In an influential body of work extending across more than three decades and drawing on data from the United States, Poland, Japan and the Ukraine, Melvin Kohn, Carmi Schooler and their associates have found that cognitive capacities are affected by experiences on the job; specifically, that working at a complex job improves cognitive functioning. These findings anticipate and parallel research on the relationships among social integration, leisure-time activities and cognitive functioning among the elderly. This paper tests the Kohn–Schooler hypothesis using different measures, models and data. Specifically, we estimate models of the reciprocal influence of work complexity and cognitive functioning at ages 53 to 54 and 64 to 65 among women and men who graduated from Wisconsin high schools in 1957. Even when adolescent academic ability test scores and high school rank have been controlled, we find moderate effects of the complexity of work on abstract reasoning ability at ages 53 to 54 (in 1993). These effects are similar among women and men and are robust to reasonable assumptions about the unreliability of measurement of adolescent academic ability. However, there were no such effects, either in 1993 or in 2004, among individuals who were still working at ages 64 to 65. Thus, the full set of findings provides only limited support for the Kohn–Schooler hypothesis.
