David Salkever

We use discrete-time survival regression to study two empirical issues relating to take-up of individual placement and support (IPS) supported employment (SE) services for persons with serious mental illness: (1) the influence of client characteristics on take-up probability, and (2) the possible impacts of a major recent initiative in one state (Maryland) to overcome barriers to IPS-SE expansion. Our longitudinal analysis of population-based Medicaid cohorts, during 2002–2010, provides tentative evidence of positive state initiative impacts on SE take-up rates, and evidence of effects on take-up for clients’ diagnoses, prior work-history, health and demographic characteristics, and geographic accessibility to SE providers.

Persons with severe and persistent mental disorders (SPMD) have extremely low earnings levels and account for 29.1 percent of all U.S. Social Security Disability Income (SSDI) disabled worker beneficiaries under age 50. Social insurance and disability policy experts pointed to several factors that may contribute to this situation, including disincentives and obstacles in the SSDI program, as well as lack of access to evidence-based behavioral-health interventions. In response, the Social Security Administration (SSA) funded the Mental Health Treatment Study (MHTS) demonstration that included 2,238 beneficiaries of SSDI whose primary reason for disability is SPMD. The demonstration, implemented in 23 different localities, consisted of two evidence-based services (individual placement and support supported employment (IPS-SE), systematic medication management (SMM)), and provision or coverage of additional behavioral-health services (OBH). This study focused on estimating MHTS intervention effects on earnings in the intervention period (two-years). The main outcome variable was self-reported average monthly earnings. Subjects were randomly assigned to intervention or control groups. Data were drawn from the baseline survey, seven follow-up quarterly surveys, a final follow-up survey, and SSA administrative data. In all surveys, respondents were asked about earnings prior to the interview. Dependent variables were average past-30-days earnings reported in all follow-up surveys, similar averages for the first four follow-ups and for the last four follow-ups, fraction of surveys with prior earnings above SSA's substantial gainful activity (SGA) threshold, and final-follow-up earnings for the past 90 days. Regression analyses compared earnings of intervention vs. control group subjects. Covariates included baseline values of: (i) beneficiary demographic and social characteristics; (ii) beneficiary physical and mental health indicators; (iii) beneficiary recipiency history; (iv) beneficiary pre-recruitment and baseline earnings; and (v) local labor-market unemployment rates. Results show significant positive MHTS earnings impacts. Estimated annual increases of earnings range from USD791 (based on the 2-year average) to USD1,131 (based on the final quarter of Year 2). Effects on the fraction of quarters with earnings exceeding SGA are positive and significant but very small in magnitude. The consistent increase in earnings impacts over the study period suggests the possibility of even larger impacts with longer-term interventions. The moderate size of the intervention impacts may partly be explained by a study population that already had an average of 9 years on SSDI, and whose labor-supply decisions continued to be affected by concerns about possible loss of benefits. Limitations are that (i) earnings effects of specific intervention components cannot be estimated since all treatment subjects received the same package of services, and (ii) study results may not generalize to the majority of the beneficiary population due to selection effects in beneficiaries' participation decisions. Replication of the MHTS on a broader scale should show similar positive earnings impacts for a substantial number of beneficiaries with characteristics similar to the study population. Future studies should consider reducing policy barriers to labor supply of persons with SPMD. Future studies should consider longer-term interventions, or at least measuring impacts for follow-up periods greater than two years.

Current arrangements for financing and delivering behavioral health services to U.S. working-age adults with severe and persistent mental disorders (SPMD) have major inadequacies in funding for and access to critical elements of a recovery-oriented, comprehensive, and coordinated package of community-based treatment and rehabilitation services. This study presents results from a nation-wide 2-year randomized trial, involving 2,238 SSDI beneficiaries with SPMD, of a comprehensive intervention including evidence-based treatment and employment services. Estimates of impacts of the MHTS service intervention package, from a variety of regression specifications, showed clearly significant treatment group reductions in four outcomes (hospital stays and days, ER visits for mental health problems, and psychiatric crisis visits); these estimates suggest annual inpatient hospital treatment cost savings in excess of approximately $900 to 1,400. Negative estimated MHTS effects on three other utilization outcomes (hospital stays and days for mental health problems, overall ER visits) generally did not achieve statistical significance. Possible study implications for cost offsets from further expansions/replications of the MHTS intervention model are considered within the context of health reform.

Electronic health data sets, including electronic health records (EHR) and other administrative databases, are rich data sources that have the potential to help answer important questions about the effects of clinical interventions as well as policy changes. However, analyses using such data are almost always non-experimental, leading to concerns that those who receive a particular intervention are likely different from those who do not, in ways that may confound the effects of interest. This paper outlines the challenges in estimating causal effects using electronic health data, and offers some solutions, with particular attention paid to propensity score methods that help ensure comparisons between similar groups. The methods are illustrated with a case study describing the design of a study using Medicare and Medicaid administrative data to estimate the effect of the Medicare Part D prescription drug program among individuals with serious mental illness.

A recent policy analysis argued that expanding access to evidence-based supported employment can provide savings in major components of social costs. This article extends the scope of this policy analysis by placing the argument within a recently developed economic framework for social cost-effectiveness analysis that defines a program's social cost impact as its effect on net consumption of all goods and services. A total of 27 studies over the past two decades are reviewed to synthesize evidence of the social cost impacts of expanding access to the individual placement and support model of supported employment (IPS-SE). Most studies have focused primarily on agency costs of providing IPS-SE services, cost offsets when clients shift from "traditional" rehabilitation to IPS-SE, and impacts on clients' earnings. Because costs and cost offsets are similar in magnitude, incremental costs of expanding services to persons who would otherwise receive traditional services are probably small or even negative. The population served by an expansion could be sizable, but the feasibility of a policy targeting IPS-SE expansion in this way has yet to be demonstrated. IPS-SE has positive impacts on competitive job earnings, but these may not fully translate into social cost offsets. Additional empirical support is needed for the argument that large-scale expansion would yield substantial mental health treatment cost offsets. Other gaps in evidence of policy impacts include take-up rate estimates, cost impact estimates from longer-term studies (exceeding two years), and longer-term studies of whether IPS-SE prevents younger clients from becoming recipients of Supplemental Security Income or Social Security Disability Insurance

The potentially serious adverse impacts of behavior problems during adolescence on employment outcomes in adulthood provide a key economic rationale for early intervention programs. However, the extent to which lower educational attainment accounts for the total impact of adolescent behavior problems on later employment remains unclear. As an initial step in exploring this issue, we specify and estimate a recursive bivariate probit model that (1) relates middle school behavior problems to high school graduation and (2) models later employment in young adulthood as a function of these behavior problems and of high school graduation. Our model thus allows for both a direct effect of behavior problems on later employment as well as an indirect effect that operates via graduation from high school. Our empirical results, based on analysis of data from the National Educational Longitudinal Study, suggest that the direct effects of externalizing behavior problems on later employment are not significant but that these problems have important indirect effects operating through high school graduation.