Beginning with time series is perhaps the best point of entree for the new methods of comparative research because its chronologically ordered data come closest to familiar notions of history. Yet rather than simply describing “one damn year after another,” time-series analysis tests for the relationships between variables. When time-series analysis is done in a sensitive, theory-driven, historically grounded manner, it can ring true with conventional historical analysis by enriching and being enriched by historical inquiry.

THE SCOPE AND LOGIC OF TIME-SERIES ANALYSIS

Time series, as currently used, most often cover variables in comparative political economy in nations or states from 20 to 100 years. Most studies of the welfare state focus on state expenditures, but participation in programs or demographic outputs have also been used (Hage and Hanneman 1980; Griffin, Devine, and Wallace 1983; Hicks 1984; Korpi 1989; Hage, Hanneman, and Gargan 1990). Political scientists have pursued the election cycle and macroeconomic policies (Hibbs 1977; Monroe 1979, 1984; Hibbs and Vasilatos 1981; Beck 1982; and Hibbs, Rivers, and Vasilatos 1982). And sociological and economic researchers have also focused on strikes (Shorter and Tilly 1974; Snyder 1975; Edwards 1981; and Isaac and Griffin 1989). Time-series methods, however, can be applied to any social or political variable that can be regularly measured over a given time frame (i.e., over standard calendar units of years, quarters, months, etc.).