When it emerged as a discipline a century ago, the historical and comparative description of political institutions dominated political science. The self-consciously rigorous analysis of behavior inspired by psychology and the physical sciences began to displace historical analysis in the 1920s. By the 1950s and 1960s the behavioral revolution came to dominate the discipline’s research agendas. Leading practitioners acknowledged that the discipline’s tone was predominantly ahistorical.

Walter Dean Burnham (1965, 1970) demonstrated that non-voting and various measures of aggregate electoral instability increased between 1900 and the late 1920s, declined somewhat during the New Deal and its aftermath, and increased thereafter. To Burnham, the behavior of these measures implied a similar evolution in the partisan strength of the electorate since individuals without strong partisan ties are characterized by lower levels of turnout and more inconsistent turnout and partisan choices, both over time for the same office and across different offices in the same year. In short, he argued that the entire twentieth century, save for the New Deal, was marked by an “onward march of party decomposition.”

Population has always been recognized as a fundamental resource of the state. In nineteenth-century Europe, however, the relationship between population—a body of people capable of working, fighting, growing, or moving collectively—and the state underwent a significant transformation as national governments found it increasingly more important to know their populations quantitatively. Satisfaction of this need required the creation of national institutions to carry out censuses and monitor the vital statistics and migratory movements of population.

Lydia J ——, daughter of a widow with five children, was admitted to the Boston Female Asylum, an orphanage run by women, in 1826, at the age of four. When she was 11 she was apprenticed to a Boston physician and his wife. On her eighteenth birthday, Lydia agreed to remain with the family as a salaried servant, but six months later she left “to learn the business of dressmaking.” Lydia’s specialized training in a needlework trade supported her until her marriage, four years later, and in all likelihood at later periods in her life.’

Married women’s entrance into the market economy proceeded at a slow but steady pace between 1890 and 1910. That, at least, is the impression given by conventional census measures of the percentage with “gainful occupations,” which practically doubled in both the United States as a whole and in the heavily industrialized state of Massachusetts (see Table 1). This impression is misleading on at least two counts. Declines in self-reporting and enumerator bias may have overstated the increase in married women with gainful occupations. More important, dwindling opportunities for informal market activities, such as industrial homework, provision of services to boarders, and participation in a family farm or enterprise, may have countervailed increases in formal market participation. In Massachusetts, at least, married women’s specialization in non-market domestic labor probably increased.