Another Look at Location Affordability: Understanding the Detailed Effects of Income and Urban Form on Housing and Transportation Expenditures

ABSTRACT

Findings from a study using the Panel Survey of Income Dynamics (PSID) and detailed urban environment and transit data support the location affordability hypothesis. Households in location-efficient places spent significantly less on household transportation, enough to offset high housing costs. Walkable blocks and good transit especially contribute to these savings. But households with very low incomes (below 35% AMI) do not see significant enough savings. Authors recommend investments in transit, sidewalks, and economic development in disinvested areas; the preservation and creation of affordable housing of all types and tenures; and more supports for households with very low incomes.

For decades, researchers have explored how location efficiency (LE) affects housing affordability, including incorporating transportation costs into a holistic housing affordability measure known as location affordability. Others have argued that estimated transportation savings from LE may be overstated because of limits in data and methods. Smart and Klein’s 2018 article in Housing Policy Debate analyzed the PSID and found “no evidence to support the location affordability hypothesis.” Considering their study’s policy implications, as well as its methodological limitations, we tested the PSID data at a smaller geography using more detailed household and urban form variables, per the LE literature. With this approach, we find statistically significant and meaningful transportation cost differences that are enough to offset higher housing prices for several income groups. However, the transportation savings for households in the lowest-income group in urban areas do not offset high housing costs. Because location-affordable places are in short supply, and the extreme shortage of affordable housing, both housing and transportation investments are needed to support households with low and moderate incomes. Expanding location affordability regionally will also help to address climate change and expand access to job opportunities, goods, services, and other amenities.

KEYWORDS:

• location affordability
• housing
• transportation
• expenditures
• urban form
• accessibility

Acknowledgement

Some of the data used in this analysis are derived from Restricted Data Files of the Panel Study of Income Dynamics, obtained under special contractual arrangements designed to protect the anonymity of respondents. These data are not available from the authors. Persons interested in obtaining PSID Restricted Data Files should contact [email protected]

Disclosure Statement

No potential conflict of interest was reported by the authors.

Supplementary Material

Supplemental data for this article can be accessed here.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. These four variables resulted in a high score on the silhouette measure of separation and cohesion (about 0.6 on a scale of −1 to +1), which indicates the cases within each cluster are close to each other and there is distance between the clusters. We specified three clusters to correspond to categories of urban form and to ensure a sufficient sample size (Norusis, 2011).

2. We also considered modeling transportation as a share of family income. However, transforming the dependent variable in this manner would introduce significant endogeneity into the model since some of the other controls are also a function of income. Taken together with the possible omitted variable bias introduced if car ownership is not included as a possible control, we rely on the actual dollar of transportation expenditures to mitigate possible correlation with the error term.

3. We also tested for multicollinearity issues across individual variables, given concerns around correlation between several of our predictors. However, we did not find evidence to suggest that we have issues of collinearity between our explanatory variables.

4. We also had the number of cars owned by each household. Approximately 89.35% of households own at least one vehicle. With a range from 0 to 10 vehicles, the mean is 1.71. Given our results, we use a dummy variable for vehicle ownership instead of the actual number for each household. The models yielded similar results to the one provided in this article.

5. We defined these as block densities at or above 0.3, square root-transformed TCIs greater than or equal to 5, and job densities greater than or equal to 100,000.

Additional information

Notes on contributors

Carrie Makarewicz

Carrie Makarewicz, PhD, is an associate professor in the Department of Urban and Regional Planning at the University of Colorado Denver. Her research interests include housing affordability, public schools, and transportation equity.

Prentiss Dantzler

Prentiss Dantzler, PhD, is an assistant professor in the Urban Studies Institute at the Georgia State University. His research interests include poverty studies, housing policy, and community development.

Arlie Adkins

Arlie Adkins, PhD, is an associate professor in the School of Landscape Architecture and Planning at The University of Arizona. His research interests include social equity, transportation, housing, and health.