Much historical work has focused on periods of abrupt climatic changes or politico-economic convulsions. But questions have been raised whether traditional approaches are based on a sufficiently rigorous methodology (de Vries, 1981). Often, studies select extraordinary events (such as revolutions or depopulations) or marginal regions and then look for explanatory events such as climate changes. Unless these crises can be shown to be typical responses to similar changes in climate, the estimated impacts are biased upwards because of statistical selection bias. Such an approach applied to banking would convince people that banking history was essentially the study of bank robberies (de Vries, 1981). To remedy this situation, statistical time-series approaches based on continuous meteorological, economic, and other social data should be used to provide more accurate information regarding the systematic relationship between weather/climate and economic activity (de Vries, 1981).