Achievement Gap in Education
Around the turn of the century, some American social scientists began to develop the concept of culture. This view universalized the capacity to be human. During the early years of the new century, most social scientists rejected the idea that racial differences were significant factors in determining cultural development.
Racism took its toll in the various institutions of American society, including education, government, and economic life. Yet, the rise of formal justifications for racism strengthened the resolve of those who discriminated.
Much has been written concerning the effects of desegregation. In nearly all these studies, however, no distinctions are made among various kinds of effects. As a result, the literature of the subject is confusing. Instead of asking whether desegregation has an effect or affects this or that, we should ask how different kinds of desegregation effects are manifested. Conceivably, a specific desegregation experience could be effective in one way and less effective or even ineffective in others.
Desegregation effects can be expressed on three levels: policy, mechanical, and instructional.
Policy effects are those changes induced by managerial decisions to alter the goals of the school system and to reallocate resources accordingly. They affect organizational units rather than individual students. Nonetheless, they may be of crucial importance. A decision to pair a severely shortchanged minority school with a well-endowed majority school sets the scene for more equal and more productive education.
But the decision to pair, while critical for further progress, is insufficient to produce actual educational change for individual students. Policies are only that–policies. They remain to be translated into action. Without the new policies, however, little more can happen.
Mechanical effects are changes in general housekeeping routines and in the physical flow of students into, through, and out of a school. Such processes have only an incidental relationship to a school’s educational function. Yet, they may come to be perceived as major elements in the desegregation process because of their physical prominence. Busing is a prime example.
Strictly speaking, its only effect can be to transport children from one place to another. Aside from that, busing can neither be assigned nor be expected to discharge an educational function. Educational research has never found educational achievement dependent upon the particular mode of student transportation to and from school–whether a student walks or rides has no effect on that student’s learning.
Consequently, it is improper to attribute any educational effect to busing–or walking, for that matter. Issues of distance, health, and comfort apply to all modes of transportation.
Instructional effects are changes in the educational effectiveness of a school in terms of individual students. These effects may be linked to changes in the climate of the school, including high expectations and goals for all the children in a school.
They may be facilitated by reorganization of curriculum and of modes of student-teacher relationships; increased use of cooperative modes of learning; improvement of textual materials; creation of an atmosphere of mutual respect and regard for all members of the school community; and, not least, employment of a firm central administrative policy of support.
Extremely few studies have been made of the policy effects of desegregation. On the other hand, numerous studies utilize the word “busing” in their titles without, in fact, measuring busing as such. The great mass of studies of instructional effects of desegregation relates to the narrowest aspect of the subject–namely, achievement test scores.
Little if any attention is paid to the classroom dynamics behind the scores, to the bearing of curriculum on them, to the organization of student learning modes, or to the possible effects of school or school system policies.
Before the 1950s, educators were not interested in studying how racial segregation affected academic achievement. For one thing, the answer seemed obvious. Boards of education made few attempts to maximize the academic achievement of minority children. Moreover, in such school systems funds intended for the use of minority children were frequently diverted for use in white schools, to the detriment of minority achievement.
None of the national organizations of school board members, administrators, teachers, or educational researchers protested against such practices. University schools or departments of education were silent, as well.
The 1954 Brown decision and the rising civil rights movement changed this. Civil rights advocates demanded that minority children be educated to an acceptable standard. The abolition of segregation was thus linked with an anticipated rise in educational achievement. Some even thought the former would automatically produce the latter.
Behind this reasoning lay a conviction that school personnel were neutral specialists, just as capable and ready to educate as to exclude minority students. As segregation gave way to desegregation, these observers expected that antagonistic attitudes as well as material inequalities would give way, and that minority achievement would prosper.
Progress came slowly, even though parents of poor and minority children expressed deep concern for improving their children’s education. In one city after another, the civil rights movement called upon school boards to publish school-by-school achievement scores so that weak spots could be identified and concerted measures undertaken to eliminate deficiencies.
Boards of education, however, resisted such demands. Even in cities in which scores were published, they were treated as isolated signposts rather than as distress signals.
In the face of reluctance of school boards to act, the question arose whether achievement in desegregated schools was any different than that in segregated schools. Ideological opponents of desegregation were certain achievement would decline in the newly desegregated schools. Academic skeptics, unused to analyzing racial factors in education, failed to see their possible relevance in the present case. And proponents of desegregation hoped for the best without much evidence or assurance of success.
During the first decade after Brown, few schools in the South were desegregated beyond a token basis. In the North, Brown brought no changes, at first. There were, therefore, few cases of actual desegregation. Persons seeking overall views of the effects of desegregation on achievement looked in vain for studies.
Between 1964 and 1970, Weinberg wrote the first three works devoted exclusively to desegregation effects. In a pamphlet written in 1964, he surveyed whatever scrappy evidence existed. He examined the few dissertations available, the testimony of superintendents, and statements of various firsthand observers. He concluded that desegregation seemed to improve the achievement of black children without adversely affecting that of white children.
That same year, 1970, Nancy St. John published a lengthy article which reviewed studies of desegregation and achievement. She criticized many of the studies as methodologically weak, a point that Weinberg had not emphasized. A number of studies, for example, did not take socioeconomic conditions into account; others did not have adequate control of factors under study; still others lacked a pretest and thus might have confounded a condition existing before desegregation with an effect of the desegregation process itself.
Desegregation and achievement were the central concern of a book published by St. John in 1975. This work dealt less with methodological shortcomings of studies. A critical aspect of her views, however, was her conviction that studies on the subject showed only that desegregation was an indeterminate process. Whether desegregation aided achievement or not depended on other factors, including the contribution of the teacher.
In 1977, Weinberg published a new book which included a comprehensive examination of evidence bearing on desegregation and achievement. 5 In it he reported on evidence of all sorts, including studies in which desegregation was found not to encourage higher achievement. Again, he concluded that desegregation helped minority achievement in more cases than not.
Overall, both St. John and Weinberg pretty much took the evidence as it was, trimmed off the most obvious shortcomings, and drew cautious conclusions. St. John tended to be somewhat more skeptical, while Weinberg placed more emphasis on empirical and historical evidence and more positive results. Indeed, writers in the field who favored one or the other emphasis tended to cite only the studies which they found acceptable, ignoring the others. The resulting distortion was unfortunate.
During 1977-1978, there were new advances in the evaluation of desegregation research and the effort became more systematized. For another, instead of continuing to make the same old queries, researchers asked new questions of the data.
In 1978, Ronald A. Krol sallied forth onto the rare heights of meta analysis, a term meaning the process of analyzing analyses. His central interest was to settle, if possible, the debate about the effects of desegregation on academic achievement.
He derived his hypotheses explicitly from the St. John and Weinberg books, although unlike these two authors, he established precise rules for admitting evidence. He located 129 analyses that satisfied the six criteria he set up. A study had to be of a before-after kind. Change in achievement could only be measured in comparison with an earlier point. Achievement had to be measured quantitatively; otherwise, exact determinations of effect-size could not be made.
Certain statistical information had to be provided in the study. This included the number of students and the extent of variation in scores within the population studied. Studies using only measures of attitudes, whether those of teachers or students, were excluded. Studies that measured a situation only once (“crosssectional”) were not considered.
Krol pointed out that many such studies analyzed the effect of different racial composition ratios on achievement at one time rather than the effects of a change from segregation to desegregation over a period of time. Studies seeking to ascertain the effect of desegregation on I.Q. scores were rejected since Krol deemed I.Q. and achievement to be different concepts.
A total of 129 analyses performed in fifty-five separate studies remained. What did these analyses reveal?
For seventy-one studies, reported Krol, “the mean of the experimental group [the desegregated group] exceeded the mean of the control group by 0.16 standard deviations.” 7 In only ten out of seventy-one cases did the experimental group score less than the control group. Krol also tested whether the positive achievement effect of desegregation was greater for younger than for older children. It was not. He found differences between average achievement in mathematics and reading and length of exposure to desegregation made no difference for achievement.
Krol thus detected “slightly positive” effects on achievement. He noted that less rigorous techniques of analysis yielded more or less the same conclusion. One such technique–which Krol calls “voting”–was employed by Weinberg when he observed that more times than not, desegregation affected minority achievement positively.
At the same time, Krol also reported: “Studies which were considered to be of a weaker design (no control group) did have a greater effect size than those studies which were considered to be stronger (with control group).” Krol, however, rejects the possibility of drawing a negative conclusion from the studies he analyzed: “One cannot say based on this study that desegregation produces harmful effects.”
Krol is not an admirer of the desegregation studies he reviewed. None, he declares, is an exemplary study; at best, they are mediocre. What is more, they are indefinite as to the kind of experimental or other “treatment” going on in the classroom as a result of desegregation.
“The treatment for all the studies in the meta analysis,” writes Krol, “is simply the act of placing the minority children in a predominantly white school.” He notes: “None of the researchers attempted to find out the atmosphere inside these schools, or if the children were actually placed in integrated classes within the desegregated school.”
Evidence from other quarters suggests that minority children in formally desegregated schools frequently find themselves in segregated classes. They are thus deprived of any positive achievement effect. Conceivably, more adequate enforcement or legal desegregation requirements might yield greater achievement.
The reluctance of desegregation researchers to investigate classroom dynamics has been noted for more than a decade. Sponsors of research, however, continue to ignore this glaring weakness.
In 1978, Robert L. Crain and Rita E. Mahard published an analysis of desegregation and academic achievement. These investigators broke new ground by asking novel questions, far beyond a simple one about the existence of overall test-score differences between students attending desegregated or segregated schools.
Does educational success depend on region? Nearly two-thirds of fifteen studies of students in the South showed positive gains for black students while only slightly over a third of twenty-six studies of northern students showed such a gain. Is desegregation at an earlier age better? Crain and Mahard found the answer “clear and unmistakable”: the earlier the grade at which desegregation occurs, the more positive the impact on achievement.
Do desegregation effects vary by curriculum? The quantitative evidence is not at all clear-cut but Crain and Mahard believe that, in many cases, desegregated black children learn more because they are using a better curriculum rather than from increased racial interaction.
Are voluntary or mandatory desegregation plans more effective in stimulating black learning? “On the whole,” the researchers report, the evidence suggests that there is a significant correlation between mandatory assignment and positive achievement outcomes. Crain and Mahard cautioned their readers against hastening to apply this somewhat unexpected finding.
All in all, Crain and Mahard maintain that desegregation has a positive effect on black academic achievement. But not always. The effect depends on what actually goes on in the classroom:
Desegregation sometimes results in better curricula or facilities; it often results in blacks having better trained or more cognitively skilled teachers; it is frequently accompanied by a major effort to upgrade the quality of education; and it almost always results in socioeconomic desegregation. When desegregation is accompanied by all of these factors, it should not be surprising that there are immediate achievement gains half to two-thirds of the time.
They estimate the average achievement gain of blacks during the first one or two years of desegregation as about one-half of a grade equivalent.
In 1977, Laurence and Gifford Bradley conducted a study which paralleled that of Crain and Mahard. They sought to discover whether existing studies established a positive achievement effect for desegregation. The Bradleys and Crain and Mahard completed their studies at about the same time although publication was delayed for a year in the latter case. Apparently neither pair knew of the other’s study.
The Bradleys asked whether open enrollment–choice of school by individual students regardless of residence–encouraged black academic achievement growth.
All four studies of this technique reported a positive achievement effect. While, according to the Bradleys, two of the studies had grave methodological weaknesses, the other two were more sophisticated. One might conclude, even if the Bradleys did not do so in so many words, that studies of open enrollment supported a positive achievement effect.
The Bradleys then examined the evidence bearing on central schools as a desegregation device. These are schools to which students are assigned after their own schools have been closed, the better to achieve a desirable racial mixture. “All of the Central Schools investigations reported increased black achievement following desegregation,” wrote the Bradleys. They described some of these studies, however, as methodologically weak.
A number of researchers have studied school closings and dispersal of students throughout a city. Their findings are mixed and, the Bradleys asserted, the studies are so weak methodologically that no firm conclusions can be fairly drawn.
There are numerous studies of the effects of desegregation plans implemented by busing but the Bradleys did not reach a definite conclusion on them. Discussing the well-known Pettigrew-Armor debate on the subject, however, they declared: “Armor’s conclusion that busing is an ineffective intervention must be considered tenuous.”
The Bradleys examined a single experimental study by Frary and Goolsby in Gulfport, Mississippi. They found what the Bradleys agreed was “relatively strong evidence that desegregation may have beneficial effects on black student achievement.”
Throughout, the Bradleys stressed heavily the presence of methodological weakness in practically all desegregation studies. They considered most of these weaknesses serious but not fatal. We are thus left with imperfect evidence. Some of the weaknesses are comparisons without an adequate control group; as a result, an achievement change in a desegregated school may also have happened in a segregated school.
In that case, a factor common to both schools may be the formative factor. In other studies, no predesegregation test was given, thus making it impossible to gauge the significance of a test score obtained at the end of a study period.
Single studies continue to be published and several are of special interest.
The first study was conducted by the Systems Development Corporation (SDC) and is an evaluation of the federal Emergency School Aid Act (ESAA). Under the law, federal funds were provided to help school districts carry out a desegregation plan (the Basic program) or to help predominantly minority districts improve the quality of their educational services (the Pilot program).
The research problem is to discover whether the federal dollars were educationally effective. A three-year research period was established, from 1973-1974 to 1975-1976.
The study was plagued by a fatal weakness not of its own making. Schools receiving ESAA funds were matched with comparable schools in the same district which did not receive such funds. Presumably, if the ESAA grants were sizable enough, certain features would emerge that favored the ESAA schools over the non-ESAA schools.
In fact, however, the non-ESAA schools “tended not only to have nearly as much money as the . . . ESAA schools, but also to spend that money for similar kinds of program activities.” It also developed that neither group of schools followed any really innovative techniques or programs. In effect, two groups of similar schools were being tested for differences.
The evaluation was, however, far from a complete loss. A comparison of Basic and Pilot schools, and an in-depth study of thirty elementary and high schools, resulted in important findings. SDC also tried to ascertain what practices improved academic achievement in either ESAA schools or nonESAA schools.
The researchers found that achievement scores of individual students who had been in a Pilot school over two years underwent “considerable growth. “ Comparable students in Basic schools exhibited “a consistent pattern of positive differences. . . .” This did not happen in Basic high schools, but, in the tenth and eleventh grades in these schools it was found that “larger reading gains were made by students in districts that undertook more activities designed to facilitate and support school desegregation.
Presumably, such an atmosphere encouraged minority children to learn more readily. When individual students were grouped by schools and studied over two years, another achievement-related finding emerged. “The strongest relationship in these data was between the teacher’s expectations for the student . . . and the reading outcome. . . . The SDC researchers emphasize that no significant achievement differences were found between ESAA and non-ESAA schools as such.
At least as significant as the main study was an in-depth study that was also made of twenty-four elementary schools. Fifteen were “successful,” that is, at least two grades in each school recorded a rise in national percentile ranks in reading or mathematics. Nine were “unsuccessful.” Observers visited each of the schools for a period of two weeks; they sat in individual classrooms and interviewed teachers and administrators.
In the in-depth study, reading and mathematics scores were maximized by the use of paid parent aides in the classroom; the utilization of behavioral or other objectives in classroom instruction; and allocation of major resources per pupil to employ remedial specialists in the classroom.
“The importance of this study’s findings,” according to Jean Wellisch, “lies in the fact that school program characteristics do appear to make a difference in student achievement.” If so, improvements in achievement are well within the realm of all schools.
It was found that ESAA expenditures for elementary schools in the Basic sample during the three years of the study helped produce a positive effect on achievement both for reading and arithmetic. A second in-depth study reaffirmed the importance of teacher planning for achievement. It failed, however, to replicate the earlier in-depth study’s finding that the presence of paid parent aides helped increase achievement.
The Educational Testing Service (ETS) undertook a second major desegregation study. It covered about 200 elementary and high schools. A survey in 1974 was repeated on a smaller scale the following year. Site visits were made to forty-eight schools the earlier year and to forty-three of these schools the next year. Fifth and tenth graders were tested; teachers, principals, and counselors filled out questionnaires; some were interviewed during site visits.
The three-year-long ETS study is preeminently practical. It is concerned with learning how to foster effective desegregation and integration rather than spinning on about the pros and cons of integration and segregation. Specifically, the researchers sought out specific school conditions that were demonstrably effective. They then asked school personnel in those real situations to suggest to others how integration might best be carried out.
Since so very few applied studies in integration have been made–itself a topic worth studying!–scholars must take care in evaluating such studies. Unlike most theoretical inquiries, the ETS project did not analyze a broad range of experiences of integration. Instead, it compared in great detail very successful instances of integration with ones that had been moderately successful. So unaccustomed are we to such studies that they are sometimes termed “biased” because they do not discuss the “failures.”
The ETS study measured successful integration in terms of academic achievement and race relations. School practices that encouraged good race relations were found to be related to conditions which fostered academic growth. Where blacks were learning, so were whites, especially in elementary schools. In high schools, “achievement is significantly related to perception of school fairness.”
The ETS researchers did not identify any school factor that, of itself, was directly related to achievement differences. Instead, various school factors, interacting with one another, significantly affected achievement. Especially in an applied study, the way a practice is implemented can be at least as important as which practice is implemented. 
The emphasis on desegregation stemmed solely from several sources. Given the history of official segregation both in the South and North, the opening of schools to all was an indispensable goal of civil rights forces. “Segregation is a denial of the equal protection of the laws,” declared the U.S. Supreme Court in 1954. It followed logically that ending segregation would restore equal protection.
Little or no attention was paid to what was to happen in the desegregated schools. Instructing lower federal courts on how to manage desegregation, the Supreme Court in 1955 had declared that school authorities have the primary responsibility for elucidating, assessing, and solving these varied local school problems arising out of desegregation.
Civil rights leaders were willing to leave the actual operation of desegregated schools in the hands of school authorities. In part, this reflected a confidence in the professional integrity of educators; besides, civil rights advocates had few ideas of how to operate a school, desegregated or otherwise. Thus, history, hope, and lack of expertise led to a concern with merely opening the school doors. Good or better education, it was presumed, would follow.
Exclusive concern with quality education, on the other hand, arose from other sources. Racial concentration, whether official or unofficial, was interpreted as a neutral factor in learning. In part, this was a tactical view. Since actual desegregation proceeded very slowly in the first years after 1954, and even up to today–especially in the North–some exponents of quality education advocated educational improvement for minority children wherever they were.
This approach was based on a faulty perception of governmental and school authorities as beneficent. Its adherents assumed that these authorities feared racial desegregation so intensely that they would gladly exchange equal sharing of resources for maintaining segregation. In general, they also assumed that segregation might have a nurturing effect on the black child’s development or that any damage caused by segregation was minimal. Its supporters also contended that actual learning gains under desegregation were nonexistent or negligible in any case.