Scattering as insurance: A robust explanation of open fields?

Abstract

McCloskey hypothesized that peasants scattered their land to diversify their portfolio of land holdings. This diversification produced insurance for peasants. Peasants paid an insurance premium in the form of lower yields on scattered holdings. McCloskey tested this scattering-as- insurance hypothesis with a small subset of Titow's study of grain yields on the Winchester estate. I use the entire data-set to test the robustness of the scattering-as-insurance hypothesis. I find that expanding the data-set lowers the robustness of the results (i.e. peasants are predicted to consolidate more often relative to the restricted data-set). The new estimates also highlight the importance of the assumed level of disaster income. All else equal, the higher is the disaster level income, the less robust is the scattering-as-insurance hypothesis. McCloskey assumed that disaster level income was half the output on a scattered field (d = 50). My empirical analysis suggests a higher level of disaster level income (d = 60). Lastly, I relax the assumption that peasants are motivated by safety-first concerns. McCloskey assumed that peasants would accept any reduction in mean incomes as long as it reduced the probability they faced of a disaster level income. I model peasants as trading off risk (variance of harvest) for consumption (mean harvests) in a more general manner. I find that even very risk-averse peasants would almost never scatter their fields to buy insurance: it was simply too expensive.

Citation

Bekar, C.T. (2001), "Scattering as insurance: A robust explanation of open fields?", Research in Economic History (Research in Economic History, Vol. 20), Emerald Group Publishing Limited, Leeds, pp. 173-221. https://doi.org/10.1016/S0363-3268(01)20006-8