The Geography of Demographic Composition in the United States

This award funds a project to study the history of displacement across American towns from 1850 to 1940. It does so by first using an innovative machine learning technology to create a comprehensive database of these displacements from historical newspaper records. It then merges this data set with individual census records to create a more detailed database of individuals so displaced and make the database available to other researchers. The researchers use this expanded dataset to study the impact of these displacements on communities and their role in shaping the populations of cities and locations in the US. The results of this research will provide educators, decision makers, and communities with detailed data to better understand the populations of US cities. The new data set contribute to making the US a global leader in research on geographical distribution of people across the US. This award funds a research project that will create a database of displacement events from 1850 to 1940 by leveraging historical newspapers, Census data, and large language models (LLMs). These tools identify and link episodes of displacements with individual census population data, enabling the analysis of their causes and impacts. The research quantifies the drivers of displacement, the intergenerational effects of displacement on affected populations, and its broader consequences for community relations and urban development. This large and highly innovative data set is made available to other researchers. By integrating machine learning with historical research, this project advances methods for systematic historical data creation and offers new perspectives on America’s community history. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.