Funding Better Data in the President's Budget

The final budget request of the Obama administration was released this week though it’s not clear that anyone in Congress is listening. It’s many initiatives, new and old, will be non-starters for a Republican-controlled congress in the throes of an election cycle, but there are some proposals that show promise as starting points for broader conversations among presidential candidates, the education community, and current and future elected officials. While better higher education data is one small component of the President’s budget proposal, it would enable more informed policy decisions about higher education going forward.

Perhaps the most interesting new proposal would include $0.5 million to fund K-12 and postsecondary “information hubs,” as part of a new InformED initiative. These hubs would integrate various sources of federal data on education and make these resources more easily accessible and understandable, circumventing many user challenges of the current data infrastructure. Currently, data can be difficult to locate, understand, and analyze, presenting significant barriers to better understanding the current state of education. In many cases, these portals would simply make answering questions easier: State officials could use such a system to better understand how their state compares to others in federal spending per student. Institutions could compare the share of students taking out Parent PLUS loans in relation to other schools. And many of the most important questions can’t currently be answered - such as how earnings vary between students who withdraw and those who complete - but might be addressed through these new systems.

The release of the College Scorecard data last August was a bold move in the right direction. It combined information from multiple, existing, sources to help paint a more detailed, if still imperfect, picture about how students are faring at our nation’s colleges. Troves of never-before-released data were combined with publically available resources. To top it all off, the newly combined data saw dramatic improvements in accessibility. The College Scorecard’s use of an automated programming interface (API) and direct downloads featured on the site made the data immediately accessible to researchers, students, high school counselors, and college administrators with widely varying levels of technical skills. The new data was even integrated into data.gov, allowing users to view and explore the data without downloading any files. These are critical components that improve the data usability greatly, particularly in comparison to other NCES data sources, such as those provided through IPEDS and FSA.

The budget proposal’s goal of combining additional data and improving access and usability of existing resources is a worthy endeavor. While the budget proposal stopped short of a specific technical recommendation, these improvements would be most effective if they include a direct link to compiled data, which would enable significantly easier access to the data sources. The Department should avoid the tedious processes necessary for selecting and processing data from sources such as IPEDS, which can be intimidating and time-consuming, creating barriers to use even among savvy data users. Similarly, this project would be best implemented if it were to include an API, which would provide software developers a key starting point for consumer facing applications for use among students, educators, and policymakers alike. All things considered, these improvements would be a great step forward regardless of which party controls the White House.

The InformED initiative also includes $81 million for an additional round of State Longitudinal Data System grants, which are competitively awarded to states to help link data on students from PreK to the workforce. The new grants would include $13 million set aside for state data liaisons, as well as a support and service center, both of which would serve to help state better utilize and improve existing systems. To date, significant funding has been invested in state-level systems, but wide variation has emerged in both the technical elements of such systems, as well as the capacity for state officials to better utilize existing data. Providing technical support could improve state’s capacity. In some states this is already working well: For example, in Utah, researchers have used their state system to understand the impact of math completion on graduation four years later, while Texas has used the data to develop a tool to help students decide where to enroll and what to study. The budget’s proposal to encourage and continue funding for this work could have a huge impact on state-level education policy.

The proposed budget also includes $2.5 million for the development of a new study on college loan default and repayment. In 2017, this would allow for the development and field-testing of a survey instrument to better understand which students default and why, as well as how different repayment plans impact students. This information is vital to protect both consumers and taxpayers, and for institutions to understand how they compare to their peers. But takes time to test, debug, and implement, meaning that policymakers and researchers will remain largely in the dark for at least a few more years. To address this concern, the survey could ld be complemented by more immediate improvements to quality and access to the data in the National Student Loan Data System.

Currently, while institutions can access their own data, quality concerns make it difficult to parse. For example, because the data are recorded for the purpose of administering loans, it’s not always clear whether students who leave a school have completed their degree or withdrawn. Clearly, this distinction has serious implications for schools, policymakers, or students hoping to better understand how they might fare at a given school. Helping schools record this data more consistently could have dramatic improvements for the usefulness of the data in explaining student loan repayment and default. Additionally, since outside researchers are unable to access to the data, largely due to privacy concerns, it can be difficult to conduct objective analytical work on these issues. Maintaining the privacy of students records is critical., which is why the Department already has rigorous processes for granting access to restricted survey data to researchers who must implement strict security protocols, adhere to limitations on data disclosure, and possess significant training and experience in data analysis and use. These could easily be applied to NSLDS and other administrative sources.These improvements would allow more immediate access to information on student loan default and repayment at the institution and state-level, as well as disaggregates of the information according to student demographics or other factors of interest. While the survey proposed in the President’s budget doesn’t suggest new uses for the NSLDS system, such an effort would complement the goals of any new study.

In addition to these, the President’s budget includes funding for cross-sectional and longitudinal surveys, including international surveys, the maintenance of existing administrative data, including IPEDS and the Common Core of Data, and cross-cutting activities including those conducted by the Current Population Survey.

Unfortunately, such a rich budget proposal will likely fall on deaf ears in Congress, where a lame duck President holds little sway amid an increasingly high-stakes election cycle. But many of these suggestions show significant promise. Ideally, these data proposals can be revisited in future years, as the leadership of both parties pursues the data needed to drive evidence-based policymaking. "