Urban Transition Project

Creating historically accurate GIS maps and linking these to data on persons and communities is a task with many kinds of challenges. This project is supervised by John Logan, and it is has included several doctoral students, postdoctoral research associates, and professionals at Brown University over the last few years. In addition, dozens of undergraduate RAs have been involved in transcribing records, cleaning street names, georeferencing map images, editing maps, and other tasks that cannot be automated. This effort has been made possible by support from the professional staff of Brown’s initiative in Spatial Structures in the Social Sciences (S4) and the Population Studies and Training Center (PSTC) and financial assistance from NICHD and NSF.

A major ongoing effort is to replicate the geocoded city maps that we initially developed for 39 cities in 1880. We selected the 69 largest cities (based on 1940 population) for this effort. Tract data for 61 of these cities with a limited set of variables were originally transcribed by Donald and Elizabeth Mullen Bogue (see ICPSR data file 2930). Census tract boundaries have been disseminated through the National Historical GIS Project (Minnesota Population Center). The availability of 100% microdata that can be readily aggregated to the enumeration district (ED) level for both 1930 and 1940 was the stimulus for creating GIS maps to allow spatial analysis of these data. Our goal is to go beyond that and geocode all households at the address level.

Major steps have included development of a historically accurate street grid for the cities in 1930 and 1940, determining census block and ED boundaries, reconciling the many ways that addresses and street names were recorded by enumerators and transcribed by Ancestry.com with the street maps, and finally geocoding the addresses of all households. A key resource that is widely used by genealogists and that greatly facilitated our work is the website stevemorse.org developed by Stephen P. Morse and Joel Weintraub. Early documentation of procedures used in this mapping were published in a book chapter: John R. Logan and Weiwei Zhang. 2017. “Developing GIS Maps for U.S. Cities in 1930 and 1940” In Don Lafreniere, Ian Gregory, and Don Debats (editors), The Routledge Handbook of Spatial History. Routledge: UK. Our methods have evolved since 2017, and we now rely heavily on methods to transcribe street indexes from historical city directories, and import information on street names, intersecting streets, and address ranges from that source into our GIS map.

At this time, the street grid with ED and ward boundaries is available to download for several cities for 1930 and 1940. Click here for the map download page. In addition, we have developed a web-based map system that allows users to view these maps interactively. Click here to enter the web-based GIS for 1930-1940.

It is possible to do spatially informed research even for cities for which no GIS map is available, taking advantage of information about nested geographies at multiple spatial scales. In our current research on residential segregation during 1900-1940, we have enhanced the 100% IPUMS microdata files for 191 cities to support analysis at the scale of households, dwellings, adjacent dwellings, streets, EDs, and wards. Details about the definitions of these variables and the steps needed to create them are described in a working paper:

John R. Logan, Benjamin Bellman, and Elisabeta Minca. “From Side Street to Ghetto: Understanding the Rising Levels and Changing Spatial Pattern of Segregation, 1900-1940.

The resulting data file is the Urban Transition Geographic Reference File (GRF). It includes the geographic variables for 191 cities in each of these decades (all cities that had a population of 30,000+ in any decade). It includes the ward, ED, block (in 1930), house number, and street name for every household, plus a dwelling, household, and person number. It also includes the MPCID, which is an identification number that can be used to link directly to the IPUMS 100% microdata. To enable checking of the linkage, it also includes for every person their race (white, black, other), nativity (U.S. or foreign born), and occupational SEI for employed persons.

Click here to download the GRF.

Click here to download the GRF codebook.