Archibald O. Haller

The hypothesis of intersocietal similarity in occupational prestige hierarchies.

An influential hypothesis in stratification theory holds that industrial society produces a unique occupational prestige hierarchy. So it is thought that as the previously non-industrialized nations develop industry their occupational prestige structures become more and more like those of the industrialized nations.2 Euro-American nations, especially the United States, are viewed as the most representative of this type. The data used to support this notion are the high correlations (+.90 or more, on the whole) among the average evaluations people of various societies make of certain occupational titles.3 But this evidence is weaker than it appears to be. Correlations such as those observed could arise when in fact there is only a slight similarity in the occupational prestige structures of any pair of societies. Moreover, no one seems to have produced evidence supporting the notion that industrialization produces the similarities that may exist. We shall present several factors that may influence the observed correlations among occupational prestige structures. Some of the factors should lead to overestimating the correlations and some to underestimating them. In the net, these suggest that the similarities may be less than has been supposed. Following this we shall present data suggesting that the complexity of the division of labor (urbanization) in general rather than industry in particular may be responsible for the reported similarities.

An Empirical Evaluation of a Scale to Measure Occupational Aspiration Level.

Abstract Data from 34,118 American high school students are used to evaluate the Occupational Aspiration Scale (OAS). The OAS is successively reevaluated for each of 16 subsamples (cells) generated by cross-classifying respondents by grade in school (9–12), sex, and socioeconomic status (SES). In each cell the OAS is found to be essentially unifactorial, and that factor is identified as level of occupational aspiration (LOA). The reliability of the OAS is slightly lower among females ( r kk = .681) than among males ( r kk = .756); it does not vary appreciably by grade or SES. The mean scores are lower for youth from low SES families than for those from high SES families, in accord with previous research. Mean OAS differences due to sex and grade are small. No important differences by age, sex, or SES are found in the standard deviations of the test scores. This and previously published data from small, local samples indicate that the reliability and validity of the OAS are sufficient for research on high school youth of both sexes and from both higher and lower SES levels.

On Lin and Yauger's Cross-National Comparison of the Process of Occupational Status Achievement

As Lin and Yauger (1975) note, comparative studies of status-attainment processes involving data from less developed countries (LDCs) are badly needed. While we think that they were correct in attempting to fill this void, we take serious objection to some of the data they present, and we question the tenability of their conclusions. Briefly, what they tried to do was to estimate national status-attainment parameters for two LDCs (Costa Rica and Haiti) and two more highly developed countries (MDCs: the United States and Great Britain). So far as we can tell, the data for the MDCs (Treiman and Terrell 1975) are appropriate. It is the adequacy of the LDC data which we doubt. Our misgivings pertain to three general problems: (1) rural sampling bias, (2) high nonresponse rates, and (3) occupational status scoring of low validity. 1. Rural sampling bias. Both of the LDC samples were drawn from rural areas of the respective countries. Surely a disproportionate number of the better educated, more prestigious, and more influential people in Costa Rica and Haiti live in the cities. So the higher reaches of the status system were almost certainly undersampled. Clearly, the only appropriate way to obtain national level status-attainment parameters is to base them upon representative national samples. There seems to be no way by which the parameters estimated from these two samples can legitimately be considered valid estimates of their respective national parameters. It follows that the comparisons drawn with the MDCs are untrustworthy. 2. High nonresponse rates. Compounding the first problem, in both LDC samples many cases were dropped from the analysis because of missing data. The intergenerational mobility analysis for Haiti was based upon only 37% of the original sample (199/544). For Costa Rica the corresponding figure is 59% (337/570). In Lin and Yaugers regression analysis the percentages are 86% (455/544) for Haiti and 59% (336/570) for Costa Rica. It is difficult ifnot impossible to assess the error in the estimates of the parameters which is due to such underrepresentation. But surely, even if they had used adequate national sampling frames, nonresponse rates as high as some of the above would make their estimates dubious. 3. Validity of occupational status scores. Both the nominal occupational codings and the numerical scores assigned to them by the SIOP (Standard Index of Occupational Prestige; Treiman, forthcoming) are probably inaccurate, for two reasons. First, from our rural Brazilian work (e.g., Haller,

Occupational Wage Differentials Among University Educated Technical Personnel in a Developing Economy.

Abstract Data taken in 1970–1971 from three university-trained occupational groups (2796 engineers, 1430 economists/business-administrators, and 162 basic scientists) in Sao Paulos manufacturing industries are used in a path analysis to draw interoccupational comparisons concerning the antecedents of occupational wage differentials. The workers total hourly wage is the dependent variable. A new variable, occupational influence level, is employed as immediately antecedent to wage, as in job experience (years in the present job). Years of advanced education, age, and seniority in the firm are treated as exogenous variables. Similarities and differences among occupations are discussed. As a whole the analysis illustrates a strategy of comparative occupational analysis.