The Spatial Dimensions of Gentrification and the Consequences for Neighborhood Crime

Abstract

This study examines neighborhood economic improvement, what is occurring in nearby neighborhoods, and the consequences for neighborhood crime rates. Negative binomial regression models are estimated to explain the relationship between the increase in average home values (a component of gentrification) and crime in Los Angeles between 1990 and 2000. We find that the spatial context is important, as gentrifying neighborhoods located on the “frontier” of the gentrification process have significantly more aggravated assaults than gentrifying neighborhoods surrounded by neighborhoods also undergoing improvement. Furthermore, this effect is stronger in neighborhoods that began the decade with the highest average home values. Our findings indicate that the extent to which neighborhoods are more or less embedded in a larger process of economic improvement, and where the neighborhood is at in the economic development process, has differential effects on neighborhood crime.

Keywords:

• neighborhoods
• social disorganization
• dynamic models
• spatial effects
• gentrification

Acknowledgments

The authors would like to thank Jon Maskaly for his research assistance and four anonymous reviewers for their helpful comments on an earlier draft.

Notes

¹ While we acknowledge that our definition is limited to examining economic improvement in neighborhoods and does not account for population displacement, cultural shifts, or changes in the social class of residents that are often associated with gentrification, we refer to the process as gentrification in order to align it with prior research and facilitate the presentation of the study.

² Glass coined the term in a pejorative manner and the term gentrification continues to evoke mixed reactions (see Lloyd, 2010; Smith, 1996).

³ Since Glass’ (1964) definition, scholars have debated the definition of gentrification. The debate largely turns on whether Glass’ characteristics 1 and 2 are necessary (see Newman and Wyly (2006) for a thorough discussion of the latter concern), as well as questions regarding how long gentrifying occurs before the transition is complete. Other scholars have argued that gentrification requires a cultural shift related to the influx of artists, “bohemians,” and young professionals (Douglas, 2012; Lloyd, 2010). For a comprehensive discussion on the gentrification debate, see the edited volume by Brown-Saracino (2010).

⁴ Gentrification can also be spurred on by intentional governmental decisions to revitalize neighborhoods or by private investors purchasing inexpensive property to renovate and flip, or use as rentals. Regardless, our measure of gentrification accounts more generally for the improvement of housing values regardless of what initiated the process.

⁵ Examples of such land value improvement include instances in which new nearby development brings desirable retail opportunities, or new quality jobs that are much closer than the existing job opportunities. Other examples include new zoning decisions, the construction of desirable amenities such as parks, or the designation of protected open land.

⁶ Not all researchers agree that super-gentrification occurs. Hackworth and Smith (2001) discount the possibility claiming that in order for gentrification to occur there must be a period of disinvestment.

⁷ Smith (1996, p. 187) describes “urban frontiers” as those located on “a line dividing areas of disinvestment from areas of reinvestment in the urban landscape.”

⁸ We do not adjust 1990 values for inflation given that this would change the estimated intercept in our equation since all neighborhoods would have shifted equally.

⁹ To account for the binning of the data (income is coded into ranges of values), we utilize the Pareto-linear procedure, which Nielsen and Alderson (1997) adapted in their prln04.exe program provided by Francois Nielsen at the following website: http://www.unc.edu/~nielsen/data/data.htm.

¹⁰ We estimated ancillary models that included only the spatial lag variables (and not the focal tract variables) and found that the r-squares were very similar to Models 1, 3, and 5 without the spatial lag variables. This suggests that the spatial lag variables explain just as much of the variance as the focal tract variables.

¹¹ For this and the other comparisons in this section, we assessed statistical significance by estimating a model on the complete sample that included the main effects of all variables, a dummy variable for high home value tracts, and interactions between the high home value dummy and all variables in the model. The t-tests for these interaction variables serve as the statistical test for significant differences between the two sub-samples.

¹² Given that some have suggested that gentrification occurs when there is both increasing home values as well as high turnover in residents, we estimated ancillary models including an interaction between the change in home values and the change in homeowner stability. This interaction was never significant in any models. It has also been suggested that gentrification only occurs in neighborhoods with sharply increasing home values: although this suggests a nonlinear effect of home values, ancillary models testing quadratic effects found no such significant results in any models.

¹³ The same significant interaction was also detected in the pooled sample, but not in the low home value subsample.

Additional information

Notes on contributors

Lyndsay N. Boggess

Lyndsay N. Boggess is an assistant professor in the department of Criminology at the University of South Florida. Her research primarily focuses on the relationship between communities and crime, with a particular interest in neighborhood change, racial/ethnic composition, and the housing market. Her work has appeared in Criminology, the Journal of Quantitative Criminology, and Social Science Research.

John R. Hipp

John R. Hipp is a professor in the department of Criminology, Law and Society, and Sociology, at the University of California Irvine. His research interests focus on how neighborhoods change over time, how that change both affects and is affected by neighborhood crime, and the role networks and institutions play in that change. He approaches these questions using quantitative methods as well as social network analysis. He has published substantive work in such journals as American Sociological Review, Criminology, Social Forces, Social Problems, Mobilization, City & Community, Urban Studies and Journal of Urban Affairs. He has published methodological work in such journals as Sociological Methodology, Psychological Methods, and Structural Equation Modeling.