Potential operational definitions of a “shrinking city” in the U.S.

Abstract

Ganning and Tighe (2021) argued that a shrinking city in the United States is a census-designated place that had a population of at least 50,000 in 1970 and experienced a loss in occupied housing units from 1970 to 2010. Using this definition, they identified a universe of 80 shrinking cities. Although it is not known how many urban scholars concur with this definition, the Ganning and Tighe (2021) publication has impacted the field, seeing that it has been cited more than forty times to date.

This author generally concurs with the Ganning and Tighe (2021) definition of shrinking cities. However, there are several critical questions that remain: Is a 40-year date range the correct timeframe for which to study decline? If so, we should now measure shrinkage from 1980 to 2020, given the updated decennial census data. If not—if we keep 1970 as a critical, fixed point—we should measure shrinkage over a 50-year period (1970-2020).

Relatedly and perhaps more importantly, when using either method of identification (loss of housing units over 40 years from 1980-2020, or 50 years from 1970-2020) does the universe of shrinking cities change? Should one account for boundary changes over time when identifying shrinking cities? By examining these questions, this paper will provide greater insight into how shrinking cities could be operationally defined. It will indicate whether city shrinkage is becoming more prevalent in the U.S. and whether there are cities that have stopped shrinking and may no longer be labeled as such.