Height and death in the Antebellum United States: A view through the lens of geographically weighted regression

Factors related to geography such as climate, natural resources or waterways often affect human activities. However, traditional approaches such as ordinary least squares (OLS) have limitations in investigating such patterns. Unlike OLS regression, geographically weighted regression (GWR) allows the coefficients of explanatory variables to differ by locality by giving relatively more weight to geographically close observations. GWR depicts spatial patterns. This paper examines the pattern of height and crude death rate in the United States prior to the Civil War by this method. The GWR results show that access to water transportation increased mortality and decreased stature in the food exporting areas of the Midwest, and the opposite pattern appeared in the food importing areas of the Northeast.

► Geographically weighted regression (GWR) allows the coefficients of explanatory variables to differ by locality. ► Water transportation increased mortality and decreased stature in the food exporting areas. ► The opposite pattern appeared in the food importing areas.