Measures Of Socioeconomic Status

Current Issues

Mary G Powers

• Publisher: Routledge
• Year: 2021
• ISBN:
  9780429724954
• Edition: 1
• Topic: Sciences sociales
• Subtopic: Sociologie

This text consists of a collection of seven papers which look at constructing measures of occupational socioeconomic status by combining census data on occupation-specific levels of education and income; including different approaches.

1.

Measures of Socioeconomic Status: An Introduction

Mary G. Powers

DOI: 10.4324/9780429049170-1

This volume reviews the major approaches to the conceptualization and measurement of socioeconomic status which have developed in North America over the past thirty-five years, and evaluates these measures in the light of current theoretical and methodological issues.

The parallel development of socioeconomic measures in Canada and the United States reflects the dominant influence of Max Weber in social stratification theory and research. Weber’s articulation of the distinct but interrelated class, status and power dimensions of stratification forms the theoretical basis for most measures of socioeconomic status and prestige which have been developed in North America. Neither the precise nature of the interrelationship among the three dimensions nor the variables underlying each dimension have been specified. There is general consensus, however, that class, or the social and economic life chances individuals experience, is closely related to their status, or the regard with which they are held by others. The logic underlying this consensus is that class or economic position generates income which results in similar life styles among individuals in the same class. Those life styles contribute to the evaluations made of individuals and to the prestige assigned to them by others. Recently, it has been argued also “… that prestige ultimately is rooted in power relations …” (Treiman, 1977:1).

This general theoretical framework strongly influenced the development of measures of socioeconomic status. Research seeking to locate individual positions in a stratification system required a summary measure or index which would accurately and simply assign individuals to such positions. Socioeconomic indices were designed to provide such summary information about the social standing of all individuals in a society in the absence of detailed information about the entire complex of variables involved in overall socioeconomic standing. Such summary measures describe an individual’s location along a single hierarchy and omit many of the complexities of group inequalities.

Two recent reviews and evaluations agree that empirical research as well as existing theory demonstrate that occupation is the most adequate single indicator of position in a complex stratification system (Haug, 1977; Treiman, 1977). Because of the overriding importance of occupation as an indicator of socioeconomic status, considerable effort has been devoted to refining the conceptualization and measurement of occupation and to developing indices of occupational status. The focus here is on the latter, the development of measures of occupational status and prestige (North and Hatt, 1947; Blishen, 1958, 1967; Duncan, 1961; Nam and Powers, 1965, 1968; Siegel, 1971; and Treiman, 1977). These status and prestige scores rank large numbers of occupations along a single hierarchy. All of them have been widely used in social science research. Duncan’s Socioeconomic Index (1961; Featherman and Hauser, 1975) has been particularly influential in studies of the status attainment process in the United States.

Once occupation became accepted as an indicator of socioeconomic status, sociologists began to inquire as to the meaning of occupational status. Although there was general consensus that occupation is the best indicator of socioeconomic position, there has not been agreement as to the characteristics of occupations to be used in ranking them. As a result, and as noted above, a considerable number of scales and indices of socioeconomic and occupational status have been developed. They tend to be highly intercorrelated, but vary in concept and calculation procedures. Historically, one can categorize two major approaches. On the one hand occupational status has been interpreted as being synonymous with prestige, or position based on a subjective evaluation that individuals make of one another. On the other band, it has been defined in terms of more objective criteria such as the educational level required for the occupation and the income associated with it. One approach measures occupational status in terms of the prestige accorded a sample of occupations by a sample of individual raters. The second assigns a status score based on some form of objective measure such as ranking the average educational and/or income levels of occupations, A review of the development of the major North American socioeconomic and prestige indices follows along with a discussion of the conceptual and methodological issues treated in the papers included in the book.

Measures of Occupational Prestige

Prior to 1950 most studies of occupational status assumed 1) that the source of prestige of an occupation lay in the opinions of people rather than in the characteristics of the occupation, and 2) that people could estimate and articulate the prestige levels. (Counts, 1925; Anderson, 1927; Hartman, 1934; Nietz, 1935), Most of them also included only a small number of occupations and raters, The raters tended to be college students, school teachers, and school children, among whom there was basic agreement in the ranking produced. The occupations at the extreme ends of the scale were ranked with a very high degree of consistency. There was somewhat less agreement about those in the middle. In effect, most of the early studies showed that people could produce some sort of rank order of occupations when forced to do so with a relatively small number of occupations.

Perhaps the first sophisticated scale of occupational prestige is the one developed by the National Opinion Research Center (NORC). In the Spring of 1947 the National Opinion Research Center sampled a cross section of the American population to determine basic public attitudes towards occupations. In the Fall of that year Cecil North and Paul Hatt wrote a preliminary analysis of the data (North & Hatt, 1947). Hence the scale is sometimes referred to as the North-Hatt scale. Because that seminal study provided a starting point for most other research for the next three decades, it is examined in some detail.

A list of 100 occupations was constructed which was presumed to be representative of the entire range of legitimate occupations in the United States. In an attempt to keep the number within practical limits, the list was reduced to 78 occupations eliminating “women’s occupations” such as private secretary, dressmaker, trained nurse, etc. Before going into the field however, twelve occupations were added and the final list included 90 occupations.

A sample of 2930 people was asked to evaluate each of the 90 occupations according to their “own personal opinion of the general standing” of a person engaged in that occupation. A five point scale was provided with ratings of 1) excellent standing, 2) good standing, 3) average standing, 4) somewhat below average standing, 5) poor standing. An “x-rating” which meant “I don’t know where to place that one” also was shown to the respondents. On the basis of the overall replies, the responses were scored 100, 80, 60, 40 and 20 respectively. The “don’t know” responses were counted but were not assigned any particular value and were not entered into the basic computation of the score. The prestige rating for each of the 90 occupations was calculated by taking the mean score assigned to the occupation by the sample. The result was a scale ranging from 96 for Supreme Court Justice to 33 for shoe shiner. The occupations were then rearranged and ranked from 1 through 90 (Reiss, Duncan, Hatt and North, 1961).

The NORC study is a landmark study for a number of reasons. It included a sizeable number of occupations, but more important, it included a large enough sample of respondents so that some attention could be paid to the effect of age, educational level, etc. on ranking. Indeed, although there was a fairly high consistency of response, there were some significant differences. For example, professional occupations, particularly scientific jobs, were ranked higher by the college educated than by all others. Those who completed no more than eighth grade tended to rank skilled jobs such as carpenter, and operative work such as railroad conductor higher than others.

Another innovation in this study was the fact that respondents were asked why they rated some occupations as they did, They were asked to name the “one main thing” about the jobs they rated as “excellent” that gave them that standing. There was no predominant answer and, in fact, no majority response, About 60 percent of the responses were equally divided among answers such as “the job pays well,” “it has social prestige,” “it requires a lot of education or hard work,” “it performs a service to humanity” and/or “it’s essential”, Responses related to the income and education associated with the job were most frequently mentioned, It has been suggested that the apparent lack of agreement on what constitutes an occupation’s standing implies either that people have a single image of occupations to which they respond intuitively, or that various criteria for judgment are highly interrelated (Svalastoga, 1965).

The study contained some clear biases, however, About one-third of all the occupations listed were professional jobs, a high proportion compared to the number in the total labor force. Another point to be noted about this study is that the scientific jobs which were ranked so high were unfamiliar to many people, “Nuclear Physicists” were singled out to be defined by respondents in an attempt to infer how many people might have evaluated an occupation without being able to define it, Only three percent of the respondents were able to give a completely correct definition of nuclear physicists and 18 percent a partially correct one, (Svalastoga, 1965; 419)

Other shortcomings of the NORC scale were also noted (Hatt, 1950; Davies, 1952). One difficulty with occupational prestige measures discussed by Hatt as early as 1950 concerns the fact that when asked to rank occupations relative to one another, people are forced to place in a hierarchy occupations which would appear on the same group level if a point score rating system were used, Hatt developed the idea of parallel status ladders which he called situses. He identified at least eight separate situses: the political, professional, business, recreation and aesthetics, agricultural, manual work, military and service occupations. This concept of situs was an important addition to the literature and used in the 1963 re-test of the NORC occupational prestige scores to analyze changes in prestige for homogeneous groups of occupations. (Hodge, Siegel, and Rossi, 1964) Beyond that, however, it has not been widely used. The NORC scores were a major contribution to the study of occupational prestige, particularly after the 1963 replication demonstrated the stability of the ratings over time.

Twenty years after the original NORC study a prestige score for occupations was developed in Canada by Pineo and Porter (1967) based on the average evaluation given a sample of occupation titles by a national sample of the Canadian population. That score has been used in Canada both independently and in conjunction with the development of other objective measures (Blishen and Carroll, this volume).

In spite of conceptual and methodological difficulties of the prestige scales, they continue to be widely used. The most recent adaptation of the NORC prestige scores in the U.S. appeared in a doctoral dissertation by Siegel in 1971. He derived a new set of occupational prestige scores by combining the occupational prestige ratings collected in three separate studies, the NORC 1963 replication and two other studies conducted in 1964 and 1965 by Siegel and his colleagues. The occupational prestige ratings collected in the surveys were based on different samples, different lists of occupations and on somewhat different methodologies but all were combined into a single list. All the studies combined included over 600 occupational titles, only 412 of which were distinct, the others having been rated more than once. This overlap of identically worded titles permitted the transformations to occur. The overlapping titles were used to predict their prestige scores in the Hodge, Siegel & Rossi metric, Regression equations were used to provide what Siegel called “prestige scores in uniform metric” for all of the titles not rated in the Hodge, Siegel & Rossi study. Prestige scores were then assigned to each of the detailed occupational titles shown in the 1960 census of population.

Perhaps because of the recency of its development and the complicated nature of the statistical procedures, the Siegel prestige scores in uniform metric have not been as widely used as the original NORC scores or the 1963 replication. The set of scores provided by Siegel has an advantage over the earlier NORC work insofar as it provides a score for each of the detailed occupational categories listed in the census. Many of the difficulties inherent in the earlier studies remain, however. Indeed the errors may be compounded because so many of the scores were derived from mathematic estimates rather than from direct ratings by respondents.

Measures of Socioeconomic Status

Dissatisfaction with the limitations of the original NORC study of occupational prestige contributed to efforts to develop a socioeconomic index of occupations based on more objective criteria. The results include some strictly objective indicators (Nam and Powers, 1965), some which combined prestige and socioeconomic dimensions (Duncan, 1961), and some which did first one and then the other (Blishen, 1958 and 1967).

In Canada Bernard Blishen developed the first sophisticated strictly objective measure of occupational status when he ranked the 343 occupations listed in the 1951 Canadian Census on the basis of the mean income and education levels of incumbents in each occupation (Blishen, 1958). In the United States Otis Dudley Duncan devised a socioeconomic index for all occupations which integrated prestige scores with objective data on education and income for a wide range of occupations (Duncan, 1961), and Nam and his associates at the Census Bureau constructed both an occupational status score based on strictly objective criteria of income andeducation and a multiple-item index of socioeconomic status (U.S. Bureau of the Census, 1963; Nam, Powers, and Glick, 1964; Nam and Powers, 1965). These measures and others in the considerable body of literature which exists on this subject have been thoroughly reviewed and evaluated (Haug, 1977). A brief review of the origin and development of the three noted above is presented here in order to provide a context for the issues raised and the modifications in measurement proposed in the papers in this volume.

In the Duncan SEI, education and income were used to classify occupations according to socioeconomic levels. The underlying logic was essentially that suggested by Hatt much earlier. If education is a prerequisite for an occupation and income is a measure of the reward bestowed on it, then occupation may be viewed as an intervening activity between these two variables and hence a good single indicator of status (Duncan, 1961). The income and educational levels of each occupation in the 1950 Census were calculated according to the proportion of incumbents found toward the “upper end” of the income and education distribution for the total male labor force. These were defined as 1) the percent in each occupation with a high school education and 2) the percent reporting incomes of $3500 or more. Age-specific education and income patterns for each occupation were incorporated as weights with indirect standardization used to obtain age adjusted indicators for income and education. As in the earlier studies, the analysis was limited to the male labor force.

The prestige dimension of occupational status was incorporated into the SEI by using multiple regression analysis to arrive at estimates of prestige based on the 45 occupations in the original NORC study which precisely matched census categories. An estimate of the prestige of occupations was based on the linear multiple regression of the ...