Abstract
Using English data, we show that geographical variation in measures of social conservatism in the 19th and 21st centuries is strongly associated with variation in measures of social conservatism in the 17th century. Our statistical model includes a range of 19th-century and 21st-century economic and demographic characteristics, so the inter-temporal persistence in social conservatism is not to be explained by persistence in these characteristics. The association is still present when we fit a model with instrumental variables for the measures of 17th-century social conservatism, which suggests that the association is not a consequence of persistence in unobserved heterogeneity across locations. Rather, there does appear to be some inter-generational transmission of attitudes. A great deal of attention has been paid to the volatility of the British electorate in recent years, but our results show that the continuity in the country's political geography should not be overlooked.