Gregory Clark
DKK 10 million. Funded by Danmarks Grundforskningsfond (01-08-2023)

Title: Danish National Research Foundation
All work on social mobility rates across time, and across countries, is bedeviled by measurement errors.  Measures of social outcomes are loose, imprecise, and of varying precision over time.  This measurement error systematically leads to overestimation of social mobility rates.  But it also makes comparing social mobility rates across time, and across countries, fraught with potential biases.

In his published work, and forthcoming book on England 1700-2022, DNRF Chair Clark has devised innovative means to correct those measurement errors.  These methods involve observations on multiple relatives and multiple generations.  Applying these methods leads to the conclusion that social mobility is extremely slow in England, has not changed its rate over more than 300 years, and has been unaffected by modern social interventions. 

The key research idea of this DNRF Chair is to apply Clark’s innovative techniques to the rich database of family connections and social outcomes over multiple generations in Denmark and Norway 1790-1920 that is being assembled by the HEDG research group (led by Professor Paul Sharp) at Southern Denmark University - the Human Capital of the Nordic Countries (HCNC) database.

Based as it is on High School and University Yearbooks, the HCNC database is a record of an upper social group in earlier Denmark, High School and College Graduates.  Further, since the source records also High School Grades we have a quite fine calibration of academic abilities of the population for those born in the years before 1920.  The database also identifies, implicitly, those with less than High School education.  The question of interest in this DNRF Chair is how fast has been the downwards social mobility of this nineteenth century educational elite in the nineteenth century, the twentieth century and now?  And how fast has been the upwards mobility of the less educated group across the same intervals?

Under the DNRF Chair the HCNC database will be extended to link families all the way from 1790 to the present.  We will thus be able to observe register data on social outcomes for people of all degrees of family connection: siblings, first cousins, second cousins, third cousins and so on.  The rate of decline in correlation of outcomes as we move to more distant relatives will be a measure of social mobility rates in Denmark, unaffected  by measurement errors, in each generation from 1790 to 2022.

Denmark and Norway, measured by conventional techniques, are among the most socially mobile countries in the world. If these new techniques show similar strong persistence as observed in England, this would fundamentally recast the way social science conceives social mobility.  If the standard social interventions, common in Nordic countries, are shown to have limited or no effect, it also has important policy implications. 

Note that in recent work, Nobel Laureate in Economics, James Heckman and Rasmus Landersø conclude that while American policy analysts point to Denmark as a model welfare state with low levels of income inequality and high levels of income mobility across generations, family influence on important child outcomes in Denmark is about as strong as it is in the United States (Heckman and Landersø, 2021).  So the supposed distinctiveness of Nordic social arrangements in terms of social mobility is ripe for examination.

A second significant aim of the DNRF Chair will be to examine the effects of the emancipation of women on social mobility rates in Denmark.  Did the admission of women to high schools after 1875 in Denmark and 1882 in Norway, led to increased marital assortment and increased correlation of mother’s attributes to child outcomes?  Using the constructed genealogies we will measure this by looking at the relative importance of paternal versus maternal grandfathers in predicting grandchild outcomes.  Did the emancipation of women in terms of education and employment from 1880 onwards increase the importance of mothers in determining child educational outcomes?