|
Article Excerpt
Les donnees sur l'espace urbain recueillies a Kitchener-Waterloo, en Ontario, jumelees avec celles d'une premiere etude realisee en 1975, sont utilisees comme point d'observation pour analyser a nouveau la stabilite historique du prestige professionnel. L'article suggere que la structure de la courbe de distribution du prestige a ete negligee en faveur de statistiques decrivant la stabilite du classement par ordre de grandeur et que le changement historique depuis environ 1975 est, du point de rue qualitatif, different de celui des periodes anterieures. Notre hypothese est que la courbe de distribution du prestige professionnel est devenue plus constante et que le classement par ordre de grandeur s'est nettement deplace.
STUDENTS OF SOCIAL STRATIFICATION trained during the 1960s have imprinted into their synapses the results of the U.S. National Opinion Research Center's (NORC) 1963 replication of the famed "North-Hatt" study of 1947. Contrary to intuitions that might have been held about social changes between the eras of Truman and Kennedy, the replication showed minimal alteration in the status ladder of occupations over this 16-year span (Hodge, Siegel and Rossi, 1966: 334). Scores from the 1947 scale correlated 0.99 with the 1963 set. So decisive a figure was believed to hold great importance for theories of social inequality and for research on work, occupations and social mobility. The ensemble of occupational mobility studies during the 1960s and 1970s, for example, rested on the premise of a stable occupational stratification within which intergenerational movement could accurately be gauged (Blau and Duncan, 1967: 119).
A number of ideas about industrial society came together and fermented within the vat of occupational prestige research. From classical sociological theory, Durkheim ([1893] 1964) had, at the close of the 19th century, emphasized the growing "division of labour." In this new economy of the 1890s, workers increasingly were performing specialized, differentiated tasks. At some level, people were conscious of the interrelatedness among specialists and that was one component of Durkheim's organic solidarity. The concept of "function" was frequently referred to in The Division of Labour, and this term was adopted by Davis and Moore (1945) in their provocative structural-functionalist paper on reasons for social inequality. They argued that the functional importance of occupational tasks cut across cultures, including cultural change over time, thus anticipating the empirical findings about prestige ratings in the 1960s. (That assumes, of course, that the prestige ratings echoed functional importance, a controversial point for critics of Davis and Moore.) Economists such as Kerr et al. (1964) emphasized the powerful standardizing force of industrialization, compared with earlier forms of society. Talcott Parsons (1951), by then nearing the height of his powers as a leading structural-functionalist theorist, stressed the high value-consensus in industrial society. Data on occupational prestige scales, with high consensus at least as computed at the aggregate-data level in comparison across time, seemed congruent with that prediction too. In short, circa 1960 sociologists believed that a natural status ordering for occupations existed and that members of the public knew the Order. These points were forcefully argued by Treiman (1977) in a study of the cross-national stability of occupational prestige. The "Treiman constant" of unvarying occupational prestige has become one of the most secure pieces of knowledge in the field of social inequality (Hout, 2002). Even though today's theoretical climate has shifted greatly since the 1960s, the Treiman constant is so entrenched as still to seem obvious.
An argument for re-examining the stability and the dynamics of occupational prestige comes from the 1989 replication in the U.S. of earlier NORC work on prestige scaling. In this period following both the disruptions of the OPEC crisis in oil supply in the mid-1970s and the recession in the early 1980s, Nakao and Treas (1994) detected some stirring in the hierarchy. Their scale correlated .97 with one collected 25 years earlier, indicating slightly greater slippage over time than the r = .99 for the 16 years between NORC 1947 and 1963. The mean occupational prestige score rose slightly between 1964 and 1989, while the standard deviation for the set of scores declined. A "startling number of occupational titles saw their prestige change" (Nakao and Treas, 1994: 17), especially with respect to gains from titles of low status in the 1960s data, but, in deference to the Treiman constant, the authors nevertheless observed that their comparison with 1964 data showed "a remarkable demonstration of the overall stability in the prestige order ..." (Nakao and Treas, 1994: 15).
It was the sense that change may be underway in the prestige ratings of occupations that prompted work on a local scale in Kitchener-Waterloo. A prestige survey had been collected there in 1975, and this work was replicated in the fall of 2000, the quarter-century anniversary. The following pages report on a re-evaluation of the stability of prestige using these data, combined with re-analysis of the earlier U.S. comparisons covering 1947-1963 and 1964-1989.
Two Assumptions and an Hypothesis
The analysis to be reported was guided by two assumptions and one hypothesis. The first assumption is that foreclosure on the question of change or cross-cultural difference in occupational prestige took place in the 1960s and 1970s, due to the already-noted theoretical orientation of the time. It is striking, looking back at earlier literature prior to the 1963 replication, what strong counterfactual intuitions researchers held about historical change in occupational prestige hierarchies. In the 1930s, Nietz (1935: 454) framed a replication of the classic Counts study of 1925 around the issue "has the depression had any effect on the social-status ratings?" Nietz believed that such changes would be detectable over three data sets collected in 1928, just prior to the 1929 crash; 1932 "which was during the heart of the depression"; and 1934, which from his historical vantage point, "represents the period of recovery" (Nietz, 1935: 455). A decade later, the Canadian pioneer in occupational prestige research Jacob Tuckman (1947: 71) recalled in connection with the Counts (1925) scale that "the question was raised regarding its validity in 1946.... " Even much later on, at the height of the structural-functionalist interpretation, Paul Siegel mused over how counter-intuitive was the seeming longitudinal stability of prestige scores: "This temporal stability of the prestige structure is especially striking when one considers the vast changes that have occurred in the occupational structure ..." (Siegel, 1971: 3). In hindsight, the period 1947 to 1963, the crucial years for which the best trend data were available when Siegel was writing, looks like a stable interval of consistent upward movement in the business cycle of mature industrial societies.
In a sense, statistical analysis of prestige data was unwittingly being set up to deliver results consistent with structural-functionalist theory. The focus on Pearsonian correlations for scores already aggregated across samples of respondents has received frequent mention (e.g., Coxon and Jones, 1978: 38-39; Guppy, 1982; Guppy and Goyder, 1984). As well, structural-functionalists were not very interested in the shape of the distribution of prestige scores, because assessment of the societal consensus about the rank ordering was the main issue. Concern with the shape of an inequality distribution comes from different subcommunities of social stratification researchers--from poverty and income inequality research, for example. So long as the "profile correlations," as Coxon and Jones (1978: 38) called them, are recognized as readings from a sociological micrometer in which shifts of a point or two may be important, they retain some usefulness. Strength of correlation should not, however, be the only criterion for assessing change in prestige. Along with correlations, measures of distribution are considered more fully herein than is the normal practice in the analysis of data on occupational prestige.
The second assumption is that the social change in the final quarter of the 20th century was qualitatively distinct from events between mid-century and the three-quarter mark. That is not to minimize the pace of events earlier in the 20th century (Caplow, 1991: 49) and, indeed, in the previous one (Schlesinger, 1922: Chap. 1). Nevertheless, works such as Toffler's Future Shock from 1970 and his later Third Wave (Toffler, 1981) were insistent that new patterns were emerging; this view is echoed in later examples such as Naisbitt (1982) and Celente (1997). New "defining technologies" (Bolter, 1984: 11) had arrived with the microelectronic chip, which shook up the skill content of jobs. Some argued that the long-run business cycle was entering a new phase (Mensch, 1979). Careers became less permanent between 1975 and 2000, but...
|
