The Imperiia Project: a spatial history of the Russian Empire

How we know what we (think we) know

The Central Statistical Committee compiled data on fires from two distinct but closely related sources: the biweekly reports submitted by provincial governors, and the annual summaries compiled by those same provincial governors. The biweekly reports were supposed to include the exact days of each incident, the location, the cause, the number of homes or other buildings burned, and the value of lost property (calculated in rubles) in urban areas and provinces. The data breaks down by month, opening the possibility of visualizing change over time and across the space of the empire in hopes of identifying patterns (as well as outliers) in the distribution, frequency, and intensity of events. 

The Committee found, however, that there were various "gaps" and inconsistencies in the reports. It turned out that the annual summaries were not, as one might have assumed, simply compilations of the information reported every other week over the course of the year. In the case of cause and value of lost property, the annual reports were more reliable (as that information was not always easy to establish within days of the fire), while the interval reports could be relied on for the sheer number of fires (in large part because that number could be corroborated through police reports). It was universally unclear whether the tally of burned buildings listed in any given report described complete losses only or included damaged property as well.