Since the early 1970s, historians have begun to devote serious attention to the working class in the United States. Yet while we now have studies of working-class communities and culture, we know remarkably little of worklessness. When historians have focused on the Great Depression of the 1930’s, the narrowness of this perspective ignores the pervasive recessions and joblessness of the previous decades, as Alexander Keyssar shows in his recent book. Examining the period 1870-1920, Keyssar concentrates on Massachusetts, where the historical materials are particularly rich, and the findings are applicable to other industrial areas. (The unemployment rates that Keyssar calculates appear to be relatively modest, at least by Great Depression standards: during the worst years in the 1870’s and 1890’s, unemployment was around 15 percent.)

Yet Keyssar rightly understands that a better way to measure the impact of unemployment is to calculate unemployment frequencies—measuring the percentage of workers who experience any unemployment in the course of a year. Given this perspective, joblessness looms much larger.

Keyssar also scrutinizes unemployment patterns according to skill level, ethnicity, race, age, class and gender. He finds that rates of joblessness differed primarily according to class: those in middle- class and white-collar occupations were far less likely to be unemployed. Yet the impact of unemployment on a specific class was not always the same. Even when dependent on the same trade, adjoining communities could have dramatically different unemployment rates. Keyssar uses these differential rates to help explain a phenomenon that has puzzled historians: the startlingly high rate of geographical mobility in the nineteenth-century United States. But mobility was not the dominant working-class strategy for coping with unemployment, nor was assistance from private charities or state agencies. Self-help and the help of kin got most workers through jobless spells.

While Keyssar might have spent more time developing the implications of his findings on joblessness for contemporary public policy, his study, in its thorough research and creative use of quantitative and qualitative evidence, is a model of historical analysis.