Panel Paper: The Impacts of the Ticket to Work Program: Estimates Based On Randomly Assigned Mail Months

Saturday, November 9, 2013 : 8:40 AM
Salon III A (Ritz Carlton)

*Names in bold indicate Presenter

David Stapleton, Arif Mamun and Jeremy Page, Mathematica Policy Research
This paper presents results from a new analysis of the impacts of the original introduction of the Ticket to Work (TTW) program. TTW was rolled out in three phases: a first set of states completed the TTW rollout in 2002, a second set in 2003, and a final set in 2004. Under TTW, the Social Security Administration (SSA) mails each eligible disability program beneficiary a “Ticket”, in essence a performance-based voucher that he or she could offer to either a state vocational rehabilitation agency or to a prequalified local rehabilitation service provider, called an employment network, in exchange for employment placement, job training, and other services. The intent of the TTW program is to help beneficiaries earn enough to forgo cash benefits. Earlier analyses produced ambiguous results on the impacts of the program, reflecting the inherent limitations of the non-experimental methodology. The new analysis substantially resolves these ambiguities by exploiting random variation in the month in which Tickets were mailed to beneficiaries within three phases over the three-year rollout period.

In this analysis we focus on new Social Security Disability beneficiaries who are younger than 40 and not also receiving Supplemental Security Income. Earlier research identified this group of beneficiaries as the most likely to actually use their Ticket. Using SSA administrative data, we estimated impacts on a number of monthly outcomes: service enrollment, start of the Trial Work Period (TWP), TWP completion, benefit suspension or termination for work (STW), and months in nonpayment status after STW (NSTW months). We examined these outcomes over a 48-month period following the start of the rollout in each phase. 

Our analysis directly estimates the effect of duration from rollout start to Ticket mail month on each outcome, in a regression framework. To address the fact that some Tickets were not mailed on the random schedule—because of benefit termination prior to the mail date, primarily due to mortality, or because a few beneficiaries obtained their tickets early under a policy called “Ticket-on-demand”—we used a modified version of the actual mail month in the model and used the intended mail month as an instrumental variable.

Because SSA eventually mailed a Ticket to essentially all eligible beneficiaries in each phase’s sample, it is not possible to directly estimate the impacts of mailing Tickets versus never mailing Tickets. Nonetheless, under reasonable assumptions, we are able to infer the total impact of mailing the Ticket versus never mailing it, by estimating the impact of duration to mail month on outcomes measured at multiple points after the rollout start (12, 24, 36, and 48 months later). We are not able to disclose the findings at this time, but will be able to well in advance of the conference.

Full Paper: