Jakob Madsen

The project extends the frameworks of Glaeser at al. (2004, 2007) and Acemoglu et al. (2014) in four dimensions to trace the most important forces that facilitated economic development such as institutions, education and culture. First, today’s per capita income is regressed on educational and institutional quality and, perhaps, culture in periods much earlier than the 1960 benchmark year considered in the literature for 150+ countries. Second, panel data will be used in supplementary regressions to deal with unobserved heterogeneity, i.e. the fact that the regressors are correlated with unobserved country characteristics. For example, 50 or 100-year differences of income growth are regressed on the change in initial values of institutions and education (the first year at which the long differences span) for a large panel of countries. Third, an instrumented variable (IV) approach will be used to deal with endogeneity since the explanatory variables may be endogenous to the extent that the regressors are serially correlated, that the regressors are capturing unobserved country heterogeneity and confounding factors, and that the initial conditions are products of earlier development. Fourth, interactions between culture, institutions and education will be allowed for; recognising that a minimum level of any of them is required for the other to have positive economic effects (Alesina and Giuliano, 2016). Educated people in countries with poor institutions, for example, are encouraged into rent-seeking activities that may render education counterproductive in the intensive margin.