The American Occupational Structure

When Blau and Duncan's The American Occupational Structure was first published in 1967, long-time students of social mobility and social stratification felt that they had been rendered obsolete. This thick book is a difficult read by any standard, but introduction of unfamiliar statistical techniques, especially path analysis, made it impenetrable to scholars not on the methodological forefront of disciplines in which they had done creditable research for decades. Blau and Duncan made things worse by failing to explain that path analysis, which quickly became a methodological obsession in sociology, was just another effort to make regression analysis more informative. Today, good-quality introductory statistics texts provide enough instruction in path analysis to enable a neophyte to understand Blau and Duncan's model of individual mobility and, given a suitable data set, to replicate it with user-friendly software such as SPSS. Brilliant though they were, I think Blau and Duncan made a serious theoretical error when they decided to treat structural mobility as a nuisance that merely blurred their focus on individual mobility. A good case can be made that level of educational attainment was the strongest predictor of offsprings' socioeconomic status primarily because the occupational distribution was opening at the top, with more and more good jobs being created to absorb the output of high schools, colleges, and universities. Education worked because there were occupational places to put educated people, not because we live in an intrinsically meritocratic world. Today things have changed markedly. No one saw it coming, certainly not Blau and Duncan, but since the early 1970's the creation of good jobs -- prospects for upward mobility through investments in education -- has been undercut by out-sourcing, down-sizing, internationalization, and technological innovation. All this in pursuit of reduced labor costs. The efficacy of investment in education as an agency of upward social mobility has not stood the test of time. Writing in the 1960's, however, Blau and Duncan saw the same happy future has everyone else: more good jobs, more opportunities for upward social mobility, unparalleled prosperity, and the professionalization of everyone. Near the end of their book they attributed their optimism to the ascendance of the norm of universalism. As with everyone else, they where caught unaware by dramatic structural changes. I've read this book twice, and learned from it each time. The most important lesson, however, is one that Blau and Duncan missed: macro-level contextual factors inevitably take precedence over individual-level characteristics. Macro-level contextual factors, however, commonly defy inclusion in statistical models.