Spatially concentrated renovation activity and housing appreciation in the city of Milwaukee, Wisconsin

ABSTRACT

This article examines the relationship between renovation activity and housing prices in Milwaukee, Wisconsin, and whether the spatial distribution of renovation activity matters for housing appreciation. We hypothesize that the renovation of housing units is a spatially diffusive process and that proximity to renovated properties increases sale price, even after accounting for renovations to the property sold and neighborhood characteristics. By adopting a modeling approach that incorporates hedonic and repeat sales methods, we find strong evidence that proximity to renovation activity exerts a positive influence on housing appreciation and that this effect extends further in space than previously believed. Our findings lend support to policy interventions that are geographically targeted and suggest that cultivating clusters of renovated housing can be a valuable lever for neighborhood stabilization and revitalization. Though appreciation was more likely in tracts with a higher poverty rate, an analysis of annual sales volume data suggests that displacement of owner-occupier households as a result of gentrification was not widespread during the study period. However, further research to better understand spatially concentrated renovation activity as a potential contributor to the displacement of existing residents is needed.

Acknowledgments

The authors thank the editor and the anonymous reviewers for their comments and suggestions. We also thank the City of Milwaukee for maintaining an extensive archive of property records and for making these data resources publicly available.

Notes

1. This term refers to a phenomenon that spreads or “diffuses” through geographic space (Cliff & Ord, 1981).

2. An example would be transfer of a residential property to a family member for price that is well below its market value. These types of transactions could bias a statistical analysis that relies on sales data.

3. Robust standard errors are calculated to address slight heteroscedasticity.

4. African Americans are the largest racial minority in Milwaukee and remain the most segregated group in U.S. metropolitan areas (Logan & Stults, 2011).

5. This is one of the suggestions offered by Helms (2003, p. 496).

6. The assumption that the unobserved characteristics of a housing unit that may impact sales price are constant from one transaction to another and therefore can be treated as fixed effects.

7. We adopt 2,000 ft. as an approximation of the three to five city blocks referenced in the literature examining the effect of new construction on appreciation (Simons et al., 1998), recognizing that the size of a block varies from city to city.

8. Recent studies (Galster et al., 2006; Helms, 2012) suggest that the effect of housing renovation on sales prices could potentially extend beyond the short distances documented by Ding et al. (2000). We estimate a regression model with distance radii of 150 ft. (Ding et al., 2000) and 1,320 ft., which is often used as a benchmark for walking distance in urban research but acknowledges that the effect of renovation is not expected to extend as far as new construction.

9. When the hypothesis of no spatial dependence is rejected, Lagrange multiplier tests are applied to select between a spatial lag and spatial error model specification (Anselin, 1988; Anselin, Bera, Florax, & Yoon, 1996).

10. A test of the residuals from initial ordinary least squares models revealed a very slight but statistically significant degree of spatial autocorrelation. To address this potential source of bias in the estimates, we fit a spatial lag regression model where the weighted average of appreciation in the four nearest neighboring observations is introduced as an independent variable denoted ρ in Equation 2.

11. Observations with a Cook’s distance greater than 4, the residual degrees of freedom were flagged as outliers, inspected, and removed (Fox, 1991).

12. Neighborhood upgrading refers to “physical improvement with the existing population remaining in place” (Varady, 1986b, p. 290) and is distinguished from gentrification as a form of revitalization by the absence of displacement.

13. The City of Milwaukee’s Targeted Investment Neighborhood program is one example.

Additional information

Notes on contributors

Bev Wilson

Bev Wilson is an Associate Professor in the Department of Urban and Regional Planning at the University of Illinois at Urbana-Champaign. He is interested in better understanding the spatial and temporal aspects of development, as well as its implications for the environment and society.

Shakil Bin Kashem

Shakil Bin Kashem is a Teaching Assistant Professor in the Department of Geography and Geographic Information Science at the University of Illinois at Urbana-Champaign. His research focuses on the use of geospatial technologies to better understand social vulnerability, climate change adaptation, urbanization, and disaster risk management.