Bismarck to no effect: Fertility decline and the introduction of social insurance in Prussia

Abstract

Economists have long argued that introducing social insurance will reduce fertility. The hypothesis relies on standard models: if children are desirable in part because they provide security in case of disability or old age, then state programs that provide insurance against these events should induce couples to substitute away from children in the allocation of wealth. We test this claim using the introduction of social insurance in Germany in the 1880s and 1890s. Bismarck’s social-insurance system provided health insurance, workplace-accident insurance, and old age pensions to a majority of the working population. The German case appeals because the social insurance program started on a large scale and was compulsory for covered classes of workers, and because fertility in Germany in this period was still relatively high. Focusing on the state of Prussia, we estimate differences-in-differences models that ask whether marriage and marital fertility reacted to the introduction or extension of the main social insurance programs. For Prussia as a whole we find little impact.