Contemporary neighborhood housing dynamics in a mid-sized US city: the policy consequences of mismeasuring the dependent variable

Abstract

This paper analyzes median assessed residential property values using three different operationalizations of a tract-level dependent variable for a mid-sized US city: Louisville, Kentucky. We estimated the impacts of accessibility, socio-demographics, and housing characteristics as well as three policy interventions (HOPE VI, historic preservation districts, and university–community partnerships) on median assessed values and changes to them over the 2000–2006 housing bubble. Our interpretation of models employing the three different operationalizations leads to different conclusions about neighborhood health and the efficacy of policy programs. Conventional operationalizations employed by advocates, such as looking at medians or raw dollar changes in median values, are likely to find that policy interventions have less of an impact compared to measuring recent percent changes in property values. Thus, we provide a methodological contribution that shows that percent changes should accompany traditional analysis in capturing the effects of contemporary policy interventions. Mismeasuring neighborhood housing markets has played a role in prematurely concluding that targeted policy programs in neighborhoods are ineffective. Based on our analysis, we invite academics and policy-makers to rethink contemporary neighborhood housing dynamics.

Keywords:

• housing policy
• neighborhoods
• HOPE VI
• historic preservation
• university–community partnerships

Notes

1. This map is recreated as Figure 2, referenced later, using 2000–2006 dollar increases. Future annual reports released by the GLP, up to 2013, do not include comparable maps or data on the local housing market. This presumably reflects their desire to focus more on other types of indicators, including education, health, and technology.

2. Gilderbloom and Appelbaum (1988, p. 95) pioneered the idea that urban regions analyzed in housing studies should be “relatively self-contained housing markets, free from the effects of adjacent market areas.” They studied only monocentric cities that were at least 20 miles from the nearest large central city over 50,000. Louisville’s larger distance to nearby major cities nearly quadruples this modest distance, making its housing market much more geographically autonomous than those in megalopolis regions.

3. Several of Louisville’s peer cities identified by GLP are also consolidated, including Indianapolis, IN, Nashville, TN, and Jacksonville, FL.

4. Other local advocacy organizations such as the Metropolitan Housing Coalition release reports based on analyses at even higher levels of aggregated geography, including Metro Council districts and ZIP codes, which do not approximate neighborhoods (e.g. Metropolitan Housing Coalition, 2008, which studied the impact of the foreclosure crisis in Louisville).

5. Three tracts (003000, 004900, and 011901) were removed from analysis due to an insufficient number of residential units, a common practice for Louisville housing studies (Ambrosius et al., 2010; Bourassa, Cantoni, & Hoesli, 2008; Gilderbloom et al., 2012; Gilderbloom, Hanka, & Ambrosius, 2009, 2012; Greater Louisville Project, 2007). Tracts 00300 and 004900 comprise the CDB, respectively, and have seen a very recent significant rise in residential development – although some of these homes are still coded as commercial by the Jefferson County PVA (as of 2006). Tract 011901 includes the airport and its surrounding commercial district, both of which were expanded very recently, thereby destroying a large portion of the housing stock in this particular census tract. This tract also includes the fairgrounds/exposition center and a theme park.

6. We also ran regression specifications with Distance to the CBD in place of the Location Inside the Inner Beltway variable (not shown) to capture proximity to the downtown in a traditional, linear fashion (Ambrosius et al., 2010). In this present analysis, we chose instead to compare the submarkets traditionally labeled “urban” with those recognized locally as “suburban,” including both inner and outer rings. The tracts labeled Inside the Inner Beltway do not perfectly align with the old city boundaries but follow Louisville Metro’s own designation of the new urban submarket (Louisville-Jefferson County Metro, 2006).

7. Our approach may be viewed by some as a limiting factor for our models.

8. We estimate that these positive changes have resulted in a housing stock that is approximately one-quarter gentrified.

Additional information

Notes on contributors

Matthew J. Hanka

Matthew J. Hanka is an Assistant Professor of Political Science and Director of the Master of Public Administration program at the University of Southern Indiana, Evansville, IN. His research interests include housing policy, community development, urban policy and governance, social capital, and historic preservation. His work has been published in American Review of Public Administration, Community Development, Housing Policy Debate, Journal of Urban Affairs, Journal of Urbanism, Local Environment, and Planning for Higher Education.

Joshua D. Ambrosius

Joshua D. Ambrosius is an Assistant Professor in the Department of Political Science and Master of Public Administration Program at the University of Dayton. His research interests include urban and housing policy, regional governance, and religious organizations. His academic work has appeared in such journals as Journal of Urban Affairs, Housing Policy Debate, American Review of Public Administration, Journal of Urbanism, Local Environment, and Interdisciplinary Journal of Research on Religion.

John I. Gilderbloom

John I. Gilderbloom is a Professor in the Graduate Planning, Public Administration, and Urban Affairs program at the University of Louisville, where he also directs the highly lauded Center for Sustainable Urban Neighborhoods (http://sun.louisville.edu). He is also a Fellow of the Scholars Strategy Network under the direction of Professor Theda Skocpol at Harvard. In an international poll of thousands of urbanists, planners, and architects, Professor Gilderbloom was ranked one of the “top 100 urban thinkers in the world.”

Keith E. Wresinski

Keith Wresinski received a BA in Geography and a GIS Certificate from the University of Missouri. He then worked toward a Masters in Planning at the University of Wyoming and is currently a Research Assistant and PhD student in Urban and Public Affairs at the University of Louisville. His primary skills are in the areas of geographic information systems (GIS) and econometric modeling.