Analyzing Employment Possibilities from Recovery Housing: Jobs Available, Job Mix, and Clustering of Houses

Background and Purpose: While remission from Substance Use Disorder (SUD) is the primary goal of treatment, many researchers consider it to be insufficient and emphasize the importance of recovery, defined as a process of continual improvement across mental, physical, and social well-being following remission (Witkiewitz et al., 2020; Witkiewitz & Tucker, 2024). Employment plays a vital role in recovery, both as an outcome of the process and a predictor of successful recovery. Evidence suggests employment is of intrinsic value to recovery, regardless of the specific job or its wage (Petry et al, 2013; Sacker et al, 2019). It is therefore understood that recovery houses, long-term residential facilities for people in recovery from SUD, should be located in areas with plentiful available jobs. Previous studies have used neighborhood unemployment as a proxy for the jobs available to recovery house residents (Callahan et al, 2021). This is a weak proxy since most individuals work outside their neighborhoods.

Methods: We use Geographical Information Systems (GIS) and spatial analysis of data from recovery houses in three different Midwestern cities, Cleveland, Columbus, and Cincinnati, all in the state of Ohio. The analysis incorporates data on the location of recovery houses, unemployment in the neighborhoods of the recovery houses, accessibility of jobs from recovery houses (using the HUD Housing & Transportation Affordability Index), and employment mix (a weighted sum of 13 categories of employment that were available to residents in each house).

Results: The vast majority of recovery houses were single-gender and single-occupancy, with a median distance between houses in all cities of less than a kilometer. The mean unemployment rate in the neighborhoods in which recovery houses were located was 11.3%. The mean number of jobs available within a 30-minute commute from each house was 41,178, but this varied considerably from city to city, with the mean in Cincinnati being 50,909 while the mean in Columbus was 38,407. The employment mix was reasonably consistent across the three cities.

Conclusion and Implications: This analysis suggests that the use of neighborhood unemployment as a proxy for the employment prospects of recovery house residents is unduly pessimistic. In these cities, there were, on average, tens of thousands of jobs available to recovery house residents within a half-hour commute, and in all of them there was a mix of jobs. Recovery houses are within walking distance of other recovery houses, indicating that it peers can also pass information on about possible jobs (Kadushin, 2011). It is possible to use these sources of data when allocating locations of recovery houses to improve the employment prospects of residents.