Using linked administrative data, the study identifies critical points where students leave STEM pathways and highlights opportunities to strengthen education-to-workforce outcomes.

New research from Coleridge offers insights into how students move through science, technology, engineering, and mathematics (STEM) education pathways and transition into the workforce, highlighting opportunities for states to improve student persistence, degree completion, and employment outcomes.

The study, Strengthening STEM Pathways from Education to Workforce: Evidence from Arkansas, uses linked administrative data from Arkansas to examine when and why students leave postsecondary STEM pathways, the factors associated with completing STEM degrees, and how educational experiences relate to workforce outcomes.

Conducted in support of the National Secure Data Service (NSDS) Demonstration project to provide a replicable framework for other states’ use of alternative data this project was sponsored and managed in consultation with the National Center for Science and Engineering Statistics within the U.S. National Science Foundation.  The research demonstrates how longitudinal administrative data can help policymakers better understand the education-to-workforce pipeline.

Key Findings: Understanding STEM Persistence and Attrition

The research found that while many students begin their postsecondary education in STEM fields, persistence remains a significant challenge. More than half of students who start in STEM ultimately leave the pathway, either completing a non-STEM degree or leaving postsecondary education without a degree.

Among the study's key findings:

• Early intervention matters. Students are most likely to leave STEM pathways during their first several academic terms, highlighting the importance of early academic support.
• Academic preparation is critical. Higher high school GPAs are strongly associated with greater persistence in postsecondary STEM programs, underscoring the importance of K–12 preparation.
• Institutional pathways influence outcomes. Students beginning at four-year institutions are more likely to complete STEM degrees than those starting at two-year institutions, where pathways may involve more transfers and interruptions.
• STEM degrees are associated with stronger employment outcomes. STEM graduates earn higher wages on average than non-STEM graduates and are more likely to work in industries associated with STEM activity.
• Student experiences vary. Differences in persistence across demographic groups, institutional settings, and enrollment patterns suggest opportunities for more targeted support.

Together, these findings point to opportunities for states and institutions to strengthen early academic interventions, improve transfer pathways, align K–12 preparation with postsecondary expectations, and better connect education programs to workforce needs.

The Power of Linked Administrative Data

Central to the research was the use of Coleridge's Administrative Data Research Facility (ADRF), which provided a secure environment for linking and analyzing data from three Arkansas state agencies:

• Arkansas Division of Higher Education: Postsecondary enrollment, degree pathways, and completion records.
• Arkansas Division of Workforce Services: Unemployment insurance wage records measuring employment and earnings.
• Arkansas Department of Education: K–12 academic preparation, coursework, and student characteristics.

By connecting these data sources, researchers developed a longitudinal view of students' educational and employment experiences, allowing them to examine pathways and outcomes in greater detail than would be possible using isolated datasets.

The research also identified opportunities to improve data infrastructure. For example, access to occupation-level employment data would enable more precise analysis of whether STEM graduates ultimately enter STEM occupations, rather than solely relying on employer industry classifications.

Informing Stronger Education-to-Workforce Pathways

Available as a shared framework within the NSDS Demonstration, this research provides a useful path for state policymakers and planners exploring similar questions. Beyond its findings for Arkansas, the study demonstrates the broader value of linked administrative data for understanding how people move through education and into employment.

The same approach can be applied to other fields, industries, and states, helping policymakers identify barriers to completion, evaluate education and training investments, and develop more effective strategies to support learners and workers.