Do Benefits and Work Incentives Counseling Improve Employment Outcomes of SSI/SSDI Beneficiaries Receiving Vocational Rehabilitation Services Under the Workforce Investment Act?
by Zafar Nazarov, Cornell University
The main goal of this study is to estimate the effect of benefits and work incentives counseling services on employment and program participation outcomes of SSI/SSDI beneficiaries participating in vocational rehabilitation programs. Only a few studies in the past have examined the associations between benefits and work incentives counseling services and the outcomes of SSI/SSDI beneficiaries (Trembley et al., 2004; Trembley et al., 2006). The main findings of these studies suggest that the beneficiaries who receive benefits and work incentives counseling services may achieve significantly greater improvements in earnings. However, the methods of measuring the effect of benefits and work incentives counseling in these studies have serious statistical limitations. They completely ignore non-random selection of beneficiaries into the services. The selection of beneficiaries into the services can be explained both by observed and unobserved factors. In order to address for the non-random selection issue, it requires using more advanced econometric techniques such as propensity score matching or an Instrumental Variable (IV) approach. To make an accurate judgment about which methods provide the most consistent results it requires a good understanding of the exact methodology that VR practitioners use in assigning beneficiaries into benefits and work incentives counseling services. In this study, I am planning to seek further insight from workforce development and vocational rehabilitation practitioners in order to understand how they assign beneficiaries into services. The use of a qualitative technique should help us to determine the most appropriate econometric method to estimate the effects of benefits and work incentives counseling on various outcomes of SSI/SSDI beneficiaries. If non-random selection is partially explained by unobserved factors, then I am planning to use the percentage of beneficiaries in local districts who do not receive these services as the instrumental variable, because the receipt of these services can be explained by a district’s capacity to provide such services (a district may have enough certified benefits and work incentives counselors to provide such services for all beneficiaries; or due to a limited number of counselors only those who show strong intention toward work may receive benefits and work incentives counseling services). This identification strategy is similar to one proposed by Aakvik, Heckman and Vyltacil (2005).