Working Paper: NBER ID: w25427
Authors: Brady P. Horn; Aakrit Joshi; Johanna Catherine Maclean
Abstract: Substance use disorders (SUDs) are a major social concern in the United States and other developed countries. There is an extensive economic literature estimating the social costs associated with SUDs in terms of healthcare, labor market outcomes, crime, traffic accidents, and so forth. However, beyond anecdotal claims that SUD treatment centers (SUDTCs), settings in which patients receive care for their SUDs, have a negative effect on property values, there is scant empirical work on this question. In this paper, we investigate the effect of SUDTCs on residential property values using data from Seattle, Washington, and SUDTC location, entry, and exit information. To mitigate bias from the potential endogeneity of SUDTC location choices, we apply a spatial differences-in-differences (SDD) model, which uses property and SUDTC location to construct treatment and comparison groups. Our findings suggest that SUDTCs endogenously locate in lower value areas, and naïve models imply that the entry of an SUDTC leads to a 3.4% to 4.6% reduction in property values. When an SDD model is applied, we find no evidence that SUDTCs affect property values. Overall, our findings suggest anecdotal claims that SUDTCs reduce property values are potentially overstated.
Keywords: Substance Use Disorder; Property Values; Treatment Centers; Spatial Differences-in-Differences
JEL Codes: H0; I1; R3
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| SUDTCs endogenously locate in areas with lower property values (R38) | lower property values (R31) |
| naive models indicate that SUDTC entry (C24) | reduction in property values (H13) |
| SUDTC entry or exit (C24) | no statistically significant evidence that SUDTC entry or exit affects property values (R33) |
| SUDTC presence (R19) | negligible or non-existent impact (F69) |
| SUDTC exits (C24) | no significant lasting impact on property values (R33) |