What Ten Years of Data Tell Us About Labor Market Projections of Bowen and Sosa Introduction: Labor Market Models, Data, and Critiques In 1989 William Bowen and Julie Sosa published Prospects for Faculty in Arts & Sciences, attempting to project balance of faculty supply and demand over a 25-year span, from 1987 through 2012. The book was last of a number of similar reports and studies (e.g., Bowen & Schuster, 1986; Lozier & Dooris, 1987; McGuire & Price, 1989), all of which predicted, from depths of a recession in academic labor markets, that a rapid turnaround was in sight. Although supply of new PhDs had far exceeded demand for new faculty throughout eighties, these works were intended as wake-up calls to fact that large numbers of faculty who had been hired during higher education's expansion in sixties would soon retire, creating a swelling demand for new faculty that would be difficult to fill at then current rate of PhD production. The intent was to spur action on part of institutions and policymakers to increase flow of students into graduate programs and rebuild supply of faculty. All of these works relied on fixed-coefficient models for calculating future supply and demand. That is, they projected current trends mathematically, without attempting to account for feedback loops or adjustment mechanisms of marketplace. Yet, there is a range of different levels of sophistication. Whereas much of previous work had been based upon relatively simple assumptions, Bowen and Sosa took a much more detailed approach, making Prospects for Faculty extremely rich in hard data. Indeed, its authors had access to levels of detail in their data that had been almost unimaginable just a few years earlier. Bowen and Schuster (1986), for example, were unapologetic in their work's aggregation of entire national academic labor market, explaining that the necessary data...[to break totals down into components such as geographic regions, types of institutions, or academic disciplines] ... are not available and we do not have resources to gather (p. 195). In their defense, Bowen and Schuster were primarily concerned with quality of faculty, not quantity. Bowen and Sosa, on other hand, focusing almost entirely on quantity, were able to obtain finer data and disaggregate their figures, both for faculty and for graduate degrees, by individual academic fields and by type of institution. Although their model required them to make many assumptions about future values of various factors and coefficients, majority of these assumptions are carefully supported by detailed analyses of data trends. Out of a universe of, at one count, 96 distinct scenarios that might be played out over quarter century of their vision, Bowen and Sosa eventually narrowed their projections down to four main models that they considered most likely to occur. With only slight variation, each of those models projected significant shortages of PhDs in academic labor market arising as early as 1992 and by 1997 at latest, reversing an oversupply of PhDs during years leading up to that point. Unlike their contemporaries, however, Bowen and Sosa did not attribute projected shortages to any sudden wave of retirements, but rather to combined effects of slight recoveries in postsecondary enrollments (after a drop in early nineties), stagnant production of PhDs (with actual declines in percentage of new PhDs who choose to pursue academic careers) and most importantly, a large, steady outflow of faculty from combined life processes of career changes, retirements, and deaths. Almost immediately after its publication, some researchers began to criticize accuracy and reliability of projections made in Prospects for Faculty (Blum, 1991; Gill, 1992), even as others rushed to proclaim arrival of first faculty shortages years ahead of when Bowen and Sosa had predicted they would begin (EI-Khawas, 1990, 1991). …
Read full abstract