Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix A Notes on Methodology, Definitions, and Needed Data and Research A CLARIFICATION OF SOME PROBLEMS AND CONCEPTS MEASUREMENT In the task of sorting out and evaluating a large set of diverse measures and indicators of conditions and their changes, many of the most interesting questions involve ~li~cult problems of scientific inference. How does one estimate the outcomes of political participation' What definitions and meas- ures should be used to evaluate changes in the economic status of blacks, What factors should be considered in an evaluation of the fairness of the nation's criminal justice system' Hundreds of such choices underlie the text of this volume. As we noted in Chapter 1, the committee's work involved four tasks: 1. ver~cat~on: critical checking of facts and analyses; 2. extension: widening of scope and elaboration of analyses; 3. discovery: funding new knowledge; and 4. assessment: evaluation of significance and implications of data and analyses. Verification involves not only ascertaining the validity of evidence, but also updating: that is, bringing forward historical series of data into the present to determine their continuing validity. Our extension involves bringing into a single report a wide range of com- plex evidence, linking together economic and political changes with changes in family structure, residence, health, and organizational and community life. We found that many widely accepted generalizations are misleading, and we often had to disa,g,gre~ate national data to see important differences among regions of the country, among individuals and families, and among other demographic groups, such as those based on age, sex, or education. 559
A COMMON DESTINY: BLACKS AND AMERICAN SOCIETY Discovery of new knowledge has been sought primarily by reanalysis of data, as in our study of the changing income distributions of black men and women compared with white men and women (Chapter 6~. Occasionally, we found no adequate data available and had to collect new information, as in the work that led to our discovery of the many organized self-help activities in black communities (Chapter 4~. Assessments of significance and of implications cannot be simple extensions of analysis; they must integrate empirical findings with knowledge of the broader sociocultural setting and with interpretations of values and potential policy options. As these comments suggest, the study confronted challenging technical problems in measuring or indexing changes in the status of black Americans; not the least has been the problem of defining what we mean by "status." MEANING OF STATUS In common parlance, status most often refers to a person's or group's relative social position within a hierarchical ranking. The rules governing rank order, formal or not, will usually be greatly determined by the specific contextual situation reflecting the values, norms, and institutions of a soci- ety. Intuitively, status means nothing more or less than this. We might initially pose the question: Whose conception of status' Unfor- tunately, there exists no universal and unchanging conception of status suspended in the human imagination like a platonic form. The idea of status in this report must be more flexible and amenable to alternative values and beliefs. Five conceptual dimensions of social status are of interest to this report: (1) social science indices of status variables; (2) white perceptions of black status; (3) black perceptions of black status; (4) black perceptions of white status; and (5) white perceptions of white status. Studies of black status too frequently discuss (1) and (2) exclusively. Omission of the last three aspects of status is a serious error: (3) and (4) are especially important because they represent a crucial and frequently cited criticism of Myrdal's (1944) monu- mental study. Myrdal's analysis of black status basically discussed external factors and circumstances as if they affected the black community in a vac- uum. Virtually no important role was given to the crucial part played by black input and autonomy in spurring black progress and the formation of independent black institutions. Black status is the creation of American social institutions and the race relations that have developed within that institutional structure. Social indi- ces of black status are concrete representations of that status. As such, measurable status indices are the primary objects of our analysis, and their study encompasses most of the material presented throughout the report. However, beliefs and attitudes-perceptions-are also important in a study of group status. People's attitudes and beliefs about one another are impor 560
APPENDIX A tent consequences of the structure of society and its race relations, as well as major determinants of race relations. One approach to defining status is to simply describe an array or composite of the kinds of statistical indices of status commonly used by social scientists. Average income, education, and deaths per 1,000 population are measures of various aspects of status. But in addition to obvious technical problems What do we mean by the average? Should we measure educational status by mean or median attainment levels' Should some other statistic be used altogether?-there is a general index problem. Educational status should refer to scholastic achievement and enrollment rates as well as to years of attain- ment. Other areas have similar problems. There exists a plethora of dimen- sions of status indices about each general area discussed in the report. When we say that status has improved or deteriorated what exactly do we mean? It is fine in the detailed presentation of facts offered throughout this report to give a descriptive analysis of a long list of status measures, some of which will have improved and some of which will have shown no improve- ment. But in our final assessment this will not do. Readers-and we-still want to know: What has happened to black status? There is no complete answer to this question. No composite measure of status serves as an index that will allow a single statement of this sort to be made. Our resolution of the problem is our major findings (see Summary and Conclusions). A major finding is a statement that is true for a wide variety of different measures and dimensions of status. Their usefulness, of course, depends on their generality-the more status measures included un- der a statement the more useful it becomes. The use of this concept does not allow us to make a single statement assessing black status, but it does classier questions concerning black status into two categories: those that are consistent with our major findings and those that are not. A discussion of these categories and the major findings then allow an intelligent and not too oversimplified assessment of the status of black Americans. RACIAL ATTITUDES In the broadest sense, racial attitudes involve any thoughts, beliefs, and feelings concerning blacks and whites as groups, as well as orientations to- ward appropriate relations between the two groups. Less globally, racial attitudes refer to consistent tendencies toward positive or negative evalua- tions of racial groups, their characteristics, and such aspects of intergroup relations as integration, equal treatment, and nondiscriminatory behavior. By and large we must infer the existence of these attitudes on the basis of replies to questions asked of sample survey respondents or of subjects in laboratory experiments. In some instances, we discuss studies that infer attitudes on the basis of systematic behavioral observations. And, we infer attitudes on the basis of opinions expressed in sources such as newspapers, books, other documents, and lectures. For all of these cases, it is important 561
A COMMON DESTINY: BLACKS AND AMERICAN SOCIETY to note that we are working levity indicators of the attitudes, rather than directly measuring the underlying attitudes themselves. There are several specific reasons for our concern with racial attitudes and beliefs. On the most general plane, Americans' attitudes about the "color line" can be understood as a test of their commitment to democratic values. Tolerance, equality, and respect for minority rights are all core democratic values. Indeed, Myrdal premised his basic analysis of-and his optimism regarding American race relations on the sharp contradiction between these enduring values and the discriminatory treatment accorded blacks. Although there are reasons to question the extent to which most Americans experi- enced psychological anguish over the "American dilemma," there is no doubt that the character of racial attitudes and related behaviors reflects on the success of American democracy. Racial attitudes and beliefs are also important elements of the general social and political climate. Prevailing norms on race can either strongly discourage prejudice and discriminatory behavior or they can encourage such patterns of thought and action. Those who lack strong prejudices may behave in a discriminatory manner if most of their peers expect or demand such behav- ior. Similarly, those who harbor animosities may refrain from acting on their inclinations if society condemns them for doing so. The prevailing norms on black-white relations are thus aspects of the broader social fabric to which people all must adapt in some way. Studies of attitudes and beliefs provide a key source of information on such norms. More concretely, however, the underlying preferences of the public are not without consequences for policy making. The will of the people is supposed to rule in a democratic society. However, public opinion on many issues is often unfocused, contradictory, and therefore not readily mobilized. Public opinion must also be mediated through the actions of elected and appointed officials at various levels of government. Yet, when public opinion on an issue is well crystallized, and when there is an overwhelming or growing majority for some issue position, it is likely that policy will in some form come to reflect those mass preferences. This process appears to operate for all policy issues, not just black-white issues. At minimum, it is difficult and costly, in practical terms and in terms of maintaining political legitimacy, for government to implement policies that large segments of the population oppose. Thus, it is important to assess the preferences of blacks and whites ^ I idyllic he much nr~f~r~nces nlav a role. even if indirect, in policy =~ ~`L`` I EVE - rat r--~ making. Although attitudes, underlying feelings, and beliefs influence individual behavior, the relationship between attitudes and behavior is seldom one-to- one. Social norms, as noted above, as well as other situational constraints such as laws, expected costs and benefits, and psychological considerations- the intensity of feelings associated with an attitude-affect whether or not an attitude influences a person's behavior. One cannot argue, however, that underlying attitudes bear no relationship to individual behavior. The rela 562
APPENDIX A tionship between attitudes and behavior is complex, yet attitudes are one important guide to a person's likely behavior. Prior efforts to assess the status of black Americans, whether focused on particular cities (Clark, 1965; Drake and Cayton, 1945; DuBois, 1899) or the nation as a whole (Cox, 1948; Johnson, 1930; Myrdal, 1944; National Advisory Commission on Civil Disorders, 1968), have attempted to under- stand the sources, character, and consequences of racial attitudes. DuBois recounted a number of incidents in which individual whites, either in re- sponse to social expectations or, more often, their own underlying attitudes, discriminated against blacks (1899:322-355~. He argued that "color preju- dice" contributed to the difficulties blacks faced in getting and keeping jobs; in raising their children in a hostile social environment; and in the ever- present potential for social rebuffs and ostracism. Johnson, recognizing race as one of many possible bases for group differ- entiation, traced "race prejudice" to the recognition of group boundaries, especially as such boundaries had been linked to economic and status com- petition (1930:355-362~. In a similar vein, Drake and Cayton's analysis of the "color line" emphasized the consequential, yet complex interweaving of folk prejudices, economic interests, and social status concerns (1945:266- 276~. Myrdal placed a number of subjective variables, such as valuations and beliefs, at the center of his work. All of these scholars thus recognized the importance of formulating an understanding of the meaning people attach to a phenomenon like race. They suggested, moreover, that the potential for progress of blacks as a group is in part a function of society's prevailing racial attitudes and beliefs. Our concern with racial attitudes should not be interpreted to mean that attitudes are a fundamental basis of the status of black Americans. Individual attitudes and beliefs are more likely to reflect the current and enduring features of an organized social environment than they are to independently shape or determine such social structures. Furthermore, the status of black Americans is powerfully determined by demographic and economic factors that have little or no dependence on racial attitudes. Yet the types of rela- tions desired by white and black Americans, and Americans' interpretations of the nature of racial inequality, are of unavoidable concern if we are to understand the character of popular discourse on these issues. In sum, prevailing attitudes and beliefs can be viewed as a set of demands on political leaders; as a set of broad constraints on viable reform agendas; as dues to likely individual behavior; and as a measure of success at fulfilling certain democratic values. Racial attitudes are thus a necessary concern of a compre- hensive attempt to understand the status of black Americans. Our understanding of the causes of change in racial attitudes is, however, far less certain than our understanding of the basic patterns of change them- selves. The processes that lead from behavioral change to attitude change, or from chat ng social norms to changes in individual attitudes, are not well understood. In particular, few systematic and longitudinal efforts to link 563
A COMMON DESTINY: BLACKS AND AMERICAN SOCIETY empirical measures of attitudes to measures of contextual factors, such as the integration of a school or workplace, have been conducted. Furthermore, attitudes and public opinion are part of a dynamic and somewhat labile social process (Myrdal, 1944:1032-1034~. Although we expect individuals' racial attitudes to be stable over any short span of time, individuals' attitudes are not entirely static. The direction and magnitude of changes in attitudes can be both large and unexpected. In addition, even though one might like to propose the operation of a process of gradual and continuous positive improvement in racial attitudes, this claim is not consis- tent with historical experience or with much of the data we present. Change is possible but it occurs in complex ways and for only dimly understood reasons. Although only a beginning has been made in the needed empirical analysis of these complex effects, we suggest that the generic problem can be seen more clearly than previously. Many of the data we review in this report refer to attitudes as indicated by specific opinions, usually expressed in interview situations. Such opinions are often short-run, volatile phenomena that can shift drastically in response both to microcontexts-situations-and to ma- croevents, such as an economic depression, national election, and war or other international crises. The constant interaction of transformative social movements, established social institutions and actors, everyday human ad- aptation and activity, and common-sense understanding affects patterns of racial attitudes and beliefs. MEAN I NO OF RACE Throughout this report we use the terms "black" and "white." These are social categories that have long been viewed as meaningful in the United States. The meaning of race is a matter of social interpretation, however, not a fact of biology or genetics. Since the beginning of European settlement in North America, the cultural models that have defined the status of blacks have gone through several great transformations. Originally the distinction was between Christians and hea- thens, then between slave and free. As these two definitions failed to distin- guish between dominant and subordinate social groups, race was invoked as a quasi-biolog~cal concept that was used to explain and justify white suprem- acy. When doctrines of racial inferiority had to be abandoned and when equal rights became a persuasive political doctrine, "race" came to be largely subsumed under "ethnicity" or "class" or both. At the extreme, it could be argued that a racial category served only as an identification or marker and that the primary dynamics of race relations really were those of class relations. However, since the mid-1960s, a substantial body of thought has developed that supports earlier rejections of the concept that blacks consti- tute an ethnic group analogous to European-origin ethnic groups. The rejection is based on objective evidence of the unique extent and harshness of discrimination against and segregation of blacks, as well as on 564
APPENDIX A the intensity and prevalence of negative racial beliefs and ideologies. Simi- larly, it is argued that racial status cannot be reduced to a matter of class position in the marketplace or of the relations to the means of economic production. Thus, an emphasis on the uniqueness of race as an "irreducible" category has emerged from the critical debates over the standard paradigms of prior interpretations. Another ingredient in the reformations of recent decades has been the reworking of "black nationalism" to include modern experiences. All these lines of conceptual analysis converge in formulations that treat the racial category "black" as a social reality that combines class, ethnicity, cultural heritage, political interests, and self-definition (see Omni and Winant, 1986~. Nevertheless, the notion of race often turns out to be vague and to have multiple meanings, and it has varied greatly over time. As Omi and Winant wrote (1986:60~: O ~ Race is indeed a pre-eminently soc~oh~stoncal concept. Racial categories and the meaning of race are given concrete expression by the specific social relations and historical context in which they are embedded. Racial mean- ings have varied tremendously over time and between different societies. In spite of its changing and uncertain meanings, the idea of race as a rigid and unchanging category has been pervasive in the United States and has received detailed legal definitions. As recently as 1970, a Louisiana statute specified that a person with at least 1/32 "Negro blood" was considered black. The definition of racial category in this instance, as in many others, is a political act. American history provides numerous examples of variation and changes in racial categories and the eventual separation of race from nationality or ethnicity and from religious categories. The concept of "race" has a long and tortuous history and its baggage of meanings is enormous and diverse. Looked at in sociocultural terms, "race" is one among many forms of categorization, a subprocess of the distribution of social identities and roles within populations. People are categorized by other people in many different ways: by age, sex or gender, occupation, intelligence quotient, conformity or nonconformity to social norms, left- or right-handedness, athletic prowess, religious piety, and so on. But some categorizations are more consequential than others. Differences in skin color, type of hair, and facial features that are biologically trivial have been used as markers for ascribing great differences in power and privilege. As Banton noted (1983:77~: "Race" relations are distinguished not by the biological significance of phenotypical features but by the social use of these features as signs identi- fying group membership and the roles people are expected to play. In the United States, race (or color) historically was linked with slavery and subsequently with a harsh and rigid system of stratification. For this reason, many observers have had the opinion that race is a peculiarly rigid basis for ethnic relationships. Comparative analysis of the world's societies 565
A COMMON DESTINY: BLACKS AND AMERICAN SOCIETY hat race is the indicator, not the substance, of superordinate- onships of distinguishable social groups (see Horowitz, y one of many traits may be used to identify group mem- ligion, language, dress, speech patterns, food preferences ares, and bodily markers, among others. What is used to Boundaries may vary greatly from situation to situation and is not the attribute that makes the group, but the group ences that make the attribute important" (Horowitz, `~ terms "race" or "racial" in this report, we are accepting conventional usage that has little precision. "Race" in the social construct that relies on common understandings and :her than scientific criteria. avoid treating blacks and whites as internally homogenous our analysis, in fact, seeks to identify sources of variation I, and similarities in the attitudes of blacks and whites are here are frequent and important differences in the attitudes rites on race-as well as on many nonracial issues. It is suit, to speak of differing tendencies between these social [ - ~ am__ F RACISM ~1 attitudes are often equated with studies of racism. We it to be precise in the use of this term. For some people, y form of race recognition, especially instances in which leged groups act in a manner injurious to a disadvantaged owever, reserve the term for patterns of belief and related tly embrace the notion of genetic or biological differences groups. Still others use the term to designate feelings of ity. And of course, some definitions include all of these manifestations of racism. Each of these uses of the racism e validity, but it is surely unfortunate that a single analytic rany different meanings. :rm racism to denote biological racism, as in the second ove. Societal racism, borrowing from Fredrickson (1971), is .egative racial attitudes or outcomes that lack a clear basis in at racial inferiority. Mere recognition of social groups based ~acteristics is not treated here as a form of racism, but as cious. " Cultural preferences that do not include the system- ~cial groups and dear hostility toward out-groups is termed if racism, however qualified and defined, involves a value m of whatever variety is undesirable; racist outcomes are pie who advocate racist ideas are typically viewed as being , if not dangerous. Some people question any scholarly use 566 shows, however, subordinate relati 1985:42-51) . An bership: color, r alla taboos, post, establish group b time to time: "It and group differ 1985:50~. When we use t for convenience a United States is a self-definitions rat In general, we groups. Much of within each grou also noted. Yet t] of blacks and w' necessary, as a re groups. MEANING O Studies of raci; think it important racism means an members of priv group. Others, ~ actions that over between human cultural superior . . . . posse ~' sties as concept has som concept has so rr We use the to . · . interpretation an used to denote a belief in inhere on "racial" cha: being "race cons atic ranking of s. ethnocentrism. The concept ~ judgment. Racis wrong; and peo morally deficient
APPENDIX A of this concept because it is so manifestly value laden. Others respond, with justification, that no complete and honest treatment of American racial attitudes and beliefs could entirely eschew use of the concept of racism. Both _~_~_~V Am_ ~_4 Ace_ ____ an ~1 and. ~_~^ ~_~ ~ V__ _~_~`L __ concerns have legitimacy. NEEDED DATA AND RESEARCH The committee had to cope with gaps and other inadequacies in the data, even for basic descriptive tasks. Recent changes in the collection and report- ing of statistical data by federal agencies, and proposed changes in the na- tional census, may seriously limit the information needed for analyzing de- mographic, social, and economic changes over time. The scientific importance of maintaining comparable and detailed time series must be strongly empha- sized. For understanding the "why" of changes such as those we discuss in this report it is crucial to have longitudinal information-measurements on the same units over periods of time long enough to detect significant changes. Correlations based on cross-sectional data are difficult to interpret and may often he mi.sleadin~ ~ Many needed analyses of the phenomena treated in this report have been limited by reason of the absence of adequate data. Current data on causes of death, for example, do not contain information that sufficiently allows re- search to pinpoint relationships between socioeconomic status and individ- uals' health practices. Analysis of the effects of geographic concentrations of poor people in cities is severely impeded by lack of information on mobility and migration patterns. Studies of unemployment and the out-of-labor-force population face an almost total absence of longitudinal data on job search behavior in relation to an individual's educational status, military service, and so on. We have very little information about the life-history transitions from school to work, or to nonwork, for young people. The data that are available for describing large-scale economic, political, and social conditions typically have been generated for reasons other than scientific relevance. The national census is mandated for legislative, adminis- trative, and other purposes, such as representation in legislatures or moni- toring the health of the economy. Valuable as they are, such data only rarely accurately represent variables as conceptualized in basic scientific hypotheses and theories. Thus, great effort and ingenuity often are evident in "making do" with proxy and surrogate variables-indicators that imperfectly corre- spond to underlying concepts. As we have examined one substantive area after another, it is instructive to observe how often apparently contradictory or anomalous findings reflect differences in indices, definitions, data samples, or statistical models. Data sources that explicate such differences aid the vigilance of research workers in identifying information that can greatly aid appraisals of factual basis for public policies. Thus figures on "unemployment" need to be interpreted 567
A COMMON DESTI NY: BLACKS AN D AMERICAN SOCI ETY along with information on persons who are not in the labor force. Failure to distinguish between "income" and "earnings" similarly may lead to mis- leading conclusions. Our committee noted that discussions found in popular works concerning alleged effects of public assistance on work and family life lacked a definitive factual base. There is much speculation and many anecdotal references, but systematic evidence from well-controlled analyses is rare (see Chapter 10~. Specifically needed is comprehensive research on what happens to so-called "cultural patterns" when new economic and educational opportunities are opened to segregated, low-income populations. Much of the available research on the effects of social contexts is at the levels of aggregation represented by school districts, zip-code areas, and other geographical units. Using such aggregate data, most of the recent research that has used post-1970 statistical methods has found few large effects dearly attributable to collective settings (e.g., mean socioeconomic status, mean proportion black, mean education of parents). This surprising result may be partly due to limitations of data and study designs that made it impossible to adequately specify relevant independent variables and outcomes for suita- ble populations. For example, many studies have only cross-sectional corre- lations; many longitudinal studies cover only short periods of time; popula- tions often are not appropriately differentiated by age, sex, race, ethnicity, income, area of residence, and so forth. There are few studies concerning the actual social structure of high-poverty neighborhoods or the attitudes and behavior patterns of residents. The few ethnographic studies that exist are very useful, but their findings cannot be safely generalized to all other settings. Needed is systematic research, over time, that collects comparable information from strategic sampling locations on changes in job search, employment, family patterns, social services, crime, informal social structures, schools, and residents' views of their situations. Similarly, there is a paucity of observational studies of behavior in different kinds of schools with varying proportions of black and white students. Evidence received by the committee shows important geographic varia- tions, both local and regional, in basic socioeconomic conditions. In this report, the need for conciseness, as well as a focus on national conditions, led to extensive reliance on aggregated data; it would be valuable if future work drew on more detailed tabulations. Some systematic accumulations of detailed longitudinal information do exist for example, the national longi- tudinal surveys of labor market experience. The original cohorts comprised 22,157 individuals (from 13,582 households), to which was added a youth cohort of persons aged 1~21 as of January 1979, an additional 12,686 persons. The scope of potential analyses is suggested by the fact that main data files for the youth cohort contain over 20,000 variables. This appendix has covered only a few of the more important issues related to methodology and data sources relevant to this report. It is our hope that future research can discover new techniques to deal with the problems noted 568
APPENDIX A here and that, indeed, this report ~11 stimulate such research as well as improved data collection. REFERENCES Banton, Michael 1983 Racial and Ethnic Competition. Cambridge, England: Cambridge University Press. Clark, Kenneth 1965 Dark Ghetto. New York: Harper and Row. Cox, Oliver C. 1948 Caste, Clues and Race: A Study in Social Dynamics. New York: Doubleday. Drake, St. Clair, and Horace Cayton 1945 Black Metropolis: A Study of Ne,gro Life in a Northern City. New York: Harcourt Brace. DuBois, William E. B. 1899 The Philadelphia Ne,gro: A Social Study. Reissued (1973), Millwood, N.Y.: Kraus- Thomson Organization Limited. Fredrickson, George 1971 The Black Image in the White Mind: The Debate on AfiwAm~ncan Character and Destiny, 1817-1914. New York: Harper and Row. Horowitz, Donald L. 1985 Ethnic Groups in Confect. Berkeley: University of California Press. Johnson, Charles S. 1930 The Negro in American Civilization: A Study of Negro Life and Race Relortions in the L'ght of Social Research. New York: H. Holt and Company. Myrdal, Gunnar 1944 An American Dilemma: The Ne,gro Phoble~n and Modern Democracy. New York: Harper and Brothers. National Advisory Commission on Civil Disorders 1968 Rest of the National Advisory Commission on Civil Disorders. New York: Bantam Books. Omi, Michael, and Howard Winant 1986 Racial Formation in the United States: From the 1960s to the 1980s. New York: RDutledge and Kegan Paul. 569
