Association for Public Policy Analysis & Management

The Retraining the Gulf Coast Workforce through Information Technology (GCIT) Pathways Consortium project was a four-year project funded by the Department of Labor’s (DOL) Round Two Trade Adjustment Community College and Career Training (TAACCT) grants program. The grant was awarded in September 2012 to Bossier Parish Community College (BPCC), which led a consortium of eight additional colleges across the states of Louisiana and Mississippi. The project’s objective was to capitalize on the region’s growing IT sector and its increased demand for skilled labor by training TAA eligible workers, veterans, and other individuals with basic skills needs for jobs. The consortium focused on three IT specialty areas: health information technology, cyber security, and industrial information IT. The project includes five inter-connected strategies to help build career pathways that allow students to earn industry -recognized credentials and access in demand job opportunities.

The Ray Marshall Center (RMC) conducted the impact evaluation of the GCIT TAACCCT grant. We conducted an outcome analysis where we tracked and studied the education and employment outcomes of GCIT participants, including number of college credits earned, student persistence,  program completion, credential achievement, employment placement, employment retention and wage gains. We also conducted a quasi-experimental impact analysis to estimate the impacts of the GCIT TAACCCT program on key education and employment outcomes. We used propensity core matching to construct a comparison group and address the key issue of the counterfactual: what would have happened absent the intervention?  

In this paper, we share the circumstances and evolution of our impact evaluation. We discuss the challenges involved in collecting the academic and employment data needed for the evaluation. We focus on: the difficulties in setting up access to the data, due to the different governing bodies and administrative structures in Louisiana and Mississippi; the complex logistics of data transfer, due to multiple agencies (7+)  involved in data collection; difficulties in ensuring accuracy and completeness of data, due to very different data systems in Louisiana and Mississippi; and, challenges in tracking participant outcomes, due to long lags in data availability and insufficient follow-up time. We note the evolution of our quasi-experimental design over the program period, in response to these realities of data availability and data collection. We also note how the challenges related to data collection limit the robustness of our impact findings. Finally, we report outcomes and impacts of the program. Lessons learned from our experience evaluating a large consortium grant spanning two states should help inform future approaches to measuring the impact of large-scale federal investments like the TAACCCT program.