Members of the Advisory Committee, thank you for the opportunity to testify today.

I am Anna Cielinski, Senior Policy Analyst at the Center for Postsecondary and Economic Success at CLASP, an anti-poverty organization that promotes effective federal and state policies for low-income people. My specific work addresses performance measures and data systems across postsecondary education, adult education, and workforce development.

My testimony will focus specifically on post-college labor market outcome data, primarily employment and earnings, and the increasing use of such data by states and potentially the federal government not only for consumer information and transparency, but also for high-stakes accountability, such as for Title IV eligibility.

CLASP supports accountability for postsecondary education that moves beyond measuring inputs and focuses on outcomes, including post-college labor market outcomes.  In Congress, the bipartisan “Student Right to Know Before You Go Act” would amend the Higher Education Act (HEA) to create a student level data collection that would provide for employment and earnings data to be collected and reported, in addition to other information. The Department of Education plans to release a data tool that would allow individuals to compare colleges on a number of measures, and it is expected that some form of employment or earnings measures will be included. In addition, a number of states are already successfully publishing labor market outcomes through College Measures websites.

As employment and especially earnings data are becoming more available for consumer information and transparency, there will be great temptation and pressure for policymakers to use these data to hold higher education institutions accountable, in ways that could have unintended consequences.  Once widely collected and published, post-college labor market outcome data are like a genie out of the bottle.

CLASP is very concerned that these data could be used for high-stakes accountability in a way that will hurt access for low-income and under-prepared students and threaten the open-access mission of community colleges. The remainder of this testimony describes these concerns and suggests policies to mitigate them. The analysis is informed by focus groups, interviews, and a survey of postsecondary education practitioners and workforce educators who are particularly attuned to the importance of post-college outcomes. It is also informed by a scan of state outcomes-based funding formulas, voluntary accountability initiatives, performance measures for other federal education and training programs, and College Measures state consumer information websites.

CLASP’s main recommendation is that accountability or funding should not be based on raw institution-wide earnings data, for four reasons:

First, a single average earnings metric for former students from an entire institution masks wide variation in earnings among graduates of different majors or programs of study.

Second, the values and missions of institutions influence institution-level earnings.  For example, the open access mission of community colleges means that students from these institutions may have lower average wages.

Third, regional labor markets differences lead to variation across institutions that is not necessarily reflective of the quality of education. For example, former students from institutions in rural areas will likely have lower average wages than those from institutions located in metropolitan areas.

Finally, institution-level earnings are influenced by program mix. Colleges with the highest wages often have large programs or majors in high-paying fields like engineering.

So, if post-college labor market outcome data is the genie out of the bottle, in the face of the aforementioned challenges, what should Congress, the Department of Education, or states do from a policy perspective? CLASP has four recommendations for using labor market outcomes in a fair way that minimizes unintended consequences for access of low-income, under-prepared students and for the institutions that serve them. For each recommendation, I provide an example of where it is already in place, to show that these are challenges that, with work and creativity, can be overcome.

First, use earnings and employment data disaggregated by major or program of study, not at the institution level. College Measures websites are already successfully doing the hard work of presenting data at the program-of-study level.

Second, take participant characteristics into account to reduce institutions’ incentives for limiting access to lower-income and under-prepared students. Federal workforce development programs use regression adjustment models to take into account participant characteristics to help reduce the incentives for creaming, or targeting services to the most prepared individuals as a shortcut to lifting graduates’ average earnings and job placement results.

Third, take into account regional labor market differences by using economic benchmarks to contextualize wage data. A prototype for this already exists in the Aspen Institute Prize for Community College Excellence, where schools are compared in part on “relative wages,” which are annual wages at 12-month follow up divided by the average annual county new hire wages.

Finally, take into account programs that provide skills that meet community and labor market needs but that may not be high-paying, for example child care workers, EMTs, social workers, and teachers. The Kentucky Community and Technical College System has designed a “social-utility index” to calculate the social good of degree programs that lead to lower-paying jobs that may nevertheless be important to communities.

These four recommendations are more easily said than done, but now is the time to make commitments to such safeguards to prevent the creation of negative incentives that would harm access for lower-income and under-prepared students to higher education. We should not wait until after labor market outcomes data are routinely and reliably available for consumer information and transparency to plan for the use of such data for accountability purposes.

Thank you.