skip to primary navigation skip to content
 

Economy, Gender, and Social Capital in the German Demographic Transition

Economy, Gender, and Social Capital in the German Demographic Transition

This research project, with Sheilagh Ogilvie as principal investigator, uses exceptional data and new econometric methods to analyse long-term, micro-level demographic decision-making in a region of southwest Germany between 1558 and 1914. Associated investigators are Tim Guinnane and Richard Smith. Research officers are Markus Küpker and Janine Maegraith. The project is supported by a generous grant from the Leverhulme Trust (F/09 722/A).

A clear understanding of the causes of fertility variation and decline would exercise a huge impact on human well-being and economic growth. Many of the unresolved puzzles about fertility can be addressed by adopting a deliberately interdisciplinary approach. Economics can provide econometric techniques and microeconomic models, demography can provide an understanding of family-building strategies and interactions among demographic variables, and sociology, anthropology and history can provide insights into the role of institutions, gender, and social networks. This project seeks to combine the strengths of these different disciplines to advance our understanding of the determinants of fertility change over the long term.

Background to the project

Fertility has a crucial impact on poor and rich countries alike. In poor countries, high fertility puts pressure on environmental resources, human capital investment, women's well-being, and economic growth. In rich countries, low fertility challenges welfare and pensions systems. The demographic transition from high to low fertility is still seen as a major transformation in human history, with dramatic effects on nearly all aspects of life. Yet social scientists still do not fully understand why fertility levels vary so greatly across societies, social groups, individuals, and time-periods. This project addresses these open questions using an unusually rich data source, innovative econometric techniques, and a deliberately interdisciplinary approach.

Objectives of the project

The project uses new data and new methods to address unresolved questions about the demographic transition. The only completed experience of such transition is contained in the history of present-day developed economies. This project selects a country - Germany - with exceptionally rich data from the sixteenth to the twentieth century. Hitherto, most studies of the demographic transition have focused on the nineteenth and twentieth centuries, adopted an aggregative approach, and used only a narrow range of explanatory variables. Nevertheless, they have given rise to important insights about patterns of fertility decline, from which wide-ranging lessons have been drawn for modern developing economies. But they have also left unresolved puzzles. First, what determined the huge variations across time, space, social strata, and individuals in fertility levels, both before and after transition? Second, what more we can learn by including key variables such as gendered work patterns and the role of social networks in creating and enforcing demographic norms? Our objective is to answer these questions.

Methods used in the project

The unresolved puzzles about the demographic transition can be addressed fruitfully by using extraordinarily rich German documentary sources from the sixteenth to the twentieth centuries. The project is engaged in constructing an innovative database using parish registers, censuses, tax cadastres, citizen and office-holder lists, guild account-books, and church court minutes. Very valuable research on the demographic transition has been undertaken using population registers, but these sources are usually best for cities and the nineteenth century, and even they do not contain all the information we will use for our study.

The project has selected three communities with different economic structures, located in the southwest German territory of Württemberg. Using parish registers, marriage, baptism, and burial records are entered into a database for each community. These data are then used to carry out family reconstitutions for all three communities. Most family reconstitution studies simply stop there, often because of source limitations, but our project uses record linkage to attach to individuals' demographic characteristics a wide array of additional variables. These include occupational background (for females as well as males), wealth (for both sexes), farm size, arable and pastoral landholding, value of craft workshop, household structure, literacy, religious background, and membership in social networks (community citizenship, guild mastership, local office-holding). The database is then used to explore the implications of socio-economic changes not usually considered in studies of the fertility transition.

Our sources allow us to surmount two important methodological limitations in previous studies. First, family reconstitution is often criticized for lacking independent population counts and excluding many individuals from analysis. But our database includes census-type listings, enabling us to determine total population and include individuals excluded by reconstitution. Second, our sources provide detailed information on wealth, landholding, occupation, gendered work patterns, and social networks.

The project also uses statistical and econometric techniques not available to earlier studies of the demographic transition. The main tool is event-history (or duration) analysis and related methods such as 'count data' regressions. Our data offer unusual opportunities for such techniques, because long observation allows us to understand the determinants of natural fertility prior to the demographic transition. Such models have been used before, but recent improvements make them more useful for historical demography. The ever-present unobserved heterogeneity that can bias event-history estimates can be dealt with through semi-parametric models that do not require strong distributional assumptions. New approaches exist to contend with the 'reflection problem' posed by efforts to test econometrically the effects of social interactions on individual behaviour. Using these tools, we will be able to let the data speak on these important issues in a precise and robust way.