Following Fido: Considering Other Measures of Gentrification

The most effective measure of gentrification will not be a magic bullet, but rather a constellation of variables aimed at addressing its processual and multi-dimensional character.
Mike Ed Fowler
By Lvova Anastasiya (Львова Анастасия, Lvova) (Own work) [CC BY 3.0 (], via Wikimedia Commons

Bearded, plaid-draped hipsters slink into newly established artisanal coffee roasteries, pop-up art galleries, and barbershop-bars in a neighborhood once dotted with vacant storefronts. To the casual observer, this overwrought caricature smacks of gentrification. But, gentrification is a bit like obscenity: blatantly obvious once observed, far more difficult to systematically define and measure.

The definition of gentrification—like any good concept in sociology—remains contested. To varying extents, many definitions suggest that gentrification is a process whereby an influx of new residents, accompanied by the service providers they attract and demand, replace and displace longstanding residents, businesses, and social institutions. The result of this process is the changing demographic profile of neighborhood residents along dimensions of wealth, race, and educational status, increasing property values (with corresponding increases in rents), and changes to the neighborhood business environment. Defined thusly, gentrification appears to be a complex process that requires not only a consideration of the shifting demographic profile within a neighborhood but also changes in local economic and built environments.

Such complexity presents a number of measurement challenges. Constructing an accurate approximation of the gentrification process is essential for quantitative assessments wishing to systematically address potential outcomes of gentrification, such as changes in crime rates, displacement, economic vitality, etc.. How researchers choose to operationalize gentrification, and construct subsequent statistical models, is of paramount importance as variation in measurement necessarily influences what is found.

Findings are particularly important in this case, beyond our lofty scientific ideals, because they are incredibly politically consequential. These findings influence the highly-visible gentrification debate as they are taken up in competing claims of who is harmed or helped by this specific process of urban change. This debate is not benign. The perception of policy-makers and the public they represent influences urban development strategies that will shape the future of American cities. As such, while researchers may not be able to control the interpretation of their results (or what they find altogether), they must strive to get measurement “more right” in the first place.

Historically, researchers have turned to Census data to measure gentrification. Most notably, the work of Elvin K. Wyly and Daniel J. Hammel has meticulously examined a litany of Census variables that are thought to indicate the process of gentrification. Many of the variables identified in this work have been taken up in subsequent quantitative assessment, specifically those regarding characteristics of residents, and, to a lesser extent, variables regarding the state of neighborhood housing, such as residential turnover, vacancies, and new-build construction. While insightful and able to capture some crucial dimensions of gentrification, Census data remains limited by decennial availability and its individualistic focus.

In response to these deficiencies, recent work has sought to augment census-based measurement with non-census-based data inspired by ethnographic accounts of gentrification. Specifically, recent work by Papachristos and colleagues utilized yearly historical business directories to operationalize the growth of coffee shops in Chicago neighborhoods as a proxy for gentrification. The authors note that coffee shops represent a salient lifestyle amenity associated with gentrifiers’ tastes and through statistical analysis show that changes in coffee shops are able to “hang with” other common Census variables, but also explain an unaccounted for corporate dimension of gentrification while overcoming the problems posed by decennial availability.

My own work on gentrification and Stop-and-Frisk in New York City has made use of coffee shop density, but why stop at coffee shops? Ethnographers like Sharon Zukin and lay-sociologists have identified numerous new consumption spaces associated with gentrifiers and gentrification. In fact, my opening vignette cited three, albeit specific, spaces that may signal to observers on the ground that gentrification is underway. Perhaps these and other businesses may offer greater explanatory power to quantitative models of gentrification, but new work should also look beyond the business world to consider lifestyle choices and habits of gentrifiers agreeable to quantification.

I offer one of such avenues open for exploration: the proliferation and spread of dog ownership in dense urban neighborhoods. Dog ownership in an environment where space is often scarce is contentious and inextricably linked to issues of class, status, and power in urban America. In cities like New York and Washington D.C., the rate of pet ownership is nearly half that of the national average, while the distribution of dogs in the densest four Boroughs of New York trends closely with resident wealth. Further, the characteristics of urban dog owners correlate with some of the primary characteristics researchers tend to associate with gentrifiers, while the formal and informal public accommodations this group carves out for their pets may be spatially assessed.

This approach may shed light on the individualistic, public, and corporate forces driving gentrification processes. Not only are the characteristics of urban dog owners correlated with some of the primary characteristics associated with gentrifiers, making them an intriguing proxy, but the accommodations municipalities make for dogs, such as sanctioned dog-runs and parks, and the businesses that crop up to service these animals—pet groomers, pet stores, and even pet clothing boutiques and massage parlors—may be quantified. Such spatial variables may be used in concert with Census data to construct a multidimensional measure, hypothesizing that growth across these areas over time signals gentrification in action. While not without limitations, such an approach incorporates non-corporate lifestyle amenities and habits associated with gentrifiers, and the role of public allocation in ways that business data alone cannot address.

While the number of residents from year-to-year calling themselves dog parents, and the accommodations they require, is and should not be the stand-alone measure of gentrification, researchers would be wise to pursue innovative variables like those suggested here in tandem with traditional census-based variables. Papachristos and colleagues have demonstrated the viability of selected business data in augmenting the solid backbone of traditional Census variables, but new approaches ought to look at quantifying other businesses, non-corporate lifestyle habits and choices, and public accommodations to bring greater explanatory power and intrigue to analyses.

The most effective measure will not be a magic bullet, but rather a constellation of variables aimed at addressing the processual and multi-dimensional character of gentrification. To accomplish this, quantitative research must be willing to use findings in ethnographic accounts in ways that are unique, sensitive, and may even account for our four-legged friends.

Mike Ed Fowler is a PhD candidate in Sociology at the University of Virginia. His work focuses on things that happen in cities and in classrooms.