire it most, across district lines and between public and private sectors, and without the unintended negative consequences that have plagued the massive mandatory busing schemes of the past.

Notes

1. The criteria for a unitary or nondiscriminatory school system were first formulated

in a brief footnote in a Supreme Court decision, Green v. New Kent County

(Va.), 430 U.S. 391 (1968). There was no discussion or explanation of the meaning of

these factors or the level of compliance that would be required.

2. Gary Orfield and Susan E. Eaton, Dismantling Desegregation (New York: New

Press, 1996), p. 53.

3. Alexander M. Bickel, “Untangling the Busing Snarl,” New Republic, September

23 and 30, 1972.

4. James S. Coleman, Sara D. Kelly, and John A. Moore,Trends in School Integration

(Washington, D.C.: Urban Institute, 1975).

5. James S. Coleman et al., Equality of Educational Opportunity(Washington, D.C.:

U.S. Government Printing Office, 1966).

6. Green v. New Kent County at 437.

7. Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1 (1971).

8. Keyes v. School District No. 1, Denver, 413 U.S. 189 (1973).

9. HEW, Office of Elementary and Secondary Education, did a school census in

1967 and published a book with enrollment by school and three ethnic groups, white,

black, and other, in several thousand school districts. The data was not kept in machinereadable

format.

10. For a more detailed description of the original national study, see Christine H.

Rossell and David J. Armor, “The Effectiveness of School Desegregation Plans,” American

Politics Quarterly 24 (1996): 267–302.

11. The formula is D=1/2 ∑ | (Wi/W)-(Bi/B) | where Wi is the white enrollment

in each school,Wis the white enrollment in the school system, Bi is the black enrollment

in each school, and B is the black enrollment in the whole school system. The absolute

sum of the difference between whites and blacks is calculated for each school, then

summed down the schools, and the total divided by two.

12. It also equals the sum of the proportion of students of one race and the

proportion of students of the other race that would have to be reassigned to attain

perfect balance. In a school system exactly half white and half black, the percentage of

each group to be assigned could be derived by dividing the index by two.

13. The number of cases for small districts was insufficient to provide stable estimates.

14. All Southern districts with both black and white students had to have some type

of plan if they had practiced segregation in 1954; some districts may have eliminated

segregation by geographic zoning, which they did not view as being a formal plan. Only

ten districts fell into this category.

15. The index of dissimilarity measures the imbalance of only two groups at a time.

In calculating Hispanic-white imbalance, black and other enrollment is ignored. Similarly,

in the calculation of black-white imbalance, Hispanic and other minority enrollment

is ignored.

16. See Coleman, Kelly, and Moore, 1975.

17. The formula is E=(∑inbipwi)/∑nbi where nbi is the number of black students

in school i, and pwi is the proportion of white in the same school. The number of black

students is multiplied by the proportion of white in each school, and this is then

summed across schools. That total is divided by the number of blacks in the school

system.

18. It is necessary to exclude districts with very small numbers of Hispanic students,

because they tend to inflate the exposure index with very high values.

19. David J. Armor, “The Evidence on Busing,” Public Interest 28 (1972): 90–126.

20. Orfield and Eaton, Dismantling Desegregation.

21. See Armor, Forced Justice: School Desegregation and the Law (New York: Oxford

University Press, 1995), pp. 79–81. Black and white scores in Norfolk actually fell after

desegregation in 1970, and neither group returned to its predesegregation achievement

levels until 1978. The black-white gap narrowed somewhat between 1970 and 1973,

but only because white achievement fell more than black achievement. Achievement

improved for both groups between 1973 and 1980, but the gap remained constant

during this time; scores fell again in 1981 when a new test form was introduced.

22. Nancy St. John, School Desegregation (New York: Wiley & Sons, 1975); Thomas

Cook et al., School Desegregation and Black Achievement (Washington, D.C.: National

Institute of Education, U.S. Department of Education, 1984); JanetWard Schofield, in

Banks and Banks, Handbook of Research in Multicultural Education (New York: Macmillan,

1994).

23. Cook et al., pp. 40–41.

24. Coleman et al., Equality of Educational Opportunity.

25. It fell to 0.5 sd’s for thirteen-year-olds in 1988 and 1990, but by 1992 this age

group was back to 0.71.

26. David Grissmer, Ann Flanagan, and Stephanie Williamson, “Why Did the

Black-White Score Gap Narrow in the 1970s and 1980s?” in Christopher Jencks and

Meridith Phillips, eds., The Black-White Test Score Gap (Washington, D.C.: Brookings

Institution, 1998).

27. Armor, Forced Justice, p. 95.

28. The adjustment uses multiple regression analysis to predict student scores solely

according to their parents’ education, family structure, household reading items, and

gender.

29. Armor, Forced Justice, pp. 77–83.

30. The break in the trend line denotes the change in 1989 from the California

Achievement Test (CAT) to the Stanford Achievement Test (SAT).

31. St. John, School Desegregation.

32. Walter G. Stephan, “The Effects of School Desegregation: An Evaluation 30

Years After Brown,” in M. Saks and L. Saxe, eds., Advances in Applied Social Psychology

(Hillsdale, N.J.: Erlebaum, 1986).

33. In fairness to the Supreme Court, they were simply referring to the testimony

of social scientists at that time, who may have misinterpreted the famous doll studies

of Kenneth and Mamie Clark. See Armor, Forced Justice, pp. 99–101.

34. Janet Ward Schofield, “School Desegregation and Intergroup Relations: A Review

of the Literature,” in Review of Research in Education 17: 335–412, ed. Gerald

Grant (Washington, D.C.: American Educational Research Association, 1991), p. 356.

35. The Supreme Court established in 1974 in Milliken v. Bradley that there can be

no cross-district remedy without a cross-district violation (i.e., collusion between city

and suburban government officials to segregate the city and suburban school districts).

Proving a cross-district violation is almost impossible, and there have been only a

handful of cross-district remedies since 1974.

36. Gordon Allport, The Nature of Prejudice (Reading, Mass.: Addison-Wesley,

1979).

==============================

The Racial Gap in Academic Achievement

ABIGAIL THERNSTROM

As early as second or third grade, [African American students] generally have much lower grades and test scores than Asians and Whites—patterns that persist over the course of their school careers.

—The College Board, “Reaching the Top: A Report of the National Task Force on Minority High Achievement”

You have to demand more of your students while providing them with the structure to meet those demands. The more difficult the curriculum, the greater the likelihood your students will be successful.

—Gregory Hodge, principal, Frederick Douglass Academy in central Harlem, New York City

In 1903, W. E. B. Du Bois, the preeminent black intellectual of his time, described “the problem of the color line [as] the problem of the twentieth century.” And four decades later he peered sixty years down the road and saw a new century in which—for blacks—there would be nothing new.

We have, of course, arrived at that new century, and much has changed. But in certain respects, not nearly enough. Although almost half of blacks are today in the middle class, racial inequality remains a fundamental fact about American life. And much of that inequality can be traced to one source: the underachievement of black students in school. The racial gap in cognitive skills is the nation’s most serious educational problem, and its most critical race-related issue.

Most discussions of the racial gap in academic achievement contrast the scores of blacks and Hispanics to those of Asians and whites. But the Hispanic and black stories are not the same. Hispanics are a classic immigrant group. With the passage of time, their academic profile (already less worrisome than that of blacks) is likely to change. African Americans are the real concern.

Skills Matter

Educational reform is sweeping the nation. Every state, except Iowa, has developed specific learning standards, and all but three are committed to statewide tests to determine compliance with those standards in at least one core subject. Thirty-eight states have aligned their standards, curricula, and assessments in all four major subjects—English, math, science, and history. More than half the states are making high school graduation conditional upon a passing grade on a state test. The standards are often rock-bottom, but even clearly defined minimal expectations are a sharp break from the past.1

Teachers, too, are facing tougher tests for certification, and many states have the power to assume control of a failing school, or even an entire district. Charter schools are bursting out all over; thirty-six states and the District of Columbia now have some sort of charter law. There are other reforms connected to high standards and accountability: early intervention programs, some of which target both parents and children; an end to social promotion; summer school, after-school, and Saturday programs; a variety of incentives (including free college tuition) designed to promote high achievement;summer workshops in content areas for teachers; the addition of more reading and writing to the curriculum; and much else.

Educational improvement means, above all, raising the performance of urban students, and particularly those who are black. The academic performance of black youngsters on standardized tests is worse than that of any other group; if educational reforms do not raise black test scores, they will have failed. And if current efforts fail, a substantial percentage of African American families will remain below the poverty line. Blacks and whites today have equal high school graduation rates and nearly equal college attendance rates. Over 85 percent of both groups finish four years of high school (although not necessarily by age eighteen), and among those who obtain a high school diploma, 60 percent of blacks (versus 68 percent of whites) go directly into college. But blacks and whites do not leave the twelfth grade equally well educated, and thus they do not have the same college graduation rates or the same earnings down the road.

Blacks are more likely to earn a college diploma than whites with the same twelfth-grade test scores.2 And blacks earn no less than whites when they are equally academically prepared. There is much talk in the media and elsewhere about the rich getting richer and the poor becoming poorer. In fact, it is those with skills who are doing well economically. In recent decades, earnings for low-skilled work have declined, while those for jobs that require a good command of language and numbers, as well as the ability to analyze complicated problems, have risen dramatically.3

Thus, it is actual knowledge—not years spent in school warming a seat—that makes the difference in an individual’s income. What often looks like discrimination is better described as rewarding workers with relatively strong cognitive skills. One study of men twenty-six to thirty-three years old who held full-time jobs in 1991 found that when education was measured in the traditional way (years of school completed), blacks earned 19 percent less than comparably educated whites. But when the yardstick was how well they performed on basic tests of vocabulary, paragraph comprehension, arithmetical reasoning, and mathematical knowledge, the results were reversed. Black men earned 9 percent more than white men with the same education—as defined by skill.4

Other studies contain similar findings. Among men thirty-one to thirty-six years old whose cognitive skills put them above the 50th percentile on the well-respected Armed Services Vocational Aptitude Battery test, the difference between black and white earnings is a mere 4 percentage points, Christopher Jencks and Meredith Phillips report in a Brookings Institution volume on The Black-White Test Score Gap.5 A 1999 College Board report came to much the same conclusion: “Differences in job performance ratings and wages among people with similar educational credentials are related in part to differences in academic achievement and skills levels, as measured by standardized test scores, class rank in college, and even high school grades or test scores.” Other factors do come into play—”motivation, perserverance,” even “plain old luck,” it went on. White racism, however, was not on the list.6

Scary Numbers

Test scores matter, and those for blacks and whites, on average, are not the same. For a while the rate of progress was heartening. During the 1970s and the better part of the 1980s, black school children were making more rapid gains than whites on the standardized tests administered periodically to a representative sample of students by the National Assessment of Educational Progress (NAEP). NAEP results are often called the nation’s report card on education, and in 1971, when the tests were first given, the average African American seventeen-year-old could read no better than the typical white eleven-year-old. By 1988, the gap had been narrowed to just 2.5 years, still too large but less than half of what it had been sixteen years before. Math and reading tests are given in different years, but the data are similar. In 1973, black students were 4.3 years behind their white peers; in 1990, the difference was down to just 2.5 years. Toward the end of the 1980s, however, black progress came to a halt. Scores began to go down. By 1996, the racial gap in math of 2.5 years in 1990 had grown to 3.4 years. The average African American high school senior thus had math skills precisely on a par with those of the typical white student in the middle of ninth grade—a huge disparity in a world in which numeracy is becoming at least as important as literacy.

The reading scores are even more dismaying. Although the black-white gap closed by a remarkable 3.5 years between 1980 and 1988, it widened by more than a year and a half in the following decade, erasing half of the previous gain. As a result, in 1998 the average seventeen-year-old African American could only read as well as the typical white child who had not yet reached age thirteen. In 1992, just 18 percent of black students in twelfth grade were rated “proficient” or “advanced” in reading, as compared with 47 percent of whites. As of 1998, those numbers were unchanged. Backsliding is evident in both science and writing as well. In the decade following 1986, the science gap widened by a half a year, and between 1988 and 1996, the difference between white and black scores on the writing assessment increased by 1.2 years. The most recent data thus show black students in twelfth grade dealing with scientific problems at the level of whites in the sixth grade and writing about as well as whites in the eighth grade.

The Middle-Class Puzzle

The racial gap in academic skills is a problem that starts before kindergarten and affects black youngsters from every social class. Thus, the median scores of black and white five- and six-year-olds who took the Peabody Picture Vocabulary Test in the years 1986–1992 were quite different. 7 In an essay in The Black-White Test Score Gap, Meredith Phillips and colleagues estimate that about half the difference between white and black twelfth-grade students is explained by the gap that is already evident in first grade.8 Could that gap be eliminated by better schooling? No one knows for sure. But schools scattered across the landscape are beating the demographic odds—educating children from whom regular public schools expect less.

More discouraging, because it’s more bewildering, is the poor academic performance of middle-class black students. Shaker Heights is a tony Cleveland suburb; its black residents are well-to-do. And its schools, as of 1998, were pouring nearly $10,000 a year into educating every child. It has had a rich program aimed at raising the level of black achievement. Successful black students mentor those doing poorly in the earlier grades; tutors work with small groups of kindergarten children who need help reading; after-school, weekend, and summer academies are open to those who want to attend; special high school classes are available for those who do not keep up; a high school counselor works with youngsters whose grades do not reflect their potential; and so forth. The results have the school administators totally baffled. About half the students in the Shaker Heights school system are black, but blacks are 7 percent of those in the top fifth of their class, and 90 percent of those in the bottom fifth.9

“At virtually all socioeconomic levels,” black students do not perform as well as those who are white or Asian, the College Board’s recent task force report acknowledged. In fact, the racial gap in academic achievement is widest among middle-class students from educated families; the NAEP scores of black and white youngsters whose parents lack even a high school degree are more alike.10 The SATs, too, paint a dismal picture. In 1995, black students from families in the top income bracket—$70,000 and up— were a shade behind whites from families earning less than $10,000 on the verbal assessment and significantly behind them in math.11

Trash the Tests

Many educators, policy makers, and others argue, of course, that staring at these numbers is a fundamentally misguided way of looking at the question of black academic achievement. Test scores tell us little, they say, and should never be used to decide who moves up a grade, attends summer school, or graduates. Their disparate racial impact “should be cause for further inquiry and examination,” the U.S. Department of Education’s Office of Civil Rights (OCR) contends. In draft guidelines issued in December 1999, it warned states and local school districts that they must be prepared to defend their “high stakes” tests in federal court. Low scores may reflect limited educational opportunities; if so, decisions that rely upon test performance are “problematic”—that is, discriminatory. Numerical cut-off scores on a statewide test as a condition of graduation are particularly suspect.12

OCR’s views square with those of many who identify themselves as members of the civil rights community. Christoper Edley Jr. was President Clinton’s main advisor on race for a number of years; the guidelines, he said, protect students from “harsh high-stakes sanctions.”13 In a resolution approved on November 19, 1999, the NAACP came out against the use of SAT and ACT exams for purposes of college admission on the grounds that they are racially biased and a poor indicator of success in college.14 In doing so, it ignored much evidence that, in fact, the SAT overpredicts black college grades; black students tend to do worse than their test scores suggest they should. The tests are indeed biased—in favor of black students.15

In 1999, the Educational Testing Service, too, took a swipe at the SATs for which it is ultimately responsible. When kids are from disadvantaged households—as measured by a host of possible criteria—SATscores should be adjusted to reflect the hurdles they must overcome, ETS tentatively suggested in September 1999. A score of 1,000 should be read as 1,200 if the student is black, for example, and has thus outperformed the black average.16 The arguments of both the NAACP and ETS logically apply to all tests at every grade level. If standardized assessments are biased and therefore misleading, then grades should be adjusted. On an eighth-grade math test, a Hispanic’s “B” is really an “A.” Scores are not scores. They are numbers to be read within a demographic context. And the uniformly high academic expectations that are integral to standards-based educational reform are a grave mistake.

That is certainly what Jonathan Kozol, a well-known writer on education, believes. The increasingly popular statewide tests, he has declared, evoke memories of “another social order not so long ago that regimented all its children . . . to march with pedagogic uniformity, efficiency, and every competence one can conceive—except for independent will—right into Poland, Austria, and France, andWorldWar II.”17 Deborah Meier, the much-celebrated founder of the Central Park East schools in New York City, condemns standardized tests as failing to “measure the only important qualities of a well-educated person.” “Life scores [not math scores] based on living” should be the educator’s concern, she has written.18 Theodore Sizer, former dean of the Harvard Graduate School of Education, is also opposed to the entire standards-based educational package. “The myriad, detailed and mandatory state ‘curriculum frameworks,’ of whatever scholarly brilliance, are attacks on intellectual freedom,” he says. “High stakes tests arising from these curricula compound the felony.”19

Kozol, Meier, Sizer, and their allies have many fans, including most members of the faculty at the leading graduate schools of education. U.S. Senator PaulWellstone has introduced legislation to ban standardized tests from being used as the sole criterion for grade promotion and high school graduation.20 But they seem to be fighting a losing war. Powerful currents are pushing education in a different direction. Asked in a 1999 poll whether they had “respect and confidence” in public schools, only 36 percent of Americans said they had a “great deal/quite a lot of confidence.” A year earlier 59 percent had listed low academic standards as a “very serious concern,” with another 29 percent choosing “somewhat serious.” Moreover, polling by the nonpartisan Public Agenda in the winter of 1997–98 revealed that 78 percent of black parents agreed that testing “calls attention to a problem that needs to be solved.”21

Peter Sacks, the suthor of a recent book on testing, dismisses those who support testing as part of a “conservative backlash advocating advancement by ‘merit,’” but important advocates for black and Hispanic children have joined the standards crusade.22 Ramon C. Cortines, formerly the interim superintendent of schools in Los Angeles, recently called the attack on assessments “insidious racism” on the part of those who “want the status quo.”23 To ignore the information that tests provide, in other words, is to perpetuate a system that has been historically indifferent to the fate of black and Hispanic students. The Education Trust is a Washington, D.C., organization devoted to promoting academic achievement among blacks, Hispanics, and low-income children. “Traditionally, of course, most advocates for such children have shied away from high standards and high stakes,” Kati Haycock, the director, has written.24 But she in unequivocal in her demand for no-nonsense standards with uncompromising consequences for those who fail to meet them. “Backing off from testing would just kick the problem of low student achievement under the rug again,” she has said.25

Peter Sacks talks about the “stigma of poor performance on mental tests.”26 But Kati Haycock and those behind the drive for state educational reform are more fearful of the consequences of letting schools ignore the racial gap in academic achievement. That fear has prompted New York, Texas, and California to consider racial and ethnic data in ranking public schools. In New York, for instance, beginning with 2001, the performance of black and Hispanic students in each school will be reported, and schools with a large racial gap in academic performance won’t earn a good grade. Thus, an affluent suburb in which the average score on the statewide tests is high will be penalized if that average masks group differences.27 New York and other states are rejecting the notion of racial double standards—lower academic expenctations for non-Asian minorities. The job of educators is to close the racial gap, they say; the tests tell a story that America ignores at its peril. The scores cannot be fudged with formulas that adjust for “disadvantage” and suggest that black children cannot keep up with their Asian peers.

An Unsolved Mystery

Why do so few black students graduate from high school with strong academic skills? Black poverty, racial segregation, and inadequate funding for predominantly black schools are standard items on the list of explanations. None withstands close scrutiny. Income inequality, for instance, appears to play only a very small role in black test performance. The rise in scores after 1970 came at a time when the poverty rate was not changing significantly.28 And when that educational progress halted, black poverty was not on the increase. In fact, Jencks and Phillips conclude, eliminating black-white income disparities would make almost no difference in the score of young black children on a basic vocabulary test. (They focus exclusively on black-white differences, ignoring the roughly similar gap between blacks and Asians.)

The huge rise in out-of-wedlock births, precipitating the steep and steady decline in the proportion of black children growing up with two parents, might seem important in explaining educational performance of African American children. Children raised in single-parent families do less well in school than others, even when all other variables, including income, are controlled.29 But marriage, per se, appears to make little difference. “Once we control a mother’s family background, test scores, and years of schooling,” Jencks and Phillips conclude, “whether she is married has even less effect on her children’s test scores than whether she is poor.”30 Moreover, the disintegration of the black nuclear family, noted by Daniel Patrick Moynihan as early as 1965, was occurring rapidly in the period in which black scores were rising.

Do urban schools—which a majority of black youngsters attend—lack the resources needed to educate children properly? Spending on education has risen dramatically in recent decades, and those new dollars have gone disproportionately to schools with high concentrations of economically disadvantaged children.31 Contrary to widespread popular belief, there is little difference in average expenditures per pupil in central cities and suburbs.32

Nor does a school’s racial mix matter after the sixth grade; it seems to affect reading scores only in the early years and math scores not at all. Scores went up from the early 1970s to the late 1980s when courts were instituting busing plans, but black students who remained in predominantly African American schools showed as much improvement as those attending integrated schools.33 Moreover, the slide in black scores beginning in 1988 does not coincide with increasing racial isolation.34

The number of teachers per student, their credentials, and pay are also unaffected by the racial identity of the children in a district, Jencks and Phillips show. Schools that are mostly black, however, have teachers with lower test scores—in part because black schools have more black teachers, they forthrightly acknowledge. In addition, to the extent that schools substitute “critical thinking” for factual knowledge, those who come to school from homes that impart very little in the way of academic skills are likely to be further disadvantaged, a problem that is further compounded when, in the name of racial sensitivity, little is expected of black children.35 Moreover, disorder in the classroom is a sure recipe for little learning.36

Good Schools Get Good Results

Many will argue, of course, that politicians, state boards of education, and often the public are asking too much of schools. And solid scholars like Jencks and Phillips make a strong case that “cognitive disparities between black and white preschool children are currently so large that it is hard to imagine how schools alone could eliminate them.” Those skills will improve when we change the way black parents “deal with their children.” 37 They suggest that social scientists take a close look at: “the way family members and friends interact with one another and the outside world”; “how much parents talk to their children, deal with their children’s questions, how they react when their child learns or fails to learn something”; “cultural and psychological differences.”38

Pessimism is seductive, but good schools with impressive results scattered across the nation suggest that educational mediocrity is the central problem. All the students at the KIPP Academy, a South Bronx public school for grades 5 through 8, are from very low-income black or Hispanic families. As the school’s literature acknowledges, in the neighborhood in which these youngsters live “illiteracy, drug abuse, broken homes, gangs, and juvenile crime” are rampant. But KIPP has the highest reading and math scores of any middle school in the entire Bronx—despite the fact that two-thirds arrive in fifth grade unable to read. In math, last year, it ranked fifth in all of New York. Its daily attendance record of 96 percent is one of the highest in the whole state.39

KIPP’s success is no mystery. All great schools have fabulous leadership, and KIPP is the brainchild of David Levin, the director, who started out as the partner of Michael Feinberg, running a sister school in Houston. The school day runs from 7:30 a.m. to 5:00 p.m., with four hours of classes on Saturdays and three weeks of mandatory attendance in the summer. (The package adds up to 67 percent more time in the classroom than the national average.) Those who chronically do not finish their homework stay even later. Teachers are available on a toll-free cell phone number from the moment they leave the school until they arrive the next morning, and many receive ten to fifteen calls a night.

KIPP aims both to prepare students for college and to change peer culture. As the school counselor put it, “We nurture feeling good about being a nerd; we tell the students, you can play basketball and read books too.” Levin adds: “We train kids to look at people and talk to them. Every moment in class, teachers are role models, treating kids with respect. They dress properly. Everything about the school is designed to teach. I never walk by a student without saying hello.” He knows the name of every student, speaks Spanish, and thus can talk directly to Hispanic parents with limited English.

KIPP parents sign a contract that commits them to getting their kids to school on time, helping their children to the best of their ability, and so forth. Parents know how their kids are doing through a system of student “paychecks” that also reward the students directly. Checks (in KIPP dollars) are issued weekly on the basis of teacher evaluations in ten specific areas of performance—including dress. Once a parent has endorsed the check, it can be redeemed at the school store for books, supplies, and even CDs. Performance is linked to awards in other ways as well. Students who get 100 on a test see their names up on the classroom bulletin board. Students of the month are listed. “You can’t have students feeling good about nothing,” Levin says. “We make kids feel successful over little things . . . writing a complete sentence.”

Not a minute of the KIPP day is wasted. When students arrive in the morning, they pick up worksheets with problems to solve while teachers are correcting homework or are otherwise occupied. If they have five minutes, they do five minutes of work. Every class is structured, and the walls are plastered with posters that, for instance, walk students through the process of writing. (Think about what you want to say, organize your thoughts, write a draft, revise, proofread.) Other posters go through the definition of an independent clause and other rules of grammar. At every grade level, students write and write and write.

The messages the school delivers on its walls and through its teachers are integral to its success. The guidance counselor talks about the dangerous culture of the streets. Everywhere there are encouraging signs: “There Are No Shortcuts”; “No Excuses”; “Be Nice!!! Work Hard!!!”; “All of Us Will Learn”; and “True Champions Always Work Hard.” Other celebrated schools tell children they are loved; KIPP says something more important: self-discipline and civility will get you where you want to go.

The KIPP students I spoke to all described their previous schools as places to which they would never want to return. As one young boy put it: “We have a closer relationship with our teachers here. We all get along. There is no fighting.We learn more. In my other school, the teachers didn’t care that much.” A girl added: “In my old school . . . there were three fights every day. Here you would get in trouble for fighting. Why would a college want someone who doesn’t behave?”

KIPP’s demand that parents sign a contract of cooperation means the school, to some extent, chooses its students. And that choice is probably integral to its success. The many parochial schools that have a track record of success with black and Hispanic children also have students whose parents want them to be there.40 But most of these same youngsters in a different educational setting would not be flourishing academically, it seems safe to say. Schools do make a difference, but they need a minimal level of parental cooperation. Michael Feinberg at Houston’s KIPP believes nine out of ten low-income parents in the community with which he works would sign the contract.41

The American Project

“The moral pulse” beats more strongly in America than in Europe, the great Swedish economist Gunnar Myrdal noted in 1941.42 It was a theme he often returned to. The country is “continuously struggling for its soul,” he wrote in his classic 1944 work, An American Dilemma.43 Myrdal did Americans the honor of taking seriously their commitment to liberty, equality, justice, and fair opportunity. “The Negro problem,” in his view, was a moral issue over which even ordinary Americans brooded in their thoughtful moments. It may not have been true then, more than half a century ago, but it is true now. Myrdal’s was a prescient voice. Today, most Americans care about racial equality. In a land of plenty, they do not want African Americans left out—second-class citizens, a deprived nation within our nation. Racial equality is a moral commitment from which few dissent.

But until the day on which blacks and whites are equally well educated, that equality will remain a dream. The good news is that most eyes are now on K–12 schooling. Can we fix the schools? If KIPP and other such schools (woefully few in number) provide ground for hope, they also deliver a sobering message: most public school systems are an obstacle course that only the most extraordinary principals can negotiate. Closing the racial gap in academic achievement will require radical educational reform against which there will be much resistance. But if racial equality is the persistent American dilemma, it is also the American project. And thus, down a long road, schools will be forced to change.

Notes

1. This information on the number of states instituting reform, etc., is taken from

the American Federation of Teachers’ most recent annual report, Making Standards

Matter, 1999, available only on the AFTWeb site (http://www.aft.org).The report notes

that “historically, states and districts haven’t organized curriculum around a clearly

defined set of expectations, nor have they developed assessment systems that measure

whether students are meeting rigorous, publicly available standards.” The AFT’s 1999

report lists twenty-two states as meeting the organization’s criteria for standards that

are academically solid, a sharp rise from only three the year before.

2. Christopher Jencks and Meredith Phillips, eds., The Black-White Test Score Gap

(Washington, D.C.: Brookings Institution, 1998), pp. 7, 8, figs. 1–4.

3. See Richard J. Murnane and Frank Levy, Teaching the New Basic Skills: Principles

for Educating Children to Thrive in a Changing Economy (New York: Free Press, 1996).

A more detailed technical version of their analysis is available in Richard Murnane,

John Willett, and Frank Levy, “The Growing Importance of Cognitive Skills in Wage

Determination,” Review of Economics and Statistics 77 (1995): 251–66.

4. George Farkas and Keven Vicknair, “AppropriateTests of RacialWage Discrimination

Require Controls for Cognitive Skills: Comment on Cancio, Evans, and

Maume,” American Sociological Review 61 (August 1996): 557–60. This is a critique of

A. Silvia Cancio, T. David Evans, and David J. Maume, “Reconsidering the Declining

Significance of Race: Racial Differences in Early Career Wages,” in the same issue of

the American Sociological Review, pp. 541–56, which had measured education crudely

by levels of schooling completed. The authors’ rejoinder to this critique (ibid., pp. 561–

64) argues that it is inappropriate to control for education by using tests of cognitive

skills because those tests are biased and really only test “exposure to the values and

experiences of the White middle class” (p. 561). Employers are apparently guilty of

class and racial bias if they want employees to be able to read a training manual or to

calculate how many bags of grass seed and fertilizer the customer will need to make a

lawn that is 60 feet long or 40 feet wide.

5. Jencks and Phillips, p. 6.

6. “Reaching the Top: A Report of the National Task Force on Minority High

Achievement,” College Board, 1999, p. 4.

7. Meredith Phillips et al., “Family Background, Parenting Practices, and the

Black-White Test Score Gap,” in Jencks and Phillips, eds., Black-White Test Score Gap,

chap. 4. The entire chapter is an analysis of the PPVT and other data bearing on the

relation between young children’s cognitive skills and family background.

8. Meredith Phillips, James Crouse, and John Ralph, “Does the Black-White Test

Score Gap Widen After Children Enter School,” in ibid., p. 232.

9. Michael A. Fletcher, “A Good-School, Bad-School Mystery: Educators Striving

to Close Racial Gap in Affluent Ohio Suburb,” Washington Post, October 23, 1998,

p. A01.

10. The College Board’s contrast is actually between white and Asians on the one

hand, blacks, Hispanics, and Native Americans on the other. But the performance of

Hispanics (and Native Americans) is a topic for another essay.

11. College Entrance Examination Board, 1995 National Ethnic/Sex Data, unpaginated.

12. The OCR draft guidelines, entitled “Nondiscrimination in High-Stakes Testing:

A Resource Guide,” were first issued on May 17, 1999. They came close to equating

disparate impact and discrimination, intimating that high-stakes tests of any sort could

be used only if the consequences of getting a low score would not be the same for

blacks and Hispanics as for whites and Asians. In mid-December a revised and softened

version was leaked to the Chronicle of Higher Education (Patrick Healy, “Education

Department Softens Its Tone in Latest Draft of Guide to Using Test Scores in Admissions,”

Dec. 13, 1999); as this essay is being written, the full text is not available.

13. Patrick Healy, “Affirmative Action Defenders Push Colleges to Back Draft

Guidelines on Testing,” Chronicle of Higher Education, June 25, 1999, p. A42. Edley’s

ringing defense of the guidelines referred to the harsher version issued in May.

14. Julie Blair, “NAACP Criticizes Colleges’ Use of SAT, ACT,” Education Week,

December 1, 1999, p. 10.

15. William G. Bowen and Derek Bok, The Shape of the River: Long-Term Consequences

of Considering Race in College and University Admissions (Princeton, N.J.:

Princeton University Press, 1998), is a brief on behalf of racial preferences and thus

argues for the admission of black and Hispanic students with academic credentials

weaker than those of their Asian and white peers accepted to highly competitive schools.

Nevertheless, the authors acknowledge the underperformance of black college students

relative to their SAT scores: “The average rank in class for black students is appreciably

lower than the average rank in class for white students within each SAT interval. . . . It

is one strong indication of a troubling phenomenon often called ‘underperformance.’

Black students with the same SAT scores as whites tend to earn lower grades”; p. 77,

italics theirs.

16. News of what ETS called its “Strivers” program surfaced in the Wall Street

Journal on August 31, 1999, and was subsequently picked up by other publications. It

was an unfinished project, although ETS was willing on the record to discuss the central

idea. For more detail, see “The End of Meritocracy: A Debate on Affirmative Action,

the S.A.T., and the Future of American Excellence” betweenNathan Glazer and Abigail

Thernstrom, New Republic, September 27, 1999, pp. 26–29.

17. Deborah Meier, Will Standards Save Public Education? (Boston: Beacon Press,

2000), foreword by Jonathan Kozol, p. xi.

18. Meier, ibid., p. 85.

19. Theodore Sizer, “A Sense of Place,” responding to Meier, ibid., p. 72.

20. Rob Hotakainen, “Calling GraduationTests Harsh,Wellstone Offers Alternative

Plan,” Minneapolis Star Tribune, April 4, 2000, p. 1A.

21. The “respect and confidence” question was asked in a 1999 Gallup poll; the

judgment about “low academic standards” was made in response to a Peter D. Hart

Research Associates 1988 survey; Public Agenda polled employers in 1988. The poll of

black parents is reported in Steve Farkas and Jean Johnson, “Time to Move On:

African-Americans and White Parents Set an Agenda for Public Schools,” A Report from Public

Agenda, 1998, p. 17.

22. Peter Sacks, Standardized Minds: The High Price of America’s Testing Culture

and What We Can Do About It (Cambridge, Mass.: Perseus, 2000), p. 6.

23. Todd S. Purdum, “A ‘Pied Piper’ Is Shaking Up the Schools in Los Angeles,”

New York Times, April 15, 2000, p. A7.

24. “Ticket to Nowhere: The Gap Between Leaving High School and Entering

College and High-Performance Jobs,” in Thinking K–16 (a publication of the Education

Trust) 3, no. 2 (fall 1999): 2.

25. Letter to the New York Times, April 16, 2000, sect. 4, p. 14.

26. Standardized Minds, p. 5.

27. Kate Zernike, “Regents Vote to Rank Schools, Weighing Performance of Minority

Students,” New York Times, May 5, 2000, p. A23.

28. U.S. Bureau of the Census, Current Population Reports, pp. 60–189, Income,

Poverty, and Valuation of Noncash Benefits: 1994 (Washington, D.C.: U.S. Government

Printing Office, 1996), table B-6.

29. See Sara McLanahan and Gary Sandefur, Growing Up with a Single Parent: What

Hurts, What Helps (Cambridge, Mass.: Harvard University Press, 1994).

30. Jencks and Phillips, Black-White Test Score Gap, p. 10 and chap. 4.

31. Between the 1969–70 school year and 1995–96, nationwide expenditures per

pupil (in constant ‘95–’96 dollars) rose from $3,337 to $6,146, a gain of 84 percent. In

the same period, per capita income (in constant dollars) rose by only 46 percent; 1998

Digest of Educational Statistics, p. 183; 1998 Statistical Abstract of the United States, p.

476. Since the passage of the Elementary and Secondary Education Act (ESEA) in 1965,

the federal government has spent $118 billion on making sure poverty ceased to be “a

bar to learning”—although with results no one celebrates. For an excellent discussion

of Title I of the ESEA, see George Farkas, “Can Title I Attain Its Goal?” paper prepared

for presentation at a Brookings Institution conference, May 17, 1999.

32. National Center for Education Statistics, Disparities in Public School District

Spending, 1989–1990, NCES 95-300 (Washington, D.C.: U.S. Government Printing

Office, 1995), p. 15. William N. Evans has also tabulated census data for 1972, 1982,

and 1992 and concluded that the average black student and average white student both

live in districts that spend just under $5,400 per pupil. Evans, Sheila Murray, and

Robert Schwab, “School Houses, Court Houses, and State Houses After Serrano,”

Journal of Policy Analysis and Management, January 1997, pp. 10–31.

33. David J. Armor, Forced Justice: School Desegregation and the Law (New York:

Oxford University Press, 1995), pp. 92–98.

34. On the stability of de facto segregation, see America in Black and White, chap.

12.

35. E. D. Hirsch Jr., The Schools We Need and Why We Don’t Have Them (New

York: Doubleday, 1996), pp. 43, 54. Shelby Steele has made the point with particular

eloquence about low expectationspackaged as racial sensitivity.Even minimal academic

demands, he argues, are often stymied by a deep and painful sense of “racial vulnerability”

that whites and blacks share. The vulnerability of whites is to the charge of

racism, that of blacks is to the claim of inferiority. In different ways both are wracked

by powerful doubts about their own self-worth. See “The Race Not Run,” NewRepublic,

October 7, 1996, p. 23. The Content of Our Character: A New Vision of Race in America

(New York: St. Martin’s Press, 1990) makes the argument at much greater length.

36. On chaos and violence in the schools, see Abigail Thernstrom, “Courting Disorder

in the Schools,” Public Interest 136 (summer 1999): 18–34.

37. Jencks and Phillips, eds., pp. 24, 45, 46. The College Board task force also refers

to the “cultural attributes of home, community, and school” and talks at length about

the attitudes toward school and toward hard work that Asian parents transmit to their

children. “Reaching the Top,” pp. 14, 17–18.

38. Jencks and Phillips, eds., p. 43.

39. All the information on KIPP comes from the author’s three visits to the school,

from the literature the school provides, and from Samuel Casey Carter, No Excuses:

Lessons from 21 High-Performing High-Poverty Schools (Washington, D.C.: Heritage

Foundation, 2000).

40. Scholars have been looking closely at the record of Catholic schools for decades;

there is solid research tracing the sources of their success with inner-city students. See,

for instance, Anthony S. Bryk, Valerie E. Lee, and Paul B. Holland, Catholic Schools

and the Common Good (Cambridge, Mass.: Harvard University Press, 1993); James S.

Coleman and Thomas Hoffer, Public and Private High Schools: The Impact of Communities

(New York: Basic Books, 1987). All the experimental voucher programs are

being carefully watched as well. For a summary of the evidence on vouchers as of 1998,

see the collection of essays in Paul E. Peterson and Bryan C. Hassel, eds., Learning from

School Choice (Washington, D.C.: Brookings Institution, 1998).

41. Personal communication with Michael Feinberg.

42. Quoted in Walter A. Jackson, Gunnar Myrdal and America’s Conscience: Social

Engineering& Racial Liberalism, 1938–1987 (Chapel Hill: University of North Carolina

Press, 1990), p. 151. The quotation was from Kontakt med Amerika, a book co-authored

with Myrdal’s wife, Alva, but Jackson attributes the sections on race solely to Gunnar

Myrdal.

43. Gunnar Myrdal, An American Dilemma (New York: McGraw-Hill paperback

ed., 1964), p. 4.

==============================

Schools That Work for Minority Students

CLINT BOLICK

Hysteria followed the plummeting admission rates of blacks and Hispanics at elite public universities in California and Texas following the curtailment of race and ethnic preferences.1 The concern was appropriate but overdue and misdirected. Critics claimed that the demise of race-based “affirmative action” meant the end of college opportunities for many blacks and Hispanics.2

In reality, the enrollment decline was a long-overdue wake-up call to a dire crisis that seemingly has escaped the attention of many public policy makers and the establishment civil rights groups: the appalling failure of the nation’s public school system to deliver quality educational opportunities to a large portion of America’s schoolchildren. Though concentrated primarily on economically disadvantaged youngsters in the inner cities, that failure manifests itself in disproportionately poor academic credentials for black and Hispanic students. In a nondiscriminatory college admissions process, that disparity leads to lower admissions rates for blacks and Hispanics; by contrast, in a system where blacks and Hispanics are leapfrogged over more qualified applicants into academic institutions for which they are not adequately prepared, it leads to disproportionately higher college dropout rates for the two groups.

No substitute exists for high standards and adequate academic preparation at the elementary and secondary school levels, and we are perpetuating a cruel hoax to assert otherwise. Race-based affirmative action in college admissions is a purely cosmetic response that allows underlying educational problems to fester and grow. An immigrant cabdriver recently distilled the policy dilemma: “They don’t understand that the problem is not in college. The problem is in kindergarten.” Removing the superficial tool of racial preferences from the policy-making arsenal means that policy makers must address at last the core problems that produce racial and ethnic disparities in higher education.

Traditional “civil rights” responses to educational inequality have focused on (1) racial balancing through forced busing or other mechanisms or (2) increased spending. Both approaches have failed utterly to close the gap in educational achievement.3 Fortunately, promising alternatives are appearing on the horizon, focusing not on social engineering but on parental empowerment. By giving parents—who have the greatest stake in their children’s success—greater power over education decisions and resources, it appears we finally can deliver on the sacred promise of equal educational opportunities for all of America’s schoolchildren.4

Educational Crisis and Systemic Failure

It seems impossible that nearly forty-five years have passed since Brown v. Board of Education. During that time, much progress has been made toward erasing the color line from education. As Stephan and Abigail Thernstrom report in America in Black and White, high school graduation rates now are nearly the same for blacks and whites. Progress for blacks in this regard has been explosive: in 1980, only 51.2 percent of blacks over age 25 had graduated from high school; but by 1995, 73.8 percent of blacks over 25—and 86.5 percent of blacks between the ages of 25 and 29—were high school graduates; college attendance rates are up, too, from 21.9 percent of blacks in 1980 to 37.5 percent in 1995.5 Those gains are important because education correlates closely with income. For instance, black women who have graduated from high school, attended some college, or graduated from college all make more money on average than their white female counterparts.6

But from there the news gets bad. Although black high school students steadily were closing the achievement gap with whites in the 1980s, that gap has widened substantially during the past decade. The typical black high school student graduates roughly four academic years behind typical white high school seniors.7 The National Assessment of Educational Progress (NAEP) reported in 1995 that only 12 percent of black high school seniors (compared with 40 percent of whites) were proficient in reading.8 The 1997 NAEP found that while 76 percent of white fourth-graders were proficient in basic mathematics skills, only 41 percent of Hispanic and 32 percent of black students demonstrated basic proficiency.9

The crisis is far more pronounced and debilitating among low-income minority children, who are consigned disproportionately to failing large urban school systems. Students often must pass through metal detectors and literally risk their lives on a daily basis for the chance to obtain a woefully substandard education. Two cities where I have litigated present the problem in especially graphic terms. In Cleveland, the numbers 1 in 14 are emblazoned permanently on my memory: children in the Cleveland Public Schools have a 1-in-14 chance of graduating on time from high school at senior-level proficiency—and an equivalent 1-in-14 chance each year of being a victim of crime in the schools. In the Milwaukee Public Schools (MPS), only 48 percent of the students graduate—a dropout rate more than seven times the statewide average—and only 15 percent of children from families on public assistance graduate. In eleven Milwaukee public high schools that enroll more than three-fourths of the city’s black students, the median grade-point average is less than 1.5 on a four-point scale.

When poor and minority inner-city students have no greater chance of graduating with basic proficiency than of being a victim of crime in the schools, we know the system is failing. With statistics like these, what is surprising is not that minority schoolchildren are admitted to colleges and universities at rates lower than their proportionate share; what is surprising is that the numbers are not much, much worse.

Apologists for the status quo search frenetically for scapegoats: standardized tests are biased, the students are too poor to educate, their parents don’t provide sufficient support, the schools are inadequately funded.10 Yet students from identical socioeconomic circumstances do much better in private schools. A recent study by University of Chicago economist Derek Neal found that although Catholic schools produce negligible academic gains for suburban and white students, they strongly improve educational outcomes for urban minority children. Holding other factors constant, Neal found that the odds of high school graduation for urban black and Hispanic children increase from 66 percent to at least 88 percent in Catholic schools. In turn, he found that three times as many black students who attend Catholic high schools go on to graduate from college. Not surprisingly, those academic gains translate into substantially higher wages in the labor market. Neal concludes bluntly, “Urban minorities receive significant benefits from Catholic schooling because their public school alternatives are substantially worse than those of whites or other minorities who live in rural or suburban areas.”11

Why do suburban public schools and urban private schools do a relatively good job in educating children, while urban public school systems as a whole are failures? That question was addressed in a pathbreaking Brookings Institution study by John E. Chubb and Terry M. Moe.12 They found that although student ability and parental background are important factors in student achievement, school organization also plays a central role. The key differences between effective and ineffective schools are autonomy, parental involvement, and a sense of mission. Urban public schools, Chubb and Moe observed, are characterized by massive bureaucracies that make it difficult for teachers to teach, for parents to exert influence, or for reform to take hold. Moreover, because poor students usually have nowhere else to go, large urban school systems are unresponsive to consumer demands and instead try to satisfy special-interest groups and politicians who control the purse strings. Meanwhile, parents have little influence, particularly on an individual basis. Private and suburban public schools, by contrast, tend to have smaller bureaucracies and to be more responsive to parental concerns, in part because parents have the resources to take their children elsewhere. Chubb and Moe found that effective schools boost student achievement by one year for every four. They concluded that greater parental choice and control over resources are necessary for low-income parents to improve their children’s education and to effectively prod public schools to improve.

The Catalyst: Parental Choice

Those were the goals State Representative Annette Polly Williams had in mind when she proposed the Milwaukee Parental Choice Program, the nation’s first school choice program targeted to inner-city, low-income families. The program initially was modest in scope: only 1 percent of Milwaukee Public School students (roughly 1,000) were eligible to use their share of the state’s education contribution as full payment of tuition in nine participating nonsectarian private schools. But the implementation of the program in the fall of 1990 set off an education revolution. For the first time ever, the program transferred control over public education funds from bureaucrats to parents and forced the public schools to compete for low-income youngsters and the resources they commanded. Predictably, the program prompted litigation by the teachers’ union and a blizzard of regulations designed to destroy the program by bureaucratic strangulation. Both counterattacks were beaten back.13 So in the fall of 1990, nearly one thousand low-income youngsters were able to cross the threshold to a brighter educational future in a dozen or so nonsectarian community private schools.

The assessments by the state’s designated researcher, John Witte, produced odd findings over the program’s first four years: parental involvement was strong, satisfaction was high, but student achievement failed to rise. Those results were seemingly contradictory, given that parental involvement and student achievement are closely linked. That conundrum was magnified by Witte’s refusal to release data to other researchers.14 The confusion dissipated when a team of researchers led by Harvard University political scientist Paul E. Peterson finally obtained the data. For the first time, they compared achievement of students who had gained access to private schools through the random selection process with students who had not. The result: little academic change over the students’ first two years in the program, but significant progress in the third and fourth years. Peterson found that over its first four years, the program narrowed the gap between minority and nonminority test scores by between one-third and one-half—an absolutely momentous accomplishment. 15

Zakiyah Courtney, director of Milwaukee’s Parents for School Choice and former principal of Urban Day School, testified:

I was glad to see . . . a study that reflected what many of us who have been working directly with the families and the children all along knew. And that is that parental satisfaction does make a difference; and that oftentimes you may not see those high achievement scores in the beginning, but if you give those children the opportunity to stay and work in the program, that you do see those differences.16

But apart from its benefits for kids for whose parents now can choose better schools, the program unquestionably has also had a positive impact on the public schools, forcing them to pursue long-overdue systemic reforms. At an evidentiary hearing on the program in 1996, both former MPS superintendent Dr. Howard L. Fuller, and his successor, Robert Jasna, agreed that the program had created a prod for long-overdue systemic reforms.17 Fuller supported the program’s expansion in 1995 to increase the number of eligible children to 15,000 and to allow religious schools to participate. Fuller explained:

I think that what it will bring into play would be, in addition to the existing schools, there will be new schools out there that will come into being that will find their niche to begin to teach kids that we are not currently reaching or that we’re losing. I think it will begin to give poor parents some capacity to have leverage over this entire discussion. And the reason they will have leverage is because they will begin to have leverage over resources, the same type of control over resources that people with money have. . . . You begin to create a synergism for change that I think is key, if the system is going to be changed, so that we . . . save these kids that we’re losing each and every day.18

The Milwaukee program’s expansion was enjoined before it could go into operation in the fall of 1995 as a result of litigation brought by the teachers’ union, the American Civil Liberties Union, the National Association for the Advancement of Colored People (NAACP), and others, who asserted that the program violated the prohibition of religious establishment in the state and federal constitutions. (Similar lawsuits are pending against school choice programs in Ohio, Vermont, Arizona, Maine, and Pennsylvania.) In June 1998, the program was upheld by the Wisconsin Supreme Court, setting up possible resolution of the constitutional issue by the U.S. Supreme Court.19 In the fall of 1998, the expansion of the Milwaukee program finally commenced, with an estimated 6,000 lowincome youngsters attending more than eighty private religious and nonsectarian schools.20

Meanwhile, a second school choice program championed by Governor George Voinovich and City Councilmember Fannie Lewis was created in 1996, providing $2,500 scholarships to approximately 3,000 economically disadvantaged children, allowing them to attend private secular and religious schools in Cleveland. Again, early results appear promising.21 Similar promising results are also reported in privately funded scholarship programs serving low-income schoolchildren in dozens of other cities across the country.22 As Paul Peterson has observed, “The choice movement is spreading in good part because its theoretical underpinnings seem more powerful than ever.”23

Other forms of parental choice are blossoming. Arizona, which boasts the nation’s most wide-ranging charter school system, approved a $500 state income tax credit for contributions to private scholarship funds, which was upheld by the Arizona Supreme Court.24 In May 1999, a bipartisan majority of the Florida Legislature enacted Governor Jeb Bush’s A public education reform program, within which parental choice is an important element. The program creates a grading system for all public schools in the state, provides financial rewards for excellent schools, gives extra help to students in failing schools, and allows families whose children are in failing schools to opt out into better public schools or private schools. In essence, the program offers the first money-back guarantee by creating both choice and competition that should improve public education for everyone. It emphasizes that the proper concern of public education is not where children are educated but whether they are educated. The threat of scholarships also is spurring spirited efforts among public school districts to improve schools that are receiving “F” grades.25

Still, a school choice program for the District of Columbia, approved in 1998 by bipartisan majorities in both houses of Congress, was vetoed by President Clinton, in spite of support from a large majority of residents, particularly blacks. Powerful special interest groups have combined to stifle parental choice all around the nation. The question is how long those defenders of the status quo can delay the day of reckoning with America’s most urgent crisis—and how many children’s lives and educational opportunities will be sacrificed in the process.

Parental Choice and Public Opinion

Public opinion is moving strongly and steadily in favor of parental choice. There are several possible explanations: (1) In spite of massive financial resources and constant excuses from the education establishment, the academic performance of public schools, even in affluent suburbs, is declining, particularly compared with schools in other industrialized nations. (2) The news about parental choice programs such as school vouchers and charter schools is generally promising; reality is debunking the fears raised by choice opponents. (3) Changing demographics are influencing public opinion: young people with children are used to making choices and are comfortable about selecting from an array of educational options.

Whatever the explanation, the trend is unmistakable. For the first time, the 1998 Phi Delta Kappa–Gallup poll found that a majority of Americans support parental choice. Asked whether they favor allowing parents to send their school-age children to any public, parochial, or private school of their choice with the government paying part or all of the tuition, 51 percent of all respondents support choice, while 45 percent are opposed. Only two years ago, the same idea was opposed by a margin of 54 to 43 percent. Among the groups who most strongly support parental options are nonwhites (68 percent), people 18–29 years of age (63 percent), and, notably, public school parents (56 percent).26

Support for parental choice is even greater in Wisconsin, which has had eight years of experience with the Milwaukee program. A 1998 poll by Louis Harris & Associates found that Wisconsin residents support parental choice by a margin of 61 to 32 percent.27 Parental choice draws majority support in all areas of the state and from whites, blacks, men, women, Republicans, Democrats, independents, conservatives, and liberals. The margin in Milwaukee is 65 to 29 percent, and residents back the Milwaukee Parental Choice Program specifically by an even greater 71 to 24 percent. The closer people reside to the program, the more likely they appear to support parental choice.

What is perhaps most remarkable,and disturbing, is the chasm between mainstream minority individuals and establishment civil rights organizations on issues over parental choice. Black Americans consider education the top national priority.28 That concern is well placed. A 1997 survey by the Joint Center for Political and Economic Studies found that while 37.5 percent of white parents considered their schools “fair” or “poor,” 64 percent of blacks and 61 percent of Hispanics gave their schools the same low grades. Not surprisingly, the poll found that while support for school choice was evenly split among whites, it was strong among both blacks (56 to 37.5 percent in favor) and Hispanics (65 to 29 percent in favor). Support for school choice was most intense among those in the age bracket most likely to have school-age children: 86.5 percent of blacks between ages 26– 35 back school choice, with only 10 percent opposed.29

Similarly, aWashington Post poll found that a large majority of District of Columbia residents backed the low-income scholarship program passed by Congress but vetoed by President Clinton and opposed by liberal black politicians such as Delegate Eleanor Holmes Norton.30 The residents supported the scholarship program by a margin of 56 to 36 percent. While whites and blacks with incomes over $50,000 split fairly evenly over the issue, lower-income blacks favored the legislation by a margin of 65 to 28 percent. “I would jump at the chance to send my son to private school,” said Janice Johnson, who lives in one of the poorer sections of the city.

Meanwhile, 1,001 low-income children from among more than 7,500 applicants received support from the Washington Scholarship Fund, which is financed by businessmen Theodore J. Forstmann and John Walton.31 In light of strong minority support for school choice, there is little wonder that establishment civil rights groups such as the NAACP are struggling for relevancy. The MilwaukeeNAACP chapter joined the lawsuit against the Milwaukee Parental Choice Program in spite of overwhelming support for the program among black city residents. The national NAACP last year announced an unholy alliance against parental choice with People for the American Way. Parental Choice is “exploitative of the black community,” contends Mary Jean Collins, PAW’s national field director. “The philosophy of the right is always, ‘Give my kid what he wants and to hell with the rest.’ For that attitude to get into the black community would be shameful.”32

Dissenters such as the Urban League of Greater Miami and former Atlanta mayor and U.N. ambassadorAndrewYoung reject such patronizing attitudes and support school choice as an essential component of civil rights. Former representative Floyd Flake, whose church in Queens, New York, operates a private school, says: “When a white person kills a black person, we all go out in the street to protest. But our children are being educationally killed every day in public schools and nobody says a thing.”33 It is time for politicians to recognize the will of the people, to reject the entreaties of special-interest groups, and to make parental choice a reality.

The Broader Context of Education Reform

Never has the climate for reform been so vibrant—nor the need for reform so urgent. In addition to parental choice encompassing private schools, promising reforms include (1) charter schools, which are autonomous public schools often operated by private or nonprofit entities; 34 (2) contracting out public schools to private management firms such as the Edison Project;35 and (3) tax deductions and credits that allow people to deduct their children’s education costs or to contribute to privately funded scholarship funds.36

Meanwhile, private philanthropy is working to meet demand from low-income parents to secure better education for their children through programs like the Children’s Scholarship Fund and CEO America. In April 1999, 1.25 million children applied for 40,000 CSF scholarships. Andrew Young likened the outpouring to Rosa Parks’s refusal to give up her bus seat and to Martin Luther King Jr.’s letter from a Birmingham jail. Declared Young:

If families were allowed to seek a quality education wherever it may be found, who would benefit? Simple: Those who aren’t getting a quality education and those who can deliver it. Certainly, some will oppose competition— just as AT&T once fought the breakup of its monopoly. Others will reflexively resist the redistribution of power to poor families. Still others will wave their worn-out ideologies to defend a system of educational apartheid while demonizing anyone who promotes a parent’s right to choose. . . . I predict that we will one day look back on the 1.25 million who applied for educational emancipation—for the chance to seek the light and oxygen of a nourishing education—not as victims, but as unwitting heroes with whom a great awakening was begun.37

Parental choice is a central facet of systemic education reform that places equal resources behind each child and allows the funds to follow children to whatever schools—public, private, or religious—their parents choose.38 Child-centered education funding transforms the focus of public education from public schools as ends in themselves to publicly funded education in whatever schools parents deem best. If parents choose public schools, the funds stay in those schools under the control of the schools themselves, rather than filtering down through an education bureaucracy. Only through a system of choice, competition, and accountability where parents are sovereign will public schools in the inner cities begin to improve. If we can do only one thing in public policy to improve prospects for minority individuals and economically disadvantaged people, there is nothing more tangible or important than making good on the promise of equal educational opportunities. As I have acknowledged before, if I were given the option, straight up or down, of abolishing racial preferences or adopting parental choice on a nationwide basis, in a heartbeat I would opt for the latter, for it would reduce the pressure for divisive race-based solutions. Unfortunately, no one is offering that choice: the same reactionary leaders who support racial preferences are blocking the schoolhouse doors for the very people whose interests they falsely claim to represent.

That won’t last long. Nothing that Jesse Jackson, Kweisi Mfume, Bill Clinton, Al Gore, or others like them can say to inner-city parents will convince them not to pursue educational opportunities their children desperately need. Milwaukee parent Pilar Gonzalez makes that plain: “I will find a way to have my children attend private school even if it means less food on the table. A quality education for my children is that important.”39 That is the primary civil rights goal of the millennium: making it possible at last for Pilar Gonzalez and millions of others like her to secure the best possible education for their children.

Notes

The author expresses appreciation to Daron Roberts, a University of Texas student

who interned at the Institute for Justice in the summer of 1998, for research assistance.

1. The end of racial preferences in California was attributable first to a decision

to curb preferential admissions policies by the Regents of the University of California,

which subsequently was extended by the California Civil Rights Initiative (Proposition

209) to all public postsecondary schools. In Texas, the cessation of preferences was due

to the decision striking down preferential admissions at the University of Texas School

of Law in Hopwood v. Texas, 78 F.3d 932 (5th Cir.), cert. denied, 116 S.Ct. 2850 (1986).

2. In any event, that prediction has now definitively been discredited. See, e.g.,

James Traub, “The Class of Prop. 209,” New York Times Magazine, May 2, 1999, p. 44.

3. For an examination of the effects of forced busing, see David J. Armor, Forced

Justice: School Desegregation and the Law (New York: Oxford University Press, 1995).

For a discussion of the lack of correlation between increased educational performance,

particularly in the context of the lavish Kansas City desegregation decree, see Blake

Hurst, “Runaway Judge,” American Enterprise, May–June 1995, pp. 53–56.

4. In Brown v. Board of Education, 347 U.S. 483, 493 (1954), the U.S. Supreme

Court declared that education, “where the state has undertaken to provide it, is a right

which must be made available to all on equal terms.”

5. Stephan Thernstrom and Abigail Thernstrom, America in Black and White: One

Nation, Indivisible (New York: Simon & Schuster, 1997), pp. 190–91, 192.

6. Ibid., p. 445.

7. Ibid., p. 355.

8. See Dennis Kelly, “Kids’ Scores for Reading ‘In Trouble,’” USA Today, April

28–30, 1995, p. 1A.

9. See “U.S. Students Make Progress in Math,” Dallas Morning News, February 28,

1997, p. 6A.

10. In fact, like many other large urban school districts, the Milwaukee and Cleveland

districts spend more per pupil than the statewide averages.

11. Derek Neal, “The Effects of Catholic Secondary Schooling on Educational

Achievement,” Journal of Labor Economics 15: 100. Neal’s findings echo similar studies

from the 1980s. See James Coleman, Thomas Hoffer, and Sally Kilgore, High School

Achievement: Public, Catholic, and Private Schools Compared (New York: Basic Books,

1982); and Andrew Greeley, Catholic High Schools and Minority Students (London:

Transaction, 1982). Recent evidence indicates that Hispanic students, too, fare better

in Catholic schools. See Anne-Marie O’Connor, “Many Latinos Fare Better in Catholic

Schools,” Los Angeles Times, August 3, 1998.

12. John E. Chubb and Terry M. Moe, Politics, Markets, and America’s Schools

(Washington, D.C.: Brookings Institution, 1990).

13. See Davis v. Grover, 480 N.W.2d 460 (Wisc. 1992).

14. Witte subsequently has concluded that the results of limited school choice are

favorable. See John F. Witte, The Market Approach to Education:An Analysis of America’s

First Voucher Program (Princeton, N.J.: Princeton University Press, 2000).

15. Jay P. Green, Paul E. Peterson, and Jingtao Du, The Effectiveness of School Choice

in Milwaukee: A Secondary Analysis of Data from the Program’s Evaluation (Cambridge,

Mass.: Harvard University John F. Kennedy School of Government, 1996).

16. Transcript of Evidentiary Hearing, Jackson v. Benson, No. 95-CV-1982 (Dane

County, Wisc., Circuit Court, Aug. 15, 1996), p. 77.

17. Ibid., pp. 48, 165.

18. Ibid., p. 51.

19. Jackson v. Benson, 578 N.W.2d 602 (Wisc. 1998).

20. See Jon Jeter, “As Test of Vouchers, Milwaukee Parochial School Exceeds Expectations,”

Washington Post, August 31, 1998.

21. See Jay P. Greene, William G. Howell, and Paul E. Peterson, “Lessons from the

Cleveland Scholarship Program,” in Paul E. Peterson and Bryan J. Hassel, eds., Learning

from School Choice (Washington, D.C.: Brookings Institution, 1998), p. 357.

22. See, e.g., R. Kenneth Godwin, Frank R. Kemerer, and Valerie J. Martinez,

“Comparing Public Choice and Private Voucher Programs in San Antonio,” in ibid.,

p. 275; and David J.Weinschrott and Sally B. Kilgore, “Evidence from the Indianapolis

Voucher Program,” in ibid., p. 307.

23. Paul E. Peterson, “School Choice: A Report Card,” in ibid., p. 8.

24. Kotterman v. Killian, 972 P.2d 606 (Ariz. 1999).

25. Carol Innerst, Competing to Win: How Florida’s A+ Plan Has Triggered Public

School Reform (2000). (This report was copublished by five groups.)

26. The 30th Annual Phi Delta Kappa–Gallup Poll of the Public’s Attitudes Toward

the Public Schools (1998). The poll surveyed attitudes about a wide range of parental

choice alternatives. The pollsters found that one proposal championed primarily by

Republicans—tax credits for private school tuition—actually is supported by a higher

percentage of Democrats (by 61 to 37 percent in favor) than by Republicans (57 to 42

percent in favor).

27. Louis S. Harris&Associates, Inc., “Wisconsin Residents Strongly FavorVoucher

System and Choice,” news release, August 17, 1998.

28. A survey conducted November 5–7, 1996, by the Polling Company found that

a large plurality of blacks (42 percent) considered education the top national priority,

compared with 25 percent of whites. Fighting crime and drugs was the second top

priority for blacks (21 percent).

29. See David A. Bositis, 1997 National Opinion Poll: Children’s Issues (Washington,

D.C.: Joint Center for Political and Economic Studies, 1997).

30. See Sari Horwitz, “Poll Finds Backing for D.C. School Vouchers,” Washington

Post, May 23, 1998, p. F1.

31. Debbi Wilgoren, “1,001 D.C. Students Win Scholarships,” ibid., April 30, 1998, p. B1.

32. Quoted in Samuel G. Freedman, “The Education Divide,” Salon Magazine,

September 30, 1997.

33. Quoted in ibid.

34. Several dozen states permit charter schools, although some jurisdictions (such

as Arizona, Michigan, and the District of Columbia) provide for far greater autonomy

than others. See, e.g., Clint Bolick, Transformation: The Promise and Politics of Empowerment

(Oakland, Calif.: Institute for Contemporary Studies, 1998), pp. 53–60; Bruno

Manno, Chester E. Finn Jr., and Louann A. Bierlein, “Charter Schools as Seen by

Students, Teachers, and Parents,” in Peterson and Hassel, eds., p. 275; and Bryan J.

Hassel, “Charter Schools: Politics and Practice in Four States,” in Peterson and Hassel, p. 249.

35. See, e.g., John E. Chubb, “The Performance of Privately Managed Schools: An

Early Look at the Edison Project,” in Peterson and Hassel, eds., p. 213.

36. In addition to the Arizona state income tax credit for contributions to private

scholarship funds, Minnesota provides tax deductions for school tuition, and both

houses of the Illinois legislature passed a tax credit for tuition in 1999.

37. Andrew Young, “Let Parents Choose Their Kids’ Schools,” Los Angeles Times,

April 29, 1999, p. 9.

38. The concept of child-centered education funding is championed most prominently

by Arizona Superintendent of Public Instruction Lisa Graham Keegan. See, e.g.,

Clint Bolick, “Charter Reformer: Arizona’s Superintendent of Schools Points theWay

to an Education Revolution,” National Review, April 6, 1998, pp. 42–44.

39. Quoted in Bolick, Transformation, p. 34. Ms. Gonzalez is one of the parents

defending the constitutionality of the Milwaukee Parental Choice Program, and I am

proud to represent her.

==============================

Preferential Admissions in Higher Education

MARTIN TROW

The historical and legal development of racial preferences in higher education, as well as their impact, have been extensively analyzed.1 This essay approaches the issue from a somewhat different vantage point, exploring how race-based preferences operate as they are actually administered in a university setting. Universities are little affected by Supreme Court decisions, the state of race relations generally, or broad public opinion about “affirmative action” and the extent of discrimination in the larger society. On campus, the leading players are the presidents, chancellors, deans, department chairmen, and the affirmative action bureaucracy that has emerged in response to the need to implement various affirmative action laws and regulations. In describing the picture inside the academy, I draw chiefly on material gathered in the course of watching the dramatic events surrounding the public policies affecting university admissions in California since July 1995, when the University of California Board of Regents voted to prohibit the use of race or ethnicity in admissions and appointments.

It is important to note that the issue of racial preferences in admissions to universities and colleges arises only in those institutions that have more applicants than they can admit. This group includes no more than a hundred or so out of the 3,700 colleges and universities in the United States, but it consists of the leading research universities, both public and private, as well as a few dozen elite private liberal arts colleges that function as feeder institutions for graduate and professional schools. The private research universities—Yale, for example—and elite colleges like Amherst and Swarthmore typically choose their students for admission largely on the basis of academic performance and promise, although decisions are made on a case-by-case basis. To secure greater “diversity,” they look with special favor on applicants from certain “underrepresented” racial and ethnic groups. By contrast, the big public research universities admit most undergraduates according to general criteria and formulas. In seeking to secure a racially and ethnically diverse student body, they must specify exactly how much weight is to be given to racial and ethnic considerations.

Admissions Procedures at the University of California

The baseline formula for admission to the University of California has been prescribed by the state’s Master Plan for higher education, virtually unchanged since it was drawn up in the 1950s. It defines “eligibility” for admission to the university as ranking in the top 12.5 percent in academic achievement of graduates of California public high schools. The top 12.5 percent is identified by a formula that combines the student’s high school grade point average and his or her combined math and verbal SAT scores. In addition, the Regents have allowed each of the eight campuses that make up the system to admit up to 6 percent of their entering classes without regard for the 12.5 percent standard. This arrangement for admission by “special action” permitted many athletes who were not top students to enter, as well as students with unusual forms of preparation or talent— for example, outstanding musicians and those who were educated at home or abroad. Before the statewide ban on racial preferences in admissions, many of these discretionary places were used to admit students from “underrepresented” racial and ethnic groups who failed to qualify under the regular requirements.

In the 1980s and early 1990s, a number of the UC’s eight “general campuses” found that they had more, and in some cases many more, eligible applicants than they could admit. The solution was to allow each individual school to adopt a policy somewhat independent of the statewide criteria, out of which came explicitly preferential policies. The formulas differed on the several campuses. With the approval of the Regents, Berkeley decided to admit half the entering class on the basis of their scores on the combined SAT/GPA. In filling the other half, all eligible minority applicants were automatically admitted. But critics charged the school with having adopted a quota or “set-aside” for minority applicants, and in 1989 it abandoned the guarantee of admission for all eligible applicants from minority groups. Yet it continued to give a huge advantage to black and Hispanic applicants. The average black student admitted, for example, had SAT scores 250–300 points lower than his or her white and Asian classmates and a substantially weaker high school grade record as well.

These students were not necessarily “disadvantaged” by any nonracial criteria. Berkeley, for example, was especially attractive to the sons and daughters of the new black middle class; in 1995, 30 percent of black undergraduates came from families earning over $70,000. It was a troubling fact. Admissions decisions, especially in public institutions, must appear as legitimate to those not admitted to their campus of choice, the parents of future applicants, and the general public. The legitimacy of the institution, supported by taxpayer dollars, depends on criteria for entrance that are perceived as fair. And yet most Americans do not believe in judging citizens on the basis of the color of their skin.

Those convictions have prevailed in California. The Regents’ vote to ban preferences in 1995 and the passage of Proposition 209 a year later brought a halt to all race-based admission procedures. Today, UC Berkeley is developing a selection policy based on a reading of all applicant files rather than a statistical formula, a more individualized system like those used by private universities.

Fair Criteria

Racial and ethnic preferences, where they are employed, rest on the assumption that color is the central feature of a person’s identity— that character, intelligence, energy, initiative, socioeconomic circumstance, and other qualities are all less important. And thus in California both economically privileged and low-income black and Hispanic students were equally eligible for preferential treatment, whereas an Asian immigrant struggling with the language and trying to support indigent parents got no break for “disadvantage.” Similarly, whereas a white youngster from a single-parent, low-income household was treated as a privileged Anglo, neither the son of a ruling family in a South American country nor the student whose remote ancestors came from Spain was asked to meet regular academic criteria. The whole Hispanic category as defining identity was particularly troubling. For inexplicable reasons, European Spaniards obviously count. But does the term also include Portuguese and Brazilians? Hispanic was not the only troubling category. The term Asian covers a wide variety of groups with little in common: old and new Chinese immigrations, Japanese, Thai, Vietnamese of various kinds, Indonesians, Koreans, among others. Crudely lumped together they outnumber whites at UCLA and Berkeley—the most selective campuses—raising questions about their “overrepresentation.” In April 1995, in Sacramento, President Clinton warned that “there are universities in California that could fill their entire freshman classes with nothing but Asian Americans.”2 The Asians had become yesterday’s Jews: an allegedly too-ambitious group eligible for exclusion on the basis of their national origins. Ironically, some of the “Asians” who were kept out in order to make room in the university for more blacks and Hispanics were the grandsons and granddaughters of the Japanese Americans who were confined to relocation camps during World War II.

Preferential admissions policies have a powerful and inherent tendency to reduce people in all their variety and complexity to their racial or ethnic identity. And yet higher education is supposed to enhance our sense of individuality, to encourage and educate distinctive qualities of mind and character. Intelligence and creativity, if allowed, burst through the constraints of social origin; nurtured by our origins, we transcend them. In our private lives we may choose to honor and celebrate the culture of a group to which we feel we belong. Or we may reject that identification. People differ, and must be allowed to differ, but the choices they make should be matters of private, not public, policy.

Counting the “Disadvantaged”

The policy of group preferences forces impossible choices on the rapidly growing number of Americans from multiracial backgrounds. The university, in effect, says to students of mixed race, choose between your mother and your father. If you decide to identify yourself as a member of a preferred group, you increase your chances of being admitted and of receiving substantial financial support. If you identify yourself as Asian, even though your father is black, you are less likely to get in. If you choose to check the box labeled “other” or refuse to choose at all, we will simply treat you as if you were white. Which parent are you prepared to reject? There is an additional problem: the whole process invites fraud— difficult to discern by large impersonal institutions like the University of California. One anecdote illustrates the form it takes. A university officer observed a student filling out his application form and checking the box labeled “Hispanic.” “Oh, you’re Hispanic,” he said, trying to make friendly conversation. “No,” came the reply. “Actually I’m Iranian, but my teacher told me to check Hispanic if I wanted to get into Berkeley.”

Although no one knows how often such advice was given, such stories were frequently told with a cynical chuckle when the University of California was still giving preferences. They led to a general cynicism about the fairness of admissions procedures and to a broader sense that the whole enterprise was a racket. I once asked an admissions officer about the problem. “We are not in the business of enforcing Nuremberg Laws,” he said indignantly. The university (rightly) was unwilling to decide whether a student one-eighth black was indeed “black” and who would determine, and by what means, just how large the fraction actually was. But without clear criteria and a settled process for determining fraud, no penalties for false representation could be imposed. And UC was certainly not prepared to set such criteria and procedures. But without the danger of identification or penalty, a “victimless crime” that carried substantial benefits was increasingly attractive, at least to high school teachers and counselors.

How Much Preference Is Enough? The Case of Boalt Hall

Racial preferences are ubiquitous in selective institutions of higher education, except where they have been banned, as in California, Texas, and the state of Washington. But no one ever defines precisely the point at which “diversity” has been reached. How large a preference should be given to reach what goal? Here again some evidence from the University of California may be enlightening, this time from a professional school. Until the Regents acted to end race-driven admissions, the policies adopted by UC schools were said to be guided by the Supreme Court’s 1978 decision in University of California v. Bakke, which outlawed quotas but permitted institutions to consider race as “one factor in admissions.” What did this mean, exactly? Some schools read the Court’s majority as having said, “other things being equal,” race could be a basis for selection at the margin. Most, though, saw the decision as a license to achieve diversity by giving heavy weight to race and ethnicity.

Berkeley’s Law School, Boalt Hall, was among those that interpreted Bakke permissively. Unlike most other graduate departments and professional schools, the admissions process used a formula that placed applicants into one of four “Ability Ranges,” A through D, on the basis of undergraduate GPAs and LSAT scores. In 1996, only 855 students were admitted to Boalt, out of 4,684 who applied, but the rate of acceptance for black and Hispanic students in every “Ability Range” below the top one A was much higher than that for whites and Asians.

For example, eighteen applicants from the preferred groups fell into the top two ranges, and all but one of them were accepted. Almost all applicants from all ethnic groups in the A range were admitted, but among those who fell in the B range, 69 percent of Asians, 62 percent of whites, and 94 percent of blacks and Hispanics were admitted. Looking at range C, only 19 percent of Asians and 17 percent of whites were admitted, while 77 percent of the blacks and Hispanics got in. In the lowest range, the disparities were even greater.

When we look at specific ethnic groups, the differences are even more striking. Of applicants in Ability Range C, ten were students of Japanese origin, and ten were black. All the blacks and none of the Japanese were accepted. In range D, a significant number of blacks but no students of Chinese origin were admitted.

These disparities were the consequence of highly race-conscious admissions processes, and when Boalt Hall’s freedom to engage in race-based admissions was curtailed, the drop in blacks admitted was dramatic. Just one black student would be entering the first-year class in 1997, the media widely reported, and that student had deferred admission after having been accepted previously, when preferences were still permissible. The end of preferences, it was said despairingly, meant the virtual end of African American students at the University of California’s top law school.

The figures for Boalt were reported before those for other professional and graduate programs had been made available, and the press presented them as the sign of things to come. In fact, Boalt Hall, together with other UC law and medical schools, was distinctive. The proportion of blacks and Hispanics entering Berkeley’s graduate programs in general was little changed by the change in admissions policy—news that was basically ignored in the media.

Between 1996 and 1997—before and after preferences—the number of African Americans in UC graduate schools increased by 2 percent, while the “other” and “declined to state” categories increased by 25 percent, the Office of the President reported. (Those figures excluded the professional schools.) One might have imagined that the UC administration would be eager to spread the good news, but there were no press conferences or statements by senior administrative officers calling attention to these surprising figures. Instead, they gave the Boalt Hall story maximum publicity. Boalt was different from most other graduate and professional programs in one important respect: it was competing for the most academically able black and Hispanic applicants with other leading law schools, especially those at the top private research universities. Unfortunately, however, it had never been very successful in that competition. Long before the Regents’ vote and Proposition 209, Boalt had consistently lost all or almost all its best-qualified non-Asian minority applicants to Harvard, Yale, Stanford, and other law schools that were both more prestigious and could offer more financial support and more attractive job prospects upon graduation. But when no minority students enrolled in Boalt’s 1997 entering class, public statements by administrators inside and outside the school blamed the results on the change in admissions criteria.

The end of race preferences at UC had dramatic effects on Boalt, in other words, because the school had gone very far down the road of racial double standards in an effort to compensate for a competitive disadvantage against more prestigious and affluent schools. But in 1997,3 it could no longer take large numbers of poorly qualified minority students over white and Asian students with higher GPAs and LSATs. By contrast, most other graduate and professional programs within the university system had not had to change their admissions criteria and practice so drastically after preferences were abolished, and their black and Hispanic enrollment was almost unaffected.

In fact, the University of California has an advantage over most other universities in being a system of eight universities whose central office keeps records of their enrollments year by year and by the old ethnic-race categories. What those records show is that after the passage of 209 the non-Asian minority students who had applied but had not been admitted to Berkeley and other UC campuses that had been using race and ethnicity as criteria for admission tended to enroll in one of the other UC campuses, which were admitting all applicants who were academically qualified for admission to the university. The result of this process of “cascading” is that by 1999, only two years after the first class was admitted color blind, the University of California as a whole showed a decline of only 3 percent in the numbers of non-Asian minority freshmen enrollees. Even that figure overestimates the decline because it does not count the minority students who refused to give their race or ethnicity now that it no longer affected their admission to the university. But 3 percent is a far cry from the predictions of the effects of Proposition 209 on minority enrollments in the University of California. By 2000, the number of non-Asian minorities newly admitted to UC was already higher than in the last year that preferences were still in effect. The numbers admitted who refused to give their ethnicity had also risen.

Consensus and Coercion

Senior administrators in the academy are solidly behind racial preferences, and they are the seemingly united voice of almost every university. But in fact there is considerable division among members of the faculty. A 1995 Roper poll asked voting members of the University of California Academic Senate which they favored: granting preferences to women and certain racial and ethnic groups, or promoting equal opportunity without regard to an individual’s race, sex, or ethnicity. A wide plurality (48 percent) chose the latter policy; only 31 percent supported preferences.4 Another survey, of a national sample of college teachers, came up with much the same result.5

These findings are consistent with American public opinion in general, which has been deeply divided on the issue for three decades. What explains this curious combination of outward consensus and internal division in the academy? A mixture of principle and pragmatism—sincere belief coupled with a keen sense of who on or around campus can make trouble— explains the unanimity among university administrators. The officials who administered UC’s large affirmative action bureaucracies were appointed on the understanding that they would support the preferential regime, and most of them probably did so genuinely. But they also know that those who are opposed to race-conscious admissions will seldom speak out and will never lead marches or sit-ins, whereas those who profit from preferences are combatants. In part, the silence on the part of opponents is a consequence of intimidation. Very few academics wish to offend both the senior administrators who govern their careers and budgets and the wellorganized affirmative action pressure groups that will quickly stereotype faculty members as “racists” or, at very least, “right-wingers.”

A distinguished federal judge who is familiar with academia has summarized the scene well:

Groups holding considerable power in the university loathe speech with the wrong content about topics important to them, and . . . those who say the wrong things will have little peer or institutional protection. . . . Many ideas may not be expressed, many subjects may not be discussed, and any discussion on matters of political salience has to avoid defending groups powerful in the university.6

The problem at the University of California went beyond a lack of institutional protection for dissenting voices. Like other universities, UC over the years developed not only a strong climate but also an organizational structure in support of preferences from the president’s office down. Every campus had administrative offices and academic senate committees to plan and enforce preferential policies; every department had an “affirmative action” officer to monitor its behavior. Needless to say, there was no equivalent organization of people and energy inside the university devoted to criticizing the preference policies or trying to reform them.

The pattern of consensual coercion brands dissenting points of view as illegitimate and deprives those who hold them of protection. Students and other faculty members need not read or listen to their arguments. Already discredited, opponents of racial preferences are then demonized. They are not merely mistaken; they are evil—and fair game for late-night calls and hate mail. But coercion does not need to reach the point of anonymous name-calling to be effective. The administrative unanimity behind the policy can itself have a chilling effect that is enough to stifle debate.

What Proposition 209 Hath Wrought

The Regents’ actions and Proposition 209 are slowly liberating UC campuses from the atmosphere of coercion. Although the president and the chancellors and their senior staffs may not have changed their views, it is now possible, indeed even necessary, for people to talk about how to admit students in ways that might preserve and enhance diversity without allowing race and ethnicity to drive admissions decisions. California voters and the Regents have forced important reforms. The university has had to abandon its categorical formulas and admit students by inspecting their folders rather than simply their scores and race. Some students have always been admitted on the basis of their individual qualities and promise, but the process has now been extended more broadly.

In implementing race-neutral policies, admissions officers need not be blind to inequalities in American society and their impact on academic performance and life prospects. California admissions officers can still give a break to students who, though their scores are a bit low, seem highly motivated and come from mediocre high schools. And they can pay special attention to the high school senior who shows academic promise, although he or she has been struggling with English as a second language.Weighing such disadvantages is quite different from racial and ethnic preferences.

The process requires schools to look at individual qualities; they cannot make stereotypical assumptions about group characteristics. And, as a result, all students admitted to the University of California today know they have earned a place in their class under uniform standards. None needs to feel like a second-class citizen, brought in to keep the “diversity” numbers up. Preference’s burden of guilt and resentment has been lifted from UC students; one can feel it in a classroom.

But Proposition 209’s greatest contribution has been its indirect impact on the lives of students before they apply to college. With racial preferences abolished, California residents can no longer ignore the problem of too few blacks and Hispanics academically prepared for seats in highly selective schools. Racial and ethnic preferences in higher education simply masked the inadequacies of California schools; elementary and secondary education must change, the university finally understands. In fact, with the help of a recent grant of $40 million from the state legislature, each UC campus now has an opportunity to create a program for improving the quality of K–12 schooling. With better primary and secondary education, the number of black and Hispanic youth who are prepared and motivated to continue their education in colleges and universities should expand. And they will be going to colleges that have looked at their talents and aspirations, not at their skin color or national origin, in admitting them. By many criteria the abolition of preferences for admission to the University of California has been a success.7 The nature and extent of that success deserve to be studied and made more widely known in other parts of the country. We no longer have to merely speculate about the effects of ending racial preferences in higher education. Some results are in.

Notes

1. Portions of this essay are drawn from the author’s “California After Racial

Preferences,” Public Interest, Spring 1999, pp. 64–85. For varying perspectives on racial

preferences in higher education, see Stephan Thernstrom and Abigail Thernstrom,

America in Black and White: One Nation, Indivisible (New York: Simon & Schuster,

1997), chap. 14; Shelby Steele, A Dream Deferred (New York: HarperCollins, 1998);

William G. Bowen and Derek Bok, The Shape of the River: Long-Term Consequences of

Considering Race in College and University Admissions (Princeton, N.J.: Princeton University

Press, 1998). Although proponents of preferential policies have hailed the

Bowen-Bok work as definitive, it has been subjected to serious methodological and

substantive criticism. The fullest critiques are the author’s “California After Racial

Preferences” and Stephan Thernstrom and Abigail Thernstrom, “Reflections on The

Shape of the River,” UCLA Law Review, June 1999.

2. Quoted by Leo Rennert, Sacramento Bee Washington bureau chief, under the

headline “President Embraces Minority Programs,” Sacramento Bee (Metro Final),

April 7, 1995, p. A1. One might have imagined that the universitywould protest publicly

that there was no such danger and, moreover, that that kind of invoking of the Yellow

Peril was deplorable. Similar remarks from a Pat Buchanan, say, would likely have

produced a flood of criticism. But no objection was forthcoming from university

officials because the remarks were made by someone on their side of the preference

issue.

3. The ban on preferences took effect in graduate and professional schools in 1997;

undergraduate admissions were not affected until 1998.

4. The Roper Center Survey of Faculty Opinion About Affirmative Action at the

University of California, October 1996, sponsored by the California Association of

Scholars, at http://www.calscholars.org/roper.html.

5. This was an October 1996 survey of academics in colleges and universities all

over the country. It found even higher proportions of respondents opposed to raceethnic

preferences in admissions and appointments than was found in the UC survey.

See http://www.nas.org/roper/exsum.htm.A recent survey of undergraduates by Zogby

International found that 79 percent said lowering the entrance requirements for some

students, regardless of the reason, was unfair to other applicants. And 77 percent said

it was not right to give preferential treatment to minority students if it meant denying

admission to other students. Ben Gose, “Most Students Oppose Racial Preferences in

Admissions, Survey Finds,” Chronicle of Higher Education, May 5, 2000, p. A52.

6. Andrew J. Kleinfeld, Circuit Judge on the United States Court of Appeals for

the Ninth Circuit, in an essay, “Politicization: From the Law Schools to the Courts,”

Academic Questions, Winter 1993–94, p. 17.

7. See Pamela Burdman, “After Affirmative Action,” Crosstalk 8, no. 1 (winter

2000): 7–8.

==============================