Assessment of Gentrification near Commuter Rail Stations in New Jersey

Devajyoti Deka - Rutgers, The State University of New Jersey

http://usj.sagepub.com/content/early/2016/08/25/0042098016664830.abstract

A couple of years ago, at a session of the Transportation Research Board Annual Meeting, two distinguished researchers presented papers on property value appreciation from rail investments. Being somewhat familiar with that stream of literature and also more recent literature showing a growing attraction of younger generations to urban centers with rail transit, I wondered why the property value appreciation studies do not go a step further to examine what happens to the people living in those areas when property value increases. Returning from the conference, I reviewed close to one hundred studies on property value appreciation from rail transit and transit-oriented developments to see how many studies examined the effect of property value appreciation on population change in areas near stations. I found very few. 

I wrote this article with two broad objectives: (a) to explain why researchers should explore the possibility that disadvantaged populations are displaced from transportation investments because they can lead to higher residential rents, and (b) to examine how population characteristics near commuter rail stations have changed in New Jersey over time. Pertaining to the first objective, I argued that transit investments may have similar impacts on disadvantaged populations as highway investments in the Interstate Era, such as the destruction of minority neighborhoods and the displacement of their residents. Such impacts could be highly consequential if we do not monitor population changes near transit stations, especially since studies have predicted a very high demand for housing near stations in the near future.  

My study focused on the commuter rail system of New Jersey since the State of New Jersey adopted the “Transit Village” program in 1999 to promote transit-oriented developments near transit stations and currently 24 of the 28 these “Villages” are located close to commuter rail stations. Through this program, municipalities receive a preferential (though conditional) treatment from the State in certain regards. I found in the literature pertaining to the Transit Village program that the interests of developers and municipalities may run counter to some of the conditions, such as the requirement to provide affordable housing. 

To examine population change near rail stations, I used census data from Geolytics at the level of census tracts instead of directly obtained data from the Census Bureau since the former uses the same tract boundaries over different censuses, whereas directly obtained census data are affected by modified tract boundaries. I made two types of comparisons: (a) a comparison of the tracts where stations are located with other tracts within the municipalities, and (b) a comparison of tracts within half-mile buffers around stations with other tracts within the counties. I call the first a micro-level comparison and the second a macro-level comparison. In both cases, the areas close to the stations were compared with the areas further in terms of proportion of low-income populations, proportion of minority populations, home value, rent, poverty rate, unemployment rate, vacancy rate, et cetera. Comparisons were made for three periods (1990, 2000, and 2013) to examine how the area characteristics changed over time.

The micro-level analysis showed low-income and minority populations have lower proximity to stations than others, but the macro-level analysis showed they have higher proximity. In terms of change over time, no significant evidence was found about a reduction of low-income and minority populations at either level. One of the reasons could be the stability of rent over time. Although home value increased significantly over time in the areas near stations, rent in those areas has remained more stable. A potential reason for the relative stability of rent is an increase in the supply of rental units. Since the reviewed literature as well as two regression models in the paper showed that rent is more likely to affect displacement than home value, I recommended continued monitoring of rent in the areas near stations.