Effects of Proximity to Multifamily Housing on Property Values in Little Rock, Arkansas, 2000–2016

ABSTRACT

A common form of not-in-my-backyard (NIMBY) activism is resistance to multifamily housing. Although NIMBY activism often targets both market-rate and subsidized multifamily development, studies of the effects of multifamily housing primarily focus on subsidized rental apartments. This study addresses this gap by analyzing the effects of condominiums and market-rate apartments as well as subsidized rental housing. Taking Little Rock, Arkansas, as a case study, this research uses a difference-in-differences approach to measure the effects of five types of multifamily housing on nearby single-family home sales prices: condominiums, market-rate rental apartments, subsidized rental apartments, senior and special needs apartments, and other multifamily housing (such as dormitories). The results suggest that most forms of multifamily housing have either no effect or a positive effect on sales prices for single-family homes within 2,000 feet of a new multifamily housing development.

KEYWORDS:

• NIMBY
• multifamily housing
• apartments
• property values
• affordable housing

Acknowledgments

I would like to thank Hunter Bacot for helpful comments on this manuscript, and Reteisha Byrd and Bruno Showers for research assistance on this project.

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1. Arms-length home sales were identified in the database’s deed transfer file as general warranty deeds with a sales price of more than $5,000.

2. Home prices are typically positively skewed. An examination of histograms and quantile–quantile plots of home prices and logged home prices indicated that logged home prices more closely match a normal distribution than the untransformed prices do.

3. Although senior and special needs housing is frequently subsidized in some fashion, a conventional belief is that such housing creates fewer negative spillovers than does subsidized housing that is available to anyone meeting a means test. Hence, this analysis distinguishes between these two forms of subsidized housing.

4. Includes college dormitories and university apartments (seven projects) and temporary shelter for the homeless (two projects).

5. These distance choices are motivated by the results of Ellen et al. (2007), Lee et al. (1999), and Schwartz et al. (2006), which indicate that the most significant effects of subsidized housing on nearby property values tend to be within about one quarter of a mile (1,320 feet) of a site, and diminish as the distance approaches 2,000 feet.

6. Values of predevelopment trend variables are expressed as a negative, and values of postdevelopment trend variables are positive, making the date of development time 0 relative to nearby sales.

7. These data were constructed using the Pulaski Area GIS Consortium’s building outline shapefile, which shows the land use for every structure in Little Rock.

8. Following similar difference-in-difference studies (e.g., Koschinsky, 2009), the interaction term notation, O_ijt * T_ijt and R_ijt * T_ijt, is used to underscore the fact that the measure of the pre- and posttrend variables is the time span between the sale date of each home and construction of a nearby multifamily development, and thus is only observed for transactions in the footprint of an multifamily development. This is equivalent to interacting each trend variable with its corresponding pre- or postdevelopment multifamily dummy variable (which takes a value of 0 for home sales outside the footprint of a multifamily development). But unlike conventional interaction models, T_ijt is not included as an independent term in the regression since it would be collinear with the interaction term.

9. Since the dependent variable is the log of sales price, the percentage effect of a dummy variable on sales price is calculated as exp(coefficient) − 1 (Halverson & Palmquist, 1980).

10. Since the regression in Equation (1)(1) $ln{P}_{it}=\alpha +{\beta }_1{O}_{it}+{\beta }_2{R}_{it}+{\beta }_3{C}_{it}+{\epsilon }_{it}$(1) also includes pre- and postdevelopment trend variables that measure the time between sale and multifamily construction, the pre- and postdevelopment dummy variables estimate what the price would be if a sale occurred at the time the multifamily building permit was issued and construction began (i.e., when T_ijt = 0).

Additional information

Funding

This research was supported through a contract with the City of Little Rock, titled “2016 Little Rock Multi-Family Housing Study.”

Notes on contributors

Michael Craw

Michael Craw is a faculty member and director of the Master of Public Administration Program at The Evergreen State College. He teaches courses in public finance, policy analysis, and urban management. His research on local public finance, social policy, and community development has appeared in journals such as the American Journal of Political Science, Publius, and Public Administration Review.