Motivation for Applying to the State IMPACT Collaborative

The Pennsylvania team applied to the State IMPACT Collaborative, a joint project of MDRC and Coleridge, to formalize and deepen emerging partnerships between agencies responsible for human services and workforce development. Leaders at the Department of Human Services and the Department of Labor & Industry had already begun breaking down silos, but much of this collaboration rested on informal relationships rather than a durable, cross‑agency structure. Participating in the State IMPACT Collaborative offered a way to institutionalize that cooperation through a formal, shared initiative. At the same time, there was a strong interest in using data more systematically for program improvement.

●      Office of Administration (which houses the Pennsylvania Longitudinal Data System)

●      Department of Human Services – Office of Income Maintenance

●      Department of Labor & Industry – Pennsylvania Workforce Development Board

●      Office, Bureau of Workforce Partnership and Operations, and Center for Workforce Information and Analysis

●      The Governor’s Policy Office

 Applied Data Analytics Project: Design and Findings

State IMPACT Collaborative participants complete an applied research project using Arkansas state data. For the Applied Data Analytics (ADA) project, the Pennsylvania team examined whether co-enrollment of Supplemental Nutrition Assistance Program (SNAP) recipients in Workforce Innovation and Opportunity Act (WIOA) Title I, Title II, and Title III programs was associated with positive employment and earnings outcomes.

The team used propensity score matching and matched-sample regression models. They compared SNAP-only participants with SNAP participants co-enrolled in various WIOA titles and examined employment and quarterly earnings one year after SNAP exit.

While the analysis was exploratory and the results were not necessarily causal, the team found statistically significant positive employment differences across all co-enrollment types. Earnings results were most positive for Title I and Title III, though the team expressed caution about whether the effect sizes could be fully attributed to program participation alone and could be related to the large sample sizes used in the analysis. They highlighted that the results might not be solely attributable to WIOA participants but might, rather, reflect unobservable differences between those who participated in the programs and those who did not.

Lessons Learned from Cross-State Collaboration

Through the State IMPACT Collaborative, the Pennsylvania team has experienced a blend of inspiration and validation as members engage with other participating states. Learning about Ohio’s partnership with The Ohio State University, for example, underscored how powerful it can be to tap into a deep bench of researchers and data analysts that state agencies often struggle to hire and retain over the long term. It also underscored that collaborating with public universities can be a smart, realistic strategy for building long-term analytical capacity, rather than trying to reinvent the state workforce. Pennsylvania currently relies on a small number of highly skilled staff within agency-based data and analytics teams, such as the Pennsylvania Longitudinal Data System, whose commitment to public service helps compensate for limited internal research capacity.

At the same time, conversations with peer states have been reassuring: challenges like slow progress in data sharing across agencies appear to be common, not a sign that Pennsylvania is uniquely behind. While the team remains committed to improving the pace and quality of data integration, it has been helpful to recognize that these hurdles are widespread and systemic.

ADA Training Benefits

The ADA training deepened the Pennsylvania team’s technical fluency with data, strengthened collaboration between policy and analytics staff, and equipped them to design, interpret, and apply rigorous analysis more confidently.

●      The ADA training substantially expanded the Pennsylvania team’s capacity to work with and understand how to use data, especially for staff from policy rather than research backgrounds. Hands-on experience with tools like SQL and R demystified the data analysis process, replacing a sense that data work happens in a “black box” with a clearer understanding of how analyses are conducted and how to interpret results.

●      The training helped non-data staff learn how to ask the right technical questions and provide the fine-grained information analysts need, improving collaboration between “data” and “policy” staff.

●      For data-focused team members, the project work and training offered a structured way to connect their technical expertise to real policy decisions, beginning to bridge two professional worlds that are often siloed in day-to-day government work.

●      Lessons from the ADA project are being actively applied to their State IMPACT project, such as limiting sample sizes to avoid finding statistical significance purely due to large scale.

●      Evidence from Arkansas reinforced that it is worthwhile examining employment and wage outcomes associated with WIOA co-enrollment and other program characteristics, such as co-location of services.

●      Beyond technical skills, the broader ADA experience—including intensive cohort convenings and opportunities to present to other states—deepened professional growth and made it easier to have “the right conversations with the right people.”

Future Goals for Analytical Work

Looking ahead, the Pennsylvania team wants to widely share the results of their State IMPACT project with state agencies, local workforce boards, and state workforce board members to spark interest in new analytic projects among colleagues who were not part of the initial effort. It also hopes to build on the ADA experience by staying connected to a broader cross-state learning community and continuing to exchange feedback and ideas with future Collaborative cohorts.