The FaMiGrowth Project

Over the last decades, theories of economic growth have expanded their scope beyond the traditional focus on capital accumulation and technology to emphasize the role of other factors like migration and fertility (Aghion & Howitt, 1998 and Galor, 2011). These two factors have, however, been explored separately, without questioning their interactions. Yet, family formation and migration decisions are strongly intertwined. For instance, the timing of family formation may affect whether an individual migrates or not, and if so where they move or even whether they do so alone or with their spouse and children. Similarly, migration affects women’s fertility, both because husbands may be absent or, when the family moves together, because of the impact of a new location on child-bearing norms or child mortality.

​

Despite of the considerable work on migration and families by historians and demographers (e.g. Green, 2002), we do not yet have a systematic economic analysis of how location and family-formation decisions interact at the microeconomic level and what these interactions imply for economic development. In particular, we lack a clear understanding of (Q1) how households members’ decisions to migrate in search of better economic opportunities depend on the ‘stage’ of family formation in which they are, (Q2) how these decisions were shaped by the major economic and social transformations that Europe experienced since the mid-17th century, and (Q3) how family formation and migration decisions contributed to economic growth and development in Western Europe over the last three centuries. 

 

Our project aims at answering these questions by providing a clear understanding of the interconnections between individual decisions about family formation, migration and economic activities. The relevance of this approach is twofold. First, quantifying the impact of migration and family formation on European growth requires understanding how these two aspects interact; second, examining those interactions in a historical context can help economists formulate policy recommendations to public stakeholders who need a long-term vision to take informed decisions on migration and social policies today. 

 

One of the main challenges of this research agenda is considering the interconnection between the various phenomena, which makes it difficult to identify and quantify causal links. Our key contribution is the identification of causal relationships by (1) developing a new integrated dataset on family formation and geographical mobility in Europe over the period from 1640 to 1940, and (2) establishing a rigorous identification strategy to understand how family and migration decisions shape economic development. The dataset we intend to develop will rely on crowdsourced genealogical data well suited to study demographic dynamics. These data come from www.geneanet.org and www.geni.com. Preliminary exploration of the raw data indicates that we will be able to follow 700 to 800 million European individuals from birth to death. We will establish their marriage and fertility history through time and space, and therefore their main migratory movements. We will use detailed information to match individual data, first across Geni and Geneanet and secondly between these datasets and nominative census data for a sub-sample of countries in the 19th century. This will allow us to identify the occupation of individuals and to observe migratory episodes between genealogical events. We would hence be able to construct the first pan-European individual panel dataset, following not only individuals over their lifetime but also their lineage across several generations, the European Historical Family Database.

 

We will develop a global identification strategy based on three powerful elements: (i) the granular observation of individuals through space and time will allow us to use at-the-frontier identification techniques relying on exclusion restrictions provided  by quasi-natural experimental settings (Boustan et al., 2012); (ii) the use of family fixed effects will enable us to account for unobserved factors correlated with family origins and genetics of individuals belonging to a common lineage; (iii) the use of indirect inference and especially the Simulated Method of Moments (SMM) will make it possible to capture how factors we cannot directly observe -like the intention to migrate and general equilibrium effects- shape the transmission of the microeconomic behavior measured in WP2 to WP4 to economic growth in Europe.