Minimizing Travel Times to Help Meet the Mandated Obligations of Returning Citizens

Background and Purpose: Over 600,000 individuals return to their communities from prison (returning citizens (RC)) each year in the United States, but most of these RC’s will return to jail or prison in a cycle known as the “revolving door”. Transportation represents a major barrier to success and a critical thread that weaves together the many needs and obligations of RC’s. We partnered with a local reentry service provider that serves 1,000 clients to produce a tool to assist case workers. The study prioritized the mobility needs of RC’s critical for successful reentry by minimizing their travel times to mandatory activities and basic needs. The following research questions guided this study: 1. Given existing transportation networks, reentry obligations, and service provider locations, what are the optimal housing locations (existing or proposed) for returning citizens? 2. Given existing residential clusters of individuals returning from incarceration, transportation networks, and reentry obligations, what are the optimal service provider locations (existing or proposed)? 3. Given existing residential clusters of individuals returning from incarceration, service provider locations, and reentry obligations, what are the optimal residential assignments (for individuals)? Methods: The three optimization models developed in response to these research questions were based on the following sources: the Texas Department of Transportation (TxDOT, 2017); North Central Texas Council of Governments (NCTCOG) data; Google Open Street Map; the Dallas Area Rapid Transit (DART) General Transit Feed Specification (GTFS, 2020), and; U.S. Census block groups and data (2018). We created two travel time matrices for the U.S. Census block groups using the automobile and transit networks. The travel time matrices provided the shortest travel time between block groups. Further, we used our partner’s community network and Dallas County parole offices and other county facilities as the set of essential activities to consider for returning citizens. Results: Model 1 provides a prioritized rank order of housing alternatives. When a RC does not have housing, the model provides a prioritized list of housing alternatives based on current services. Model 2 identifies optimal housing locations to select for new housing projects or housing partners based on the primary community network partner locations, Dallas County parole offices, and other county facilities. Model 3 identifies optimal service locations to meet a maximum travel time constraint for all (as many as possible) Dallas County residents. It is optimal for agencies to serve as much of the population as possible, and not assume that RC’s must be concentrated in a few areas of the city (although this may be the current situation) to receive reentry support. Conclusions and Implications: We will discuss our efforts to install these models into the practices of our community partner agency and how this framework may be extended to other vulnerable populations who experience transportation disadvantage to optimize the locations of housing and services and assist social work practitioners.