GAO Studies State Data
The U.S. Government Accountability Office (GAO) recently released a report showing to what extent states match data between educational institutions and workforce sectors. The report’s findings are consistent with WDQC’s survey of states, and are relevant because longitudinal data systems may improve research and policymaking on workforce outcomes.
For the report, "Education and Workforce Data: Challenges in Matching Student and Worker Information Raise Concerns about Longitudinal Data Systems", the GAO analyzed data from 47 states and the District of Columbia which have received Statewide Longitudinal Data Systems (SLDS) and/or Workforce Data Quality Initiative (WDQI) grants, and which also participated in the 2013 Data Quality Campaign (DQC) survey. The GAO report relied primarily on the DQC survey results. The GAO also interviewed personnel from five of those grantee states (Ohio, Pennsylvania, South Dakota, Virginia and Washington) as well as officials from the U.S. Departments of Education and Labor. These departments provided over $640 million in SLDS and WDQI grants from 2006-2013.
The GAO found that most grantees that match data do share between sectors, but few share all possible types of data from all programs between sectors. Almost two-thirds of the grantees examined “had the ability to track individuals between all sectors from early education to workforce to at least some degree.” The GAO noted that in general, however, as “the percent of unique individual records reliably connected between databases increases, the number of grantees able to match data decreases.”
GAO also found that more grantees reported being able to match data among the education sectors than between the education and workforce sectors. Matching K-12 with workforce data appeared to remain especially difficult, with only nine of the 48 showing a 95% or higher match rate.
A low “match rate” doesn’t always mean that a state is doing a poor job of linking data. When matching education records with wage records, many individuals may not show up in the wage records because they’re working out of state or are unemployed, not because the matching method is faulty.
The report pointed to a number of reasons why states are facing challenges in developing and using their longitudinal data systems:
Restricted use of Social Security Numbers: Collecting SSNs in K-12 education data is prohibited in some states. One workaround has been the provision of another unique identifier for use by educational institutions that still allows for matching with workforce records while addressing privacy concerns.
Varied levels of data governance: States have been at different stages of developing data sharing agreements.
Flexibility of SLDS and WDQI grant requirements: These grants have provided some options in how states may use funds to develop their data systems.
State officials interviewed by GAO emphasized that federal funding has been vital to building longitudinal data systems and are worried about the sustainability of these systems after their grants end.
The GAO report also noted that more grantees are moving beyond building the data systems to actually using the data for research and policy development. Ultimately, the quality of these analyses and policies will in part depend on having more complete and comprehensive data systems, and successfully sharing data while addressing privacy concerns.