Changes in Access to Urban Mental Healthcare
Access to affordable substance abuse treatment has become an increasingly relevant policy issue in large American cities. Annual drug-induced deaths in the United States has more than doubled in the past decade, with the majority of these cases being concentrated in the country’s urban centers. As city leaders race to identify a viable strategy for addressing this concerning trend, a strong majority have aimed to mitigate the demand of illegal substances by providing access to Substance Abuse Treatment Centers (SATCs). Evidence supports this strategy, with recent national studies indicating that the presence of one additional substance abuse treatment facility reduces the annual rate of drug abuse and violent crime by 7% and 5% within the region, respectively. While many city leaders have adopted some variation of this strategy, others have detrimentally reduced access to their city’s mental healthcare resources through sweeping healthcare policy reforms.
In April 2011, the Chicago Department of Public Health (CDPH) operated 12 Community Mental Health Centers (CMHCs). By April 2012, that number had dwindled down to 6. This came as a result of a social service consolidation plan crafted by the city’s newly elected Mayor, Rahm Emmanuel, who aimed to cut the municipal budget by approximately $2.6 million. My analysis attempts to understand the causal relationship between Emmanuel’s plan and health-related outcomes in the city of Chicago following the closure of the 6 CMHCs. More broadly, however, this analysis explores the use of synthetic control methodology as a strategy for evaluating municipal-level policies.
Data Collection and Methods
The data collected for this analysis can be grouped into two categories: a treatment and donor series. The treatment series consists of what are considered “treated” outcomes, or data values that are predicted to experience a significant effect from the event or policy being analyzed. In the case of this analysis, the treatment series consists of annual drug mortality outcomes collected in the city of Chicago from 2000 to 2018, with the analyzed policy being Emmanuel’s consolidation plan. Conversely, the donor pool consists of multiple untreated data series that do not experience the analyzed policy and serve to predict how the treated series would have behaved had the policy never been enacted. In other words, they serve to create a hypothetical set of treated outcomes that, when compared to the actual treated outcomes, allow us to understand the causal effect of the policy decision. This analysis’ donor pool consists of 603 annual arrest- and mortality-related series from 49 states and 35 large American cities spanning from 2000 -2018.
Analysis of Treatment Series
Figure 1 depicts observed annual drug mortality rates in the city of Chicago between 2000 and 2018. The treated series depicts a clear increase in the rate of drug overdose mortality in the city of Chicago in the years following the 2012 clinic closures. This supports the hypothesis that the 2012 health clinic closures maintained a significant treatment effect on health-related outcomes for Chicagoans. However, this data simply serves to suggest that given the observed rise in drug overdose mortality in Chicago, it is plausible that an observable treatment can be identified through further analysis.
Figure 1: Treated Series Over Time
Results
The main results of the analysis appear in Figure 2, where the synthetic prediction is plotted against the treatment series. Based on these results, it is clear that an accurate synthetic prediction was not able to be made from the donor pool, indicating that an accurate match between the pre-treatment donor coefficients and the treated data series was unable to be made. A number of explanations can be offered concerning the failure to create an accurate prediction of drug overdose mortality outcomes using the SCUL procedure. First, the donor pool covariates, which included 603 annual city- and state-specific rates for arrests and mortality related to drug abuse, are simply ineffective as predictors of the actual pre-treatment outcomes in the city of Chicago. Second, the 2012 clinic closures may simply have caused too unique of a treatment effect for this method to effectively create a synthetic prediction. Thirdly, the treated data series may simply be too volatile to form a significant match with the covariates selected from the donor pool.
Figure 2: Treated Series Vs. Synthetic Prediction
Conclusions and Implications
Rahm Emanuel’s 2012 social service consolidation plan can easily be interpreted as the most consequential decision of his tenure as the city’s mayor. This research aims to identify the causal treatment effect of the 2012 clinic closures through the application of synthetic control methodology as a strategy for understanding the decision’s impact on mental health-related outcomes in Chicago. The application of synthetic control methodology in this analysis is an entirely nuanced concept that has the potential to fundamentally change the way in which municipal level policies are evaluated. Further, the results of this analysis serve as an indication that future analyses on Rahm Emmanuel’s 2012 Social Service Consolidation Plan is essential in understanding the policy’s long term effect on health-related outcomes in Chicago.
Riley Melton is a senior at Indiana University.
Leave a Reply