Placing Segregation Robert Shepard robert-shepard@uiowa.edu University of Iowa Libraries, United States of America Residential segregation and socioeconomic inequality among neighborhoods is common in many American cities. Many large American cities became modern industrial urban places during the mid- to late-nineteenth century, so it is critical to understand the reality of human dimensions in cities during this period. Popular narratives tracing the history of the phenomenon back to that time have been constructed from analyses that used city wards or other large unit areas that aggregate data for many hundreds or even thousands of families. For example, prominent works like American Apartheid have concluded that black Americans were not particularly segregated from whites before 1900, and even then they were only slightly more separated from the native white population as the Irish and German immigrants (Massey and Denton 1993). By relying on aggregate datasets from the past, urban historians have been missing out on a level of granularity that would provide better measurements of segregation as well as critical context of spatial separation, given the relative smallness of American cities during their earlier years. Mid-nineteenth century U.S. Census records contained an abundance of potential information relating to socioeconomic realities of each home during this key formative period in American history. The 1860 and 1870 censuses were especially unique in that they collected each individual's personal estate value and real estate value (these economic variables were omitted from subsequent censuses). However, American census takers did not record address numbers of those they visited before 1880, which was after the Bureau of the Census stopped collecting these critical economic data about individuals and households. Furthermore, even if address data were available on manuscript census records for the time period, much of the built environment has changed since the mid nineteenth century, including the complete renaming of roads and - most importantly - the renumbering of houses along the street grid. Rather than using easily-available aggregate data collected at city ward levels to make inferences about past urban geographies, this work has combined city directories and period advertisements with census records to reconstruct digitally the historical address systems of cities and geolocate every possible family in the 1860 census for the cities of Washington, D.C., Nashville, Tennessee, and, for the 1870 census, the city of Omaha, Nebraska (additional cities will be added as the project expands). Notably, Donald DeBats and Mark Lethbridge had pioneered a similarly intensive approach for the much smaller cities of Newport, Kentucky and Alexandria, Virginia (DeBats and Lethbridge 2005). Because of the extensive details collected by these censuses, these geolocated individuals provide rich new datasets for historical researchers to explore. By geolocating individuals, this research also highlights specific individual living situations and helps to complement known (and more easily accessible) anecdotal accounts from common primary sources such as journals and letters. Placing Segregation is the primary result of this research effort - a new, open access digital project at the University of Iowa Libraries' Digital Scholarship and Publishing Studio that explores research questions about housing segregation and socioeconomic disparities across nineteenth century American cities through a series of fully interactive maps and scholarly interpretations derived from the geolocated census data. This presentation introduces core functionality of the digital exhibit and also explains in detail the process of developing the data and the website. Built primarily upon Leaflet.js, an open-source JavaScript library for mobile friendly interactive maps, as well as other openly available JavaScript libraries such as Fuse.js, this digital map exhibit gives public audiences direct access to detailed census data for the years 1860 and 1870 in these select U.S. cities. A search feature allows visitors to quickly locate the historical residences of persons of interest and read complete census information about them, while a number of simple layer filters permit users to explore potentially infinite topics by giving them direct control over the geographic data. The primary digital exhibit gives audiences without GIS experience the power to ask research questions in a spatial environment. Initial research findings from the project indicate that urban historians have been substantially underestimating the degree of housing segregation experienced by free black residents, even though some past research built on ward-level analyses closely approximated the extent of housing segregation between the native white population and various immigrant groups. The data generated for this project illustrate American cities were in fact significantly divided according to social class and wealth, long before the rise of the automobile. However, residents were less divided by wealth differences than they were separated spatially by skin color in Washington D.C. or by German and Irish origin in Nashville and Omaha.