Prior studies of the political economy of the welfare state have focused on programs that automatically respond to demographic or economic pressures: students cause educational expenditures, unemployment causes unemployment compensation, poverty causes public assistance, and persons over 65 cause old-age pensions. At the very least, such demographic pressures largely determine short-run fluctuations in such spending. However, active labor market policy (ALMP) – direct government intervention into labor markets to decrease unemployment through job-placement, job-training, and job-creation programs – is preponderantly discretionary and not dominated by demographic fluctuations. Thus, ALMP can be either a major recipient of fiscal resources – Sweden spent 2.10 percent of GNP in 1984 – or relatively ignored by politicians and budget makers – the United States spent only 0.17 percent of GNP in the same year (see Table 3.1 for additional years and countries). As testimony to the discretionary character of ALMP, countries as similar as West Germany and Austria have ALMP/GNP figures that differ by factors of 4 to 10.

Sweden, West Germany, and the United States effectively represent this range of ALMP efforts with high, moderate, and low ALMP expenditures, respectively. Moreover, they are also good examples of Esping– Andersen's social democratic, conservative, and liberal regime types (1990, pp. 69–78). “Regimes” represent basic characteristics of the welfare state system – institutional arrangements, program rules, expenditure developments, and problem definitions – and to Esping Andersen, they also represent basic aspects of political economy (p. 80).