This project aims to measure the geographical extent of social, economic, and political activities, that is, the geographic outline of markets in the United States from the year 1849 to the year 1871 by using 191, 928 records of town-level postal revenues. The geographical dimension of economic activities had been understudied for this period, even within the domain of Economic Geography. Economists usually think of economic activities as bounded by artificial borders such as the state borders or the county borders, or vague geographical features such as “the Mississippi Valley” or “the Great Plains”. In this paper, the geographic extent of markets is measured as the spatial clustering of high levels of human activities at a highly refined geographical scale and high frequency, rather than a subset of artificial borders. In other words, the methodology proposed in this paper for the explanation of how the markets coalesced, that is, how they came together, and cohere, that is, how they stayed together over time.
This paper presents a statistically defensible technique for using postal data for generating the geographical outlines of markets. Postal data is particularly useful as an input to that statistical process because it is micro-analytic, carefully created by the government agency, uniform across time and space, at least within the boundary of the United States, made available to be used in computer systems, and unexplored as an analytical tool. A bi-annual, national-scale postmaster compensation dataset is digitized, cleaned, geocoded, and analyzed in this paper. Postal revenue could be inferred from this postmaster compensation dataset because the Federal Government’s formula directly links local postmaster compensation to postal revenue. The entire dataset covers each town, city, and populated place in the country that has a post office during the period from 1813 to 1881.
