Series # :
  ETAOP 2013-10
Title :
  Estimated Impacts for Participants in the Trade Adjustment Assistance (TAA) Program under the 2002 Amendments, Impact Estimates Regarding Non-Participants in the Trade Adjustment Assistance (TAA) Program Under the 2002 Amendments, and Methodological Notes
Release Date :
  January 18, 2013
Abstract :

These two reports from the Evaluation of the Trade Adjustment Assistance (TAA) program provide impact estimates, for both participants and nonparticipants, of the TAA program under the 2002 amendments.  Detailed information on the research design, statistical methods, data sources, sampling, and survey response rates can be found in the Methodological Notes Regarding the Impact Analysis. 

The impact study involved a quasi-experimental comparison group design, since workers eligible for TAA could not be denied benefits and services.  The study used two treatment groups:  participants and nonparticipants (i.e., workers who were eligible for TAA but did not receive significant services and benefits from the program).  Each treatment group had its own matched comparison group of UI claimants who lost their jobs in manufacturing firms around the same time (mostly in 2005 and 2006) in 26 states.  Data used in the impact analysis was primarily from the initial and follow-up surveys of both treatment and comparison groups.  “Nearest neighbor” propensity score matching was used to initially match comparison group members to members in each treatment group.  “Kernel matching” was then used to “rematch”, i.e., statistically adjust, the treatment and comparison groups, based on information collected in the initial survey.

Estimated impacts are provided in the reports on:  1) the receipt of re-employment and training services, 2) employment and earnings, 3) characteristics of jobs (including hourly wages and health benefits), and 4) health and family status.  The report on estimated impacts for participants includes many additional sensitivity analyses, including those for employment and earnings for different subgroups, such as by age and service receipt (including training), as well as using different specifications of the comparison group, including one of a matched group of just UI exhaustees. 

Publication Author(s)
Social Policy Research Associates
/ Mathematica Policy Research
  • Ronald D’Amico
  • Peter Z Schochet
  • Nathan Wozny
  • Sarah Dolfin
Full Text Document(s)
Additional Information

Hard copy available: No