The Commodity Effects of Decommodification: Community Land Trusts and Neighborhood Property Values

ABSTRACT

This article explores the impacts of community land trust (CLT) properties on the real estate prices of nearby homes through a case study of a relatively large CLT in Minneapolis, Minnesota. We use hedonic regression and a difference-in-difference estimation with spatial error correction to measure price effects. The number of developments citywide is insufficient to yield significant results. However, we find evidence that clustering CLTs stemmed the decline in sales prices during the foreclosure crisis. The introduction of the first nearby CLT had no measurable price impact, but each additional CLT was associated with a 5% higher sales price in North Minneapolis, and 3% higher in Central Minneapolis. In the postrecession period, we estimate that the introduction of CLTs in North Minneapolis was associated with a 10.9% increase in nearby sales prices. These results suggest that, contrary to common assumptions, price effects are strongest when affordable properties are spatially clustered.

KEYWORDS:

• community land trust (CLT)
• community development and revitalization
• housing
• neighborhoods
• foreclosure

Acknowledgments

The authors would like to thank Eric Seymour, Edward Goetz, and Sam Ratick for comments on an earlier draft of this article. All errors remain our own.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Supplementary Material

Supplemental data for this article can be accessed here.

Notes

1. Citywide home sale prices began to rebound by 2010. However, in North and Central Minneapolis, where the CLTs were primarily sited, prices rebounded a bit later. For this reason, we ran the analysis using 2011 rather than 2010 as the start of the recovery period. To test the impact of this decision, we also ran citywide results using 2010 as the cutoff, and results were similar.

2. Whereas the shared ownership model in the UK has many similar programmatic features to the CLT—particularly its making the property affordable by dividing the ownership of the property into pieces owned by different entities—it is also rather different in some ways, as there is not permanent affordability built into that model, nor is there any community engagement (certainly not a legacy of community control) as there is for CLTs. For more on shared ownership in the UK, see Munro (2007) who situates it in the context of UK housing policy more generally or Wallace (2012) who focuses on the ways in it works, or does not work, for the people in the housing.

3. Minneapolis has designated community areas.

4. Koschinsky (2009) and Deng (2011) call this approach the adjusted interrupted time series model with a difference-in-difference estimation, or the AITS-DID modeling approach. The AITS terminology comes from Galster et al. (2004). However, the DID approach has been popular in the economics literature for more than 40 years.

5. We considered several other neighborhood variables, including crime rate and school quality, but none added anything to the explanatory power of the model.

6. The standard in the literature is 500 or 1,000 feet, and given the scattered nature of CLCLTs portfolio, the smaller threshold was more appropriate. We considered both thresholds, and descriptive analysis suggests that they would yield similar results.

7. There was one multifamily project, which was not included in this analysis.

8. We considered several other density specifications, including the distance to the nearest three CLTs; the density of CLTs per square mile (measured using a raster density process that calculated a smooth surface using the number of CLTs within ¼ mile); and the density of CLTs per 1,000 housing units (calculated using the city’s parcel file and census data). Ultimately, we decided to stick with the first specification because the results were generally consistent across models and because of the intuitiveness of the coefficients.

9. Variable definitions inspired by Deng (2011).

10. Galster et al. (2004) consider a spatial lag model, but determine that the fixed effects and spatial trend terms effectively control for spatial autocorrelation.

11. Foreclosure data were acquired from Hennepin County and analyzed by the authors.

12. We conducted the density analysis using a 1,000-foot threshold in GIS. This is why the density variable uses this threshold, whereas the difference-in-difference analysis is based on a 500-foot distance to a new CLT.

13. See Pace and LeSage (2008) for a discussion of the spatial Hausman test as it relates to the SEM.

Additional information

Funding

This work was supported by the National Science Foundation [Grant No. 1359826].

Notes on contributors

Katharine Nelson

Katharine Nelson is a doctoral candidate at the Bloustein School of Planning and Public Policy at Rutgers University. Her research focuses on housing finance, neighborhood change, and racial inequality.

James DeFilippis

James DeFilippis is a Professor of Urban Planning at Rutgers University. His research focuses on the political economy of cities and communities, with a particular interest in the processes of social change, and questions of power and justice in cities.

Richard Kruger

Richard Kruger is a PhD student at the Graduate School of Geography at Clark University in Worcester, Massachusetts. His research examines the urban environmental politics of housing, infrastructure, and community development planning in informal communities in San Juan, Puerto Rico.

Olivia Williams

Olivia Williams is an independent activist-scholar based in Madison, Wisconsin. Her research and advocacy revolve around land, housing, collective ownership, and cooperative movements.

Joseph Pierce

Joseph Pierce is an urban geographer who focuses on sustainability, place-making, micropolitics, non-state governance, and housing. He is currently researching the effects of very large-scale development projects on neighborhood governance and identity.

Deborah Martin

Deborah Martin is an urban geographer with interests in place identity, local politics, legal geography, qualitative methodologies, and social movements (particularly neighborhood activism). 

Azadeh Hadizadeh Esfahani

Azadeh Hadizadeh Esfahani is a PhD Candidate at the Graduate School of Geography at Clark University in Worcester, Massachusetts, currently based in Teheran, Iran. Her research interests include place-making and identity building, neighborhood-based urban planning and development, and citizen empowerment and capacity-building.