Exclusionary Zoning: Policy Design Lessons From the Mount Laurel Decisions

ABSTRACT

Exclusionary zoning takes many forms, but always aims to limit economic integration within certain communities. Understanding the effectiveness of programs designed to reduce exclusionary zoning yields insight for future policy design, and the program that followed the Mount Laurel decisions in New Jersey remains relatively unexplored. The program created the Council on Affordable Housing (COAH), which used an incentive-based structure to implement affordable housing requirements. Municipalities that volunteered to meet their requirement received legal protection from zoning lawsuits. They could also engage in a regional contribution agreement (RCA), which allowed them to pay another municipality to complete up to 50% of their affordable housing obligation. Using probit and multinomial logit models, I investigate two questions concerning the program’s design: (a) Did COAH’s incentive-based structure succeed in attracting those municipalities with the greatest need for affordable housing? And (b) Did RCAs exhibit a pattern of high-income municipalities sending their affordable housing obligations to low-income municipalities? I find that the program succeeded in attracting high-income municipalities to participate, but that these municipalities were also likely to use RCAs to send housing units to low-income municipalities. I argue that the program’s design undermined the Mount Laurel decision’s original intent by limiting economic integration in high-income municipalities.

KEYWORDS:

• low-income housing
• municipalities
• policy
• programs

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1. For more background information, refer to: Mount Laurel Doctrine. Fair Share Housing Center. Retrieved from http://fairsharehousing.org/mount-laurel-doctrine/

2. So. Burl. Cty. NAACP v. Tp. of Mt. Laurel, 67 N.J. 151 (1975).

3. This general pattern was confirmed by a telephone conversation with a representative of the New Jersey Department of Community Affairs.

4. For more on the design on the program, refer to the New Jersey Council on Affordable Housing (1987) Annual Report.

5. NJSA 52:27D-312.

6. So. Burl. Cty. NAACP v. Mt. Laurel Tp., 92 N.J. 158 (1983).

7. Assembly Committee Substitute for Assembly, No. 500, State of New Jersey, 213th Legislature (2008).

8. I attempted to find more reports, as they are titled annual reports. However, I was unable to find evidence of any reports other than the six years listed.

9. For a summary of some of these details and a proposed alternative methodology moving forward, refer to Econsult (2015).

10. Since there were a number of municipalities for which new construction and rehabilitation obligations were 0, I log transformed these variables using ln(x + 1).

11. Given the small number of regions and the difficulty in correcting for heteroskedasticity in maximum likelihood models with simple robust standard errors, I report unclustered, nonrobust standard errors.

12. As noted by Keane (1992), identification in the multinomial probit model is difficult without an exclusion restriction. Moreover, Long and Freese (2014) discuss the conflicting results of Independence of Irrelevant Alternative (IIA) tests and highlight multinomial logit’s usefulness when the alternatives are distinct and not simply substitutes for each other, which I believe is an easy case to make here. As such, I believe the multinomial logit model is the best model to answer this question with the available data.

13. For more information on AMEs and average partial effects, refer to Wooldridge (2012).

14. The first two differences are confirmed by a t-test of means to be statistically significant at the 1% level, and the third is statistically significant at the 5% level.

15. T. Murakami (2000, May), Going to Court over Housing “Builder’s Remedy” Suits Force the Issue with Towns that Won’t Plan for Affordable Housing. Retrieved from philly.com

16. Recall that a municipality’s rehabilitation obligation and its total population are highly correlated, and that almost 74% of the variation in rehabilitation obligation is explained by total population, which is why population is omitted from the model.

17. Assembly Committee Substitute for Assembly, No. 500, State of New Jersey, 213th Legislature (2008).

Additional information

Notes on contributors

Thomas J. PlaHovinsak

Thomas J. PlaHovinsak II is an Assistant Professor of Economics at Longwood University and completed his PhD in Economics at Northeastern University. His research interests include housing policy, the spatial distribution of poverty, and urban economics. His previous work has examined a variety of urban economic issues and involved program and policy evaluations, as well as economic impact studies of public and nonprofit entities.