Like many other industrialized countries, Germany experienced two pronounced stages of fertility decline in the last 150 years. During the first fertility decline, total fertility rate fell from about 5.5 children per woman in 1885 to less than 2.5 in the 1920s. The beginning of the second fertility decline is usually associated with the introduction of the contraceptive pill in the middle of the 1960s. Since then, the total fertility rate in Germany has decreased to less than 1.5 children per woman. There are many explanations for this development. Scholars consider among other things the improvements in general education that led parents to invest their resources in better schooling of fewer children, the increase in female labor participation due to rising wages during economic growth, or the changing role models of women in the course of female emancipation. Beatrice Scheubel argues in her dissertation that the introduction and extension of the social security system played a much bigger role in Germany's long-term fertility decline then it is usually assumed. Her main argument is that the Bismarckian pension system decoupled the motive to provide for old age from the decision to have children. In economic terms, compulsory investment in public insurance crowds out any type of investment in private insurance: Dependent employees that are subject to social insurance contributions reduce both private savings for old age and the number of children who were traditionally supposed to support their disabled parents. In Scheubel's opinion, the latter substitution effect will be intensified by the externalities of a pay-as-you-go system in which the children of other people have to pay for the pensions of childless people. That is why, in the long run, the Bismarckian pension system might have undermined its own financial sustainability by creating incentives to have fewer and fewer children. To solve this problem, Scheubel's rather radical policy advice is to reestablish the link between old- age provision and fertility by making the level of future pensions contingent on the number of children an individual has.

To test her hypotheses empirically, Scheubel relies on cross-sectional observations about the demographic development in the 41 regions of the German Empire between 1878 and 1914—the period in which the Bismarckian social insurance system was step-by-step implemented. The social insurance system's three pillars (health insurance, accident insurance, and pension insurance) came into effect in 1884, 1885, and 1891 respectively, Scheubel also consider the pension system's transition to a pay-as-you-go system in 1900, the tightening of the law on child labor in 1903, and the completed review of the code of practice of the Regional Insurance Agencies in 1904 as major changes in the institutional environment. Her identification strategy makes use of the fact that the extent to which the population was covered by the pension system differed across provinces and over time. Based on this observation she employs a difference-in-difference approach in which the treatment group consists of all provinces where the share of insured people was higher than the sample mean plus one standard deviation. Controlling for various socioeconomic factors Scheubel concludes that up to a third of the observable decline in crude birth rates reflected the introduction and extension of the pension system. Scheubel further claims that industrialization (which she measures by the share of population working in mining) had an independent, negative effect on fertility too (p. 158). This deduction might be wrong. Tobias Jopp (Insurance, Fund Size, and Concentration. Berlin: Akademie Verlag, 2013) has shown that German miners` had been covered by a sector-specific pension system (very similar to the Bismarckian one) since the middle of the nineteenth century. They therefore faced incentives to reduce fertility long before other industrial workers. That is why the variable “share of population working in mining” measures not only the degree of industrialization but also identifies regions with a very early treatment.

In contrast to today, in the late nineteenth and early twentieth centuries, the most important policy parameters of the social insurance system such as coverage, contribution rates, or benefits varied considerably across German regions and insurance carriers. Moreover, German statistics offers considerable information on the demographic and economic regional development needed to exploit this variation statistically. Thus it is surprising that there has been so little empirical research on the German welfare state's impact on individual behavior. From this perspective, Scheubel's dissertation is groundbreaking work.