Unsettling the case study in gentrification research: Leveraging data science for new geographies of urban theory

Abstract

Case studies have long been the dominant methodological approach in critical urban theory. But reliance on case studies limits the ability to do systematically comparative research. One alternative is a turn toward data science in critical urban research. Such emerging methods can expand the data geographies of urban research, building complementarity between “thick” geographies of information available in case studies—which delve into local nuances but leave all other places unexamined—and “thin” geographies of information available through data science approaches—which offer less information on any one place, but allow a far more diverse range of places to be analyzed. This is illustrated with research on gentrification in United States cities. Text analysis of a large sample of gentrification literature finds that a heavy reliance on case studies leads to over-representation of large global cities and use of circumscribed units of analysis (especially individual neighborhoods or municipalities), geographies that may not capture spatial patterns of gentrification. An alternative approach is presented, based on mapping gentrification intensity across over 56,000 census tracts in over 800 metropolitan and micropolitan regions. Our National Gentrification Intensity Map demonstrates important but often under-researched patterns, including gentrification in smaller cities and suburban gentrification.