Yes or Not in My Backyard (YIMBY vs. NIMBY)? The Impact of New Social Housing Construction on Single-Family House Prices in Quebec City (Canada)

Abstract

The development of new social housing faces important resistance by local population, a phenomenon knows as the “not in my backyard” movement. One argument from residents to oppose such project is the idea that new construction will negatively impact property values. This is what this paper aims to investigate. The analysis is based on a complete recension of the new social housing projects built between 2000 and 2020 and on single-family house transactions that occurred between 2004 and 2020 in Quebec City (Canada). A repeated sales model integrating a difference-in-differences estimator is developed to isolate the net price premium related to the emergence of a new social housing building while accounting for the possible heterogeneity impact related to characteristics of the building, including the number of apartments and the type of clientele hosted as well as the local characteristics, such as the spatial concentration of social housing buildings and distance to the city center. The results show a complex net price premium rent function that leads to mixed conclusions and has important implications for the development of new social housing projects.

Keywords:

• Social housing
• public policy evaluation
• hedonic pricing model
• repeated sales approach
• difference-in-differences (DID) estimator

Acknowledgments

The Ville de Québec is not responsible for the accuracy, timeliness, or reliability of the content of this article. The views expressed in this article are the sole responsibility of the authors and do not necessarily coincide with those of the administration.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 However, this approach suffers from two specific problems: (a) it does not consider the fact that the spatial lagged dependent variable is endogenous (therefore OLS or GLS estimation methods are not appropriate); and (b) it does not decompose the marginal effect into the direct, indirect and total effects (LeSage & Pace, 2009).

2 Including the matrix D_it into the matrix Z_it allows us to directly test for the common trend assumption, similar to an event study design.

3 It should be noted that the correlation between the vector (T_r − T_s) and the matrix (D_ir − D_is) is usually very high, if not perfect. For this reason, the vector (T_r − T_s) is usually not included in the regression model.

4 This includes the introduction of eigenvectors (Griffith & Peres-Neto, 2006) or other spatial trend variables such as geographical coordinates and their quadratic terms (Galster et al., 1999, 2004) or other transformations such as Fourier expansion functions (Dubé et al., 2012).

5 This assumption is also made with the HPM application.

6 Unfortunately, the information on the projects’ announcement date is not available for about half of the new buildings. Thus, it makes it hard to decompose the price effect to account for the anticipation effect. For the buildings where the information is available, we know that it takes about 3 years, at the mean, from a project’s announcement to its opening.

7 The inverse Mills ratio is defined by ϕ(Z)/φ(Z), where ϕ is the marginal normal distribution function, φ is the cumulative normal density function, and Z is a matrix of independent variables included in the probit model.

8 To do so, it is possible to use the Gaussian transformation of the distance to the nearest new social housing building by applying the following the mathematical transformation: [1 – (d_ih/d_c)²]², where d_ih is the distance to the nearest social housing building h (in meters) and d_c is the maximal distance within the treatment area (600 m).

9 With a Gaussian transformation of the distance: [1 – (d_iCBD/d_c)²]², where d_iCBD is the distance to the CBD (in meters) and d_c is the maximal distance recorded.

Additional information

Funding

This study was funded by the Ville de Québec.

Notes on contributors

Jean Dubé

Jean Dubé is a full professor (since June 2020) at the École Supérieure d'aménagement du territoire et de développement régional (ÉSAD). His research interests include impact assessment of public policies related to planning and development, measurement of urban externalities through the real estate market, spatial analysis and spatio-temporal econometrics.

François Des Rosiers

François Des Rosiers is a full professor in urban and real estate management at the Faculté des Sciences de l’administration (FSA). His expertise covers urban and real estate economics, analysis of real estate markets and housing policies, valuation, property taxation and local finance, feasibility, market and profitability studies in real estate, economic impact studies and econometric modeling of real estate markets.

Nicolas Devaux

Nicolas Devaux is a professor at the Département sociétés, territoires et développement since 2018. His research interests concern regional and urban development through transport infrastructures, residential and property values analysis. He is also interested in issues of regional resilience and spatial mismatch in a non-metropolitan context.