Better Information for Better College Choice & Institutional Performance U.S. Department of Education September 2015 2 Table of Contents Introduction .................................................................................................................................................. 3 Context of Postsecondary Education ........................................................................................................ 4 The Diversity of the Higher Education System...................................................................................... 5 Misaligned Incentives in Other Consumer Information Systems ......................................................... 6 Identifying Performance Metrics That Matter.............................................................................................. 7 Overview ................................................................................................................................................... 7 Access........................................................................................................................................................ 9 Percentage of Pell Students .................................................................................................................. 9 Alternative Access Measures .............................................................................................................. 10 Affordability ............................................................................................................................................ 11 Net Price.............................................................................................................................................. 11 Borrowing and Debt ............................................................................................................................ 13 Outcomes ................................................................................................................................................ 15 Completion Rate ................................................................................................................................. 15 Transfer Rates ..................................................................................................................................... 18 Labor Market Outcomes ..................................................................................................................... 18 Loan Performance Metrics.................................................................................................................. 21 Composite Metrics .............................................................................................................................. 22 Learning and Other Outcomes ............................................................................................................ 26 Other Metrics for Consumers ................................................................................................................. 27 Steps Forward ............................................................................................................................................. 27 Improving the Data ................................................................................................................................. 27 Improving the College Scorecard ............................................................................................................ 28 Changing the Public Discourse ................................................................................................................ 29 3 Introduction In today’s economy, higher education is no longer a luxury, but a necessity for individual economic opportunity, as well as America’s competitiveness in the global economy. At a time when jobs can go anywhere in the world, skills and education will determine success, for individuals and for the health of our democracy and our nation. Over this decade, employment in jobs requiring education beyond a high school diploma will grow more rapidly than employment in jobs that do not; of the 30 fastest-growing occupations, more than half require postsecondary education.1 With the average earnings of college graduates at a level that is twice that of workers with only a high school diploma, higher education is now the clearest path into the middle class. There are a variety of ways that postsecondary institutions prepare students for diverse personal and career goals in their future. Many institutions offer high-quality, affordable educational experiences that expose students to new fields of thought and prepare them to be engaged and productive citizens in their communities. However, some schools do not serve their students well; for instance, they may charge prices that make higher education increasingly out of reach or fail to support students through to completing their education and obtaining well-paying jobs. With such great variation in the types of educational opportunities available throughout the country, it is increasingly important for students and families to have the best information about the educational experiences and outcomes they may expect at different institutions. To that end, the Administration is releasing new information to provide unprecedented transparency about the costs and quality of institutions of higher education: 1  An updated College Scorecard redesigned to provide students, families, and their advisers with a truer picture on college cost and value, and includes the most reliable national data on the earnings of former college graduates and new data on student debt. Rather than highlighting traditional rankings that are constructed to drive colleges to care more, for example, about how many students they reject, this new College Scorecard can empower Americans to compare colleges’ performance based on what matters most to them; highlight colleges that are serving students of all backgrounds well; and keep the focus on ensuring that a quality, affordable education remains within reach.  A new technical site for researchers, policymakers, and others interested in delving more deeply into institutional performance. After exploring several methods for assessing the extent to which institutions contribute to students’ growth and future opportunities, the Administration has produced data and published analyses that share lessons learned and provide considerations for researchers and others in the field regarding factors to consider when building models for evaluating institutional performance. This release represents the best national data on higher education, ranging from demographic information to student outcomes, “Employment Projects: 2012-2022 Summary.” Bureau of Labor Statistics. 19 December 2013: http://www.bls.gov/news.release/ecopro.nr0.htm. 4 from the Department of Education’s National Center for Education Statistics, the Federal Student Aid office, and the Department of the Treasury. With greater transparency around student outcomes at various colleges, the tools needed to identify and promote high-performing institutions no longer rest solely with policymakers and accreditors. On the other hand, students, parents, researchers, and others in the higher education community can serve as better-informed ambassadors and advocates in the conversation about strengthening the higher education system for all. This paper describes the measurements included in the updated College Scorecard and explores how the data can be combined to measure the tradeoffs that exist among outcomes and costs of different institutions of higher education. It accompanies a technical paper that describes the data, and explores their use and limitations in greater detail. Context of Postsecondary Education American higher education comprises a diverse range of colleges and universities that vary significantly in terms of quality and cost, making it challenging to evaluate college performance and difficult for students and families to understand which college options are most suitable to them. Indeed, surveys of Americans reveal that they are looking for more and better information to help evaluate their options.2 Existing college ranking systems focus attention on resources spent, rather than outcomes achieved, and often emphasize selectivity over inclusiveness. At a time when our nation needs more college graduates, and credentials at an affordable cost, these are the exact wrong characteristics to encourage. Moreover, existing rankings do little to focus colleges and universities on improving the effectiveness of academic offerings, strengthening supports that help students to persist in and complete college, and providing increased opportunities for disadvantaged students to earn a college degree. In August 2013, at the State University of New York at Buffalo, President Obama announced that his Administration would work to combat rising college costs, expand opportunity, ensure quality, and make college more affordable for American families. He committed to focus on improving college performance across the critical dimensions of access, affordability, and outcomes—the key goals and expectations for the higher education community, regardless of school mission, location, size, or student body. The new College Scorecard provides free, transparent, and nationally comparable data on the full universe of higher education institutions and their performance on student outcomes, such as graduation rates, student debt repayment, and post-college earnings prospects – information that can help students apply to and enroll in colleges that serve them well. The website will also provide states, colleges, and the public with access to a large database suitable for in-depth analyses to examine 2 “Is College Worth It?” Pew Research Social & Demographic Trends, 15 May 2011. http://www.pewsocialtrends. org/2011/05/15/is-college-worth-it/2/#fn-7679-1; and Fishman, Rachel. “Deciding to Go to College: 2015 College Decisions Survey part 1.” New America, 2015: https://static.newamerica.org/attachments/3248-deciding-to-go-tocollege/CollegeDecisions_PartI.148dcab30a0e414ea2a52f0d8fb04e7b.pdf. 5 questions related to the quality of academic offerings, student supports, factors affecting student outcomes, and other key areas for improvement. With access to better information, the public can engage in a shared effort to strengthen educational opportunities and resources for students from all backgrounds. The College Scorecard represents the Administration’s continued commitment to expanding college opportunities for all students. Since the President took office, the Administration has made historic investments to help Americans pay for college and to reduce the burden of student debt by increasing the maximum Pell Grant by over $1,000, creating the American Opportunity Tax Credit worth up to $10,000 over four years of college, and letting borrowers cap their student loan payments at 10 percent of income. The Administration has also worked to promote innovation and competition to improve the overall performance of our nation’s colleges and universities to ensure they are working to reduce costs, improve quality, and help more students complete their education. A critical part of that strategy has been to help students and families obtain reliable information about college performance to help them select schools that provide the best value, and to encourage colleges to improve by making them publically accountable for the outcomes of the students they enroll. With the resources available through the College Scorecard, college leaders and policymakers now have access to free, high quality, comprehensive, and accurate information that can help inform their efforts to raise graduation rates, bring down costs, and help colleges improve. The Diversity of the Higher Education System Comparing and evaluating the performance of diverse institutions of higher education in order to identify those that provide good value to students based on objective and valid measures presents an array of challenges. Institutions serve students from a wide array of backgrounds, with varied levels of academic preparation and different goals for their education. Differences in students’ needs and institutional resources across higher education create a challenge in assessing institutional performance using shared measures of student success.3 Moreover, institutions have varied strengths. For instance, many community colleges serve and are closely connected to the populations in their area; build strong partnerships with local employers, tie their curricula and program offerings to local labor market needs, and tend to offer skill-building opportunities as well as educational experiences. In many cases, they provide students with affordable opportunities for success at an impressive value. Some colleges excel in preparing students for important careers in public service, such as social workers and teachers. Still others produce graduates 3 Notwithstanding the complexity of comparing the performance or value of institutions in general, in some cases a more focused assessment is appropriate. In particular, title IV of the Higher Education Act requires vocational programs to prepare students for gainful employment in a recognized occupation. 20 U.S.C. §§1001(b)(1) & 1002(b)(1)(A)(i). As the Department has indicated, to meet this obligation, such programs at the least should be enabling students to earn enough money to pay the debts they incur in purchasing their education. 6 in the Science, Technology, Engineering, and Mathematics (STEM) fields and may, as a result, boast high earnings for alumni as a result. Yet despite the diverse higher education landscape, all colleges should meet baseline expectations and advance values that the public generally shares: whether the institution is affordable; the degree to which the institution supports students to and through graduation and prepares them to earn at least a minimal wage and repay their educational debts in the future; and the extent to which the school serves low-income students and serves them well. Data aligned with these expectations will serve as a starting point for meaningful public discourse, and promote a collective effort to understand institutions’ performance and for various audiences. Misaligned Incentives in Other Consumer Information Systems As President Obama noted in his August 2013 speech, many of the incentives in higher education do little to promote an affordable, high-quality education – and often even work against promoting affordability. For instance, the U.S. News and World Report ranking weights spending and school resources as nearly thirty percent of the evaluation, scored six times greater than how students fare after their educational experience. Although a few college ranking systems have attempted to value access and affordability, they focus on only a fraction of the highest-regarded institutions in the U.S. rather than providing information for the majority of colleges. For instance, MONEY magazine’s college rankings consider only about 700 of more than 5,000 total degree-granting institutions, leaving many students unable to access information relevant to their own college selection process. Due in part to pressure from distorted incentives such as those created by some ratings systems, many leaders say that to show their institution is a good choice, they would have to increase selectivity by rejecting more students who apply, admitting fewer disadvantaged students, and implementing policies that drive up costs. Schools face severe pressure to climb the existing rankings, to succumb to the “higher education arms race” of raising tuition and growing more selective as ways to compete with other institutions for higher scores.4 One holdout to the U.S. News and World Report rankings, Reed College, has declined to participate to avoid those misaligned incentives; said the school’s former president Steven Koblik, “The best college is what’s best for the individual student.”5 The Administration’s efforts with the redesigned College Scorecard focus on driving the conversation and the incentives toward what is most essential for students, with their families, in making a decision about where to go to college—what it costs and whether students at the college graduate with more opportunities. As Secretary of Education Arne Duncan has said, “[t]he degree students truly can’t afford 4 Burd, Stephen. “How public colleges use merit aid to compete in the out-of-state student arms race.” The Hechinger Report, 18 May 2015: http://hechingerreport.org/how-public-colleges-use-merit-aid-to-compete-inthe-out-of-state-student-arms-race/. 5 Watson, Harriet. “U.S. News and World Report Hat Trick.” Reed Magazine, November 1997: http://www.reed.edu/reed_magazine/nov1997/news/3.html. 7 is the one they don’t complete, or that employers don’t value.”6 The College Scorecard provides a critical improvement over the information currently available to students and families. And it does something even more important—it holds colleges accountable to the public. The new College Scorecard, accompanying data, and research analyses are the result of teamwork from federal staff and the American public, who engaged with thoughtful ideas and a shared concern for students. In addition to pulling together experts from across the federal government – including the White House’s Council of Economic Advisers, Domestic Policy Council, Office of Management and Budget, and the U.S. Digital Service; the Department of Education; and the Department of the Treasury– the team traveled the country to hear from thousands of students, families, advocates, institutions, researchers, and other stakeholders through bus tours, technical review panels, conferences, and consumer testing to develop the most relevant and responsive college website tool. This collective work will strengthen national efforts to develop meaningful measures of college success for all students in ways that are easy to understand. For instance, we encourage additional thoughtful efforts, like that of the New York Times list of “the Most Economically Diverse Top Colleges,” to ensure that rankings also consider and recognize how institutions provide educational opportunities to support the success of all students, regardless of their family income, geography, or personal background.7 Also recognizing the value of diversity, the Jack Kent Cooke Foundation, a philanthropic organization, introduced its annual economic diversity award – a $1 million prize to an institution that shows a proven track record of enrolling and helping to graduate low-income high-achievers. The College Scorecard will contribute to the Administration’s vision for a sustained national commitment to strengthen college opportunities for all. Particularly in the coming weeks and months, we welcome continued dialogue with students, parents, counselors, colleges, and other stakeholders to further develop, consumer-test, and expand its potential in order to build upon and improve the resources available to help students and families make good college choices, and encourage institutions to improve their performance. Identifying Performance Metrics That Matter Overview A college degree or postsecondary certificate is more important than ever, particularly for low-income students exiting high school and looking to enter the workforce. However, many students, and especially 6 Duncan, Arne. “Toward a New Focus on Outcomes in Higher Education.” Remarks at the University of Maryland— Baltimore County. 27 July 2015. http://www.ed.gov/news/speeches/toward-new-focus-outcomes-highereducation. 7 Leonhardt, David. “The Most Economically Diverse Top Colleges.” The New York Times, 8 September 2014: http://www.nytimes.com/interactive/2014/09/09/upshot/09up-college-access-index.html?_r=0. 8 underserved students, feel unprepared for the choices they need to make and may not even fully understand that college is within reach.8 In exploring the metrics that best represent those categories of information, the Department of Education evaluated all available data sources, from publically available data, including the Integrated Postsecondary Education Data System (IPEDS), as well as newly produced data from the National Student Loan Data System (NSLDS)9 and from the Treasury Department’s federal wage records, and nonfederal data sources like the Student Achievement Measure (SAM).10 We identified particular elements that represent schools that are providing affordable, high-quality educational opportunities, particularly to students from low-income families. As detailed below, the elements we selected revealed several exemplar institutions that serve students well. For example, the graduation rate at Georgia Tech—a predominantly four-year institution located in Atlanta, Georgia—is in the top 10th percentile of four-year schools (80 percent), with median earnings 10 years after entering the school of more than $74,000. The lowest-income students, at Georgia Tech, pay an average of $7,364 per year, and nearly one-fifth of students are Pell Grant recipients. A very different school, State Technical College of Missouri in Linn, Missouri, stands out among predominantly two-year colleges.11 More than 40 percent of its students receive Pell Grants; the lowest-income students pay $7,783 per year, on average, and borrow less than half the federal student loan debt of Georgia Tech graduates; and six years after entry, more than seven in 10 students earn more than the national average annual earnings for high school graduates aged 25 to 34, exceeding many other two-year institutions (see Table 1). Table 1. Georgia Tech (Atlanta, GA) State Technical College of Missouri (Linn, MO) 8 Net Price for the LowestIncome / HighestIncome Students $7,364 / $14,114 $7,783 / $10,382 Median Completion Median Debt of Rate Earnings Completers $22,750 80% $74,000 Share of Fmr. Students Earning More than HS Graduate 86% $10,500 60% $36,400 71% Share of Pell Recipients 42% 19% “Is College Worth It?” Pew Research Social & Demographic Trends, 15 May 2011: http://www.pewsocialtrends. org/2011/05/15/is-college-worth-it/2/#fn-7679-1. 9 Several of the new NSLDS measures mirror similar measures for Gainful Employment programs, but data are now available at an institution level for all schools; for instance, the cumulative median loan debt of graduating students is now being released for all institutions. Newly constructed NSLDS completion and transfer rates and federal student loan repayment rates are also produced across all institutions. 10 “A New System of College Ratings—Invitation to Comment.” U.S. Department of Education, December 2014: http://www2.ed.gov/documents/college-affordability/framework-invitation-comment.pdf. 11 This institution was formerly known as Linn State Technical College. 9 Access The primary goal of the federal student aid system is to provide access to high-quality higher education for low-income populations. The legacy of the Title IV aid programs rests on the millions of low- and middle-income students who have successfully completed degrees, found well-paying and rewarding careers, and can support their families as they pursue their own educational opportunities.12 Recognizing and rewarding institutions that play a critical role in providing educational opportunities to hard-working, low- and moderate-income students—and noting those that have not succeeded in ensuring access to low-income students, and/or that have not served low-income students well—is an important element in examining college performance. Some institutions that admit more Pell Grant students than others may provide more aid to low-income students, or offer more support to help lowincome students complete their education. On the other hand, schools that fall short in these areas can negatively impact a student’s chance of completing college and transitioning to the workforce. Percentage of Pell Students For the College Scorecard, we measure low-income students’ access to education based on the share of Pell Grant students that the institution enrolls using IPEDS data. The Pell Grant program, which has provided grants to low- and moderate-income students since its inception in the 1970s, forms the cornerstone of efforts to increase access for disadvantaged students. This metric is widely recognized and understood within the field as a proxy for the financial circumstances of enrolled students and their families. Some institutions do well by the disadvantaged students that they do enroll, but serve only a small number of them. Ivy League schools like Harvard University, Columbia University, and Princeton University have some of the lowest net prices for students in the bottom two quintiles of family income ($0 to $48,000), low typical loan debt for students, and high graduation rates and earnings. However, these selective institutions tend to be among those in the bottom 10 percent of all four-year institutions whose students receive Pell Grants (see Table 2).13 Table 2. Harvard University Stanford University Columbia University Princeton University 12 Share of Pell Recipients 10% Net Price for the LowestIncome Students $3,897 Median Debt of Completers $6,000 Completion Rate 97% Median Earnings $87,000 16% $3,516 $12,224 95% $81,000 22% $8,086 $19,435 94% $73,000 12% $5,932 $6,810 96% $75,000 Title IV of the Higher Education Act authorizes the federal loans and grants administered to students by the U.S. Department of Education. 13 These data measure the outcomes from the most recent cohort available. 10 Many public colleges are pledged to an historic mission of serving low- and moderate-income students, and some stand out as serving them especially well. For instance, a large share of students at the University of California—Los Angeles (UCLA) receive Pell Grants (36 percent), and the school has a high overall cohort graduation rate (91 percent) among first-time, full-time students and below-average debt. Similarly, about 20 percent of undergraduates at the University of Illinois at Urbana-Champaign receive Pell Grants, and the school reports excellent outcomes, including an 84 percent completion rate for firsttime, full-time students and median earnings of more than $57,000 10 years after entering the school (see Table 3). Table 3. Pell Recipients UCLA University of Illinois at UrbanaChampaign 36% 20% Net Price for the Lowest-Income Students $8,883 $7,954 Median Debt of Completers Completion Rate Median Earnings $15,900 $20,950 91% 84% $59,000 $57,000 Alternative Access Measures Students can use the College Scorecard to identify schools that serve disadvantaged students. These student populations, which may be geographically constrained in the locations they can consider, often have less guidance available to them, may do less research prior to selecting a school, and are vulnerable to choices that may lead them to enroll in a school with fewer opportunities for them to succeed.14 The Administration is releasing additional data that can help inform higher education stakeholders about the quality of educational services those low- and moderate-income students received, which is a critical component of ensuring equitable access to a high-quality education for all students. Several of the data elements published through the technical page of the College Scorecard disaggregate key metrics—completion rate, federal loan repayment rate, and median debt, for instance—by family income and for Pell Grant recipients. These metrics offer additional details on how well schools serve specific subgroups, like first-generation or low-income students, rather than simply meeting an average bar for the entire student population. The disaggregated data may also help schools to identify their own shortcomings and develop solutions. The data produced also include disaggregated enrollment information addressing the income breakdown of federal financial aid recipients at the institution, those who are first-generation students, 14 Fishman, Rachel. “Deciding to Go to College: 2015 College Decisions Survey part 1.” New America, 2015: https://static.newamerica.org/attachments/3248-deciding-to-go-tocollege/CollegeDecisions_PartI.148dcab30a0e414ea2a52f0d8fb04e7b.pdf. 11 the racial/ethnic makeup of the student body, and more, all of which can provide important evidence of the degree to which schools serve historically disadvantaged populations. Indeed, these elements can contribute to a more complete picture of institutional performance. Affordability Affording college is one of the concerns at the forefront of students’ and parents’ minds as they explore the college selection process. Families’ out-of-pocket costs have continued to rise, in part because of the economic downturn, which precipitated further declining state investments in public higher education.15 In Wisconsin, state higher education appropriations per student as of fiscal year 2014 were reduced nearly 20 percent since 2008, before the recession. Over the same period, tuition at Wisconsin state institutions increased by more than 30 percent. Similarly, Florida reduced its per-student appropriations by 32 percent from fiscal year (FY) 2008 to FY 2014, and tuition rose 53 percent over that time period.16 At the same time, fewer public institutions are helping make up the difference in costs for low-income students. Many public colleges and universities—including well-resourced ones—are reacting to budget constraints, contracting enrollment, and college rankings that emphasize spending over outcomes by diverting their institutional aid to attract high-performing students, which can drive up costs without improving quality.17 Net Price Figure 1 The combination of Stanford University provides low average costs for falling state the lowest-income students investments, redirected institutional $70,000 $58,408 aid, and rising costs $60,000 force many students to $50,000 $40,323 wonder which—if $40,000 any—colleges are $30,000 worth the cost. But $20,000 finding information about the true costs of $10,000 $3,516 college can be difficult, $0 and the information Cost of Attendance Net Price for the Net Price for the Lowest-Income Title Highest-Income Title that exists for students IV Students IV Students can be misleading. Prospective students are often presented with a school’s tuition and fees, which can understate the costs of attendance by excluding the living costs and additional costs of books and other supplies. Other 15 “Out-of-Pocket Net Price for College.” U.S. Department of Education, April 2014: http://nces.ed.gov/pubs2014/2014902.pdf. 16 “State Higher Education Finance: FY 2014.” State Higher Education Executive Officers Association, April 2015: http://www.sheeo.org/sites/default/files/project-files/SHEF%20FY%202014-20150410.pdf. 17 Burd, Stephen. “Undermining Pell: Volume II.” New America, September 2014: https://www.newamerica.org/downloads/UnderminingPellVolume2_SBurd_20140917.pdf. 12 students may see the full cost of attendance, which may overstate the cost because it does not account for federal, state, local, or institutional aid—possibly substantial—that students may receive.18 For consumers looking to compare college costs, the best publically available, comparable information is the net price of the school across students’ income brackets, as it provides a more personalized number that allows students to better gauge the actual price they will need to pay to attend the school. These data also help policymakers, researchers, and institutions to identify inequities in the distribution of need- and merit-based aid. The metric used here captures the total cost of attendance, less federal, state, local, or institutional aid, for students, based on the income of the student and/or his family, on average.19 Presenting this information tailored to each student’s income can show that schools they thought were too expensive may actually be affordable for them. For instance, students who received federal financial aid at Stanford University and who come from families earning $30,000 per year or less can expect to pay, on average, $3,516—just a fraction of the overall cost of attendance for the 20122013 academic year ($58,408). Even Title IV-receiving students from families earning over $110,000 annually pay less than the full, posted cost of attendance ($40,323) (see Figure 1). The Department of Education already helps students understand costs in terms of net price, recognizing it as the most accurate available measure of what students can actually expect to pay. In 2008, Congress required institutions of higher education to create net price calculators that generate individualized cost estimates for the freshman year. 20 The Department’s Net Price Calculator Center compiles the information from all institutions, allowing students to easily access their top schools’ calculators. For the purposes of college comparison on the redesigned College Scorecard, the Administration chose average net price calculated across the five income quintiles, an element reported through the Integrated Postsecondary Education Data System (IPEDS) for every institution and that provides a reasonable expectation of what Title IV students (those eligible for federal grants and loans) pay. More detailed institutional pages also offer net price information by income quintile, to help lower-income students especially get a more accurate sense of their cost of attendance.21 To date, this is the most consistent and individualized source of information for each school. Furthermore, the data are listed on the College Scorecard’s profiles for each institution, where students can weigh the metric side-by-side with other metrics that illuminate such important questions as their odds of graduating, their prospects in the job market, and the amount of debt they are likely to take on. 18 Low-income students and families, particularly Latino parents, frequently overestimate the costs of college. “Paving the Way: How Financial Aid Awareness Affects College Access and Success.” The Institute of College Access and Success, October 2008: http://ticas.org/sites/default/files/pub_files/Paving_the_Way.pdf. 19 The net price calculation used on the College Scorecard is derived from a weighted average of the five income quintiles reported in IPEDS. We have also provided some of the information by income bracket to aid students in identifying the most granular information possible. 20 Levine, Phillip. “Transparency in College Costs.” Brookings Institution, 12 November 2014: http://www.brookings.edu/research/papers/2014/11/12-transparency-in-college-costs-levine. 21 Income quintiles are defined as follows: $0-$30,000; $30,001-$48,000; $48,001-$75,000; $75,001-$110,000; and $110,000 or more. 13 For instance, at Columbia University in New York City, students in the lowest income quintile can expect to pay an average net price of $8,086 per year. That price is substantially lower in cost than neighboring institution New York University, where the net price for students in the lowest income quintile is $25,441. Moreover, Columbia has a completion rate for first-time, full-time students of 94 percent, and median earnings 10 years after entry of nearly $73,000. At NYU, though outcomes are still relatively positive, the completion rate is lower at 84 percent, its median earnings average closer to $58,800, and graduating students carry a similar amount of loan debt at both schools (see Table 4). Knowing that information, students who are weighing the two schools stand a far better chance of making an informed, carefully considered decision. With that information, students can be better prepared to delve deeper into the programs and other specifics of the two schools. Table 4. Columbia University NYU Net Price for the Lowest-Income Students $8,086 $25,441 Median Loan Debt of Completers $19,435 $23,250 Completion Rate Median Earnings 94% 84% $73,000 $58,800 It is also crucial to note that the total cost of attending and graduating from college depends largely on the number of years a student takes to complete. If a student takes longer to graduate, and pays for more credits or terms, he will face higher costs than a more efficient completer. An $8,000 per year bachelor’s degree program that takes six years to complete still costs more in the end ($48,000) than a $10,000 per year, bachelor’s degree program in which most students complete on time ($40,000), despite a $2,000 additional annual cost. Given that most students who default on their federal student loans are those who never complete a degree22, a marginal additional cost each year to attend an institution that provides substantially greater academic supports and has higher completion rates and employment outcomes is one that may be a wiser choice for many students. Borrowing and Debt The Administration has also worked to reduce the burden of student loan debt. The Financial Aid Shopping Sheet is designed to help students disentangle the types and amount of loans the typical student can expect to borrow each year. The interest rate reform that the Administration signed into law has resulted in lower interest rates on federal undergraduate student loans. And reforms and expansions to income-based repayment options, like the establishment of Pay As You Earn to limit loan payments to 10 percent of a borrower’s income over 20 years, will help make loan payments more affordable, particularly for low- and middle-income alumni. Students and families can and should consider their expected levels of indebtedness when searching for and selecting a school. While federal student loans enable millions of students to enroll in and complete 22 McCann, Clare. “College completion is the best default aversion.” The Hill, 13 October 2014: http://thehill.com/blogs/pundits-blog/education/220532-college-completion-is-the-best-default-aversion. 14 higher education, excessive amounts of debt should serve as a warning to students. In a consumerfacing college selection tool, the debt levels students who borrow can expect to bear through graduation, and the payments they can expect to owe each month thereafter, along with information on these other metrics, are relevant. The College Scorecard site includes the median debt of graduates from each institution who borrow, and the typical monthly payments due on that debt level, based on a 10-year repayment plan, though payments could be lower on an income-driven or another repayment plan. The Department previously reported debt of all former students, which may perversely make an institution appear more affordable if its students are less likely to persist. Knowing the full debt typically required of borrowers to earn a college degree, rather than just the cost of taking classes that fall short of a degree, provides a more realistic picture of affordability to students. Details of student debt—presented as the overall loan balance borrowed and as monthly payments—are included for each institution. In addition, the College Scorecard’s technical page includes even more information about the debt levels of students who borrow at each institution, including the total median debt that incorporates data for both borrower completers and non-completers. The data are disaggregated in several ways, including family income, first-generation status, and Pell Grant recipient status. The new information will provide additional nuance to researchers, counselors, and advocates exploring the borrowing behaviors at a school to help hold institutions to high standards across all of the students they serve.23 Clear information on debt is essential as students consider their choices. For instance, at the predominantly two-year Bellevue College outside Seattle, Washington, graduates who take out loans typically leave with just $12,224 in federal student loan debt—about $136 per month in monthly payments; and the institution has a completion rate of about 26 percent and median earnings of nearly $42,000. But at the neighboring ITT Technical Institute in Seattle, a two-year, for-profit institution, students take on far more debt for similar completion rates (37 percent) and post-school earnings ($42,500). Debt levels for ITT Tech students total around $27,833 for students who graduate, or about $309 per month in student loan payments—more than twice what Bellevue students owe after completing with similar post-school outcomes (see Table 5). The College Scorecard will help students to evaluate their expected federal student loan debt against the payoffs of attending the school. Table 5. Bellevue College 23 Share of students borrowing federal loans 10% Median loan debt $12,224 ($136) Completion rate 26% Median earnings $41,300 One element that we are unable to provide at this time is the amount of private student loan debt that students take on per institution. These loans frequently offer less-generous terms for students who fall on hard times or find their degree cannot bear its own weight in the job market; and they may have much higher interest rates or require a credit check that many undergraduates are unlikely to pass. Data on private student loan borrowing can offer an important measure of students’ expected financial health upon leaving the school. 15 ITT Technical InstituteSeattle 75% $27,833 ($309) 37% $42,500 Outcomes We received hundreds of comments and spoken testimony suggesting particular elements we should examine to determine a student’s success. One of the driving factors in announcing the College Scorecard project was the concern that too many students have their futures shortchanged because they attend schools that do not serve them well—either because the institution doesn’t get the student to graduation, or because the student completes the course of study but finds their credential or degree is not valued in the labor market—and then are left saddled with debt but with few opportunities. These considerations should be a major factor in students’ college choices. At 53 percent of institutions, more than half of alumni are not even earning more than a typical high school graduate within six years after starting at the school.24 Completion Rate One widely recognized metric of student success is the rate at which students are able to earn their degrees. Student borrowers who fail to obtain a degree are three times as likely to default on their federal student loans, and are less likely to find well-paying jobs, than those who complete their programs.25 Completion rates are helpful indicators of institutional quality for students, families, and researchers. Furthermore, for subgroups of students, such as first-generation and low-income students, completion rates can help determine how well the school is supporting its neediest populations. The Administration has taken major steps to highlight the importance of completion on improving students’ ultimate outcomes. The Department of Education has proposed and implemented a number of reforms that incent institutions and other higher education partners to emphasize the importance of college completion through proven, evidence-based strategies and interventions. For instance, the, Administration’s First in the World program, as well as the 2015 TRIO Student Support Services grant competition, emphasize the importance of completion. Both competitions have helped to identify interventions that show positive outcomes, and that meet standards for high levels of evidence, suggesting they can have a real impact on students’ likelihood of finishing their degrees. America’s College Promise, proposed by the Department of Education in its fiscal year 2016 budget request, would encourage and support state and institutional efforts to improve student outcomes through, for example, requiring community colleges to adopt promising and evidence-based institutional reforms and innovative practices to improve student outcomes. In addition, the Administration has proposed reforms to campus-based student aid programs, which would target those institutions that enroll and graduate higher numbers of Pell-eligible students, and offer affordable and quality education and training such that graduates can obtain employment and repay their educational debt. And in the Department’s Gainful Employment regulations, we recognized that “it is important to hold institutions 24 This percentage defines the institution at the six-digit OPE ID level, and includes any institution with Treasury data, regardless of predominant degree type or other characteristics. 25 “Fact Sheet: Focusing Higher Education on Student Success.” U.S. Department of Education, 27 July 2015: https://www.ed.gov/news/press-releases/fact-sheet-focusing-higher-education-student-success. 16 accountable for the outcomes of students who do not complete a GE program,” and are requiring institutions to disclose the completion rates for any of their programs covered by the regulation.26 Institutions themselves are also generating many new strategies to support student completion. The Accelerated Study in Associate Programs (ASAP) program implemented by the City University of New York (CUNY) is one example. CUNY’s educational model provides additional resources and supports to help community college students attend school full time, through funding the cost of attendance and through enhanced advising efforts. Rigorous studies have shown that students participating in the ASAP program earned more credits, graduated at nearly twice the rate of completion, and transferred to fouryear degree programs at higher rates than similar students at the institution. It is important to consider what a graduation rate does and does not tell a student about a school. For a school at which programs are typically of short duration—such as six or nine months—the graduation rate may be higher than at a school that typically confers four-year degrees, where the time commitment to graduate is significantly higher. But a four-year degree typically also has higher pay-offs than short-term programs. Both shorter- and longer-duration programs can be right for some students, and the graduation rate tells students what share of their peers finish the programs offered at that school. However, graduation rates must be considered alongside other contextual information, such as the kinds of programs or degrees offered and the employment outcomes of individuals who go to those schools. Importantly, many less-than-two-year schools have high completion rates but poor labor market outcomes. This suggests that the credentials or degrees offered by a school do not necessarily help them in the job market. Consider, for example, Golf Academy of America, which offers online or brick-and-mortar options for associate’s degrees in Golf Complex Operations and Management. Despite a completion rate of 85 percent at the school’s Myrtle Beach, South Carolina location, median earnings 10 years after entering the program fall barely above that of the average high school graduate at $26,400 (see Table 6). Table 6. Golf Academy of America Median Loan Debt of Completers $19,000 Completion Rate Median Earnings 85 % $26,400 Similarly, schools that serve large numbers of disadvantaged students may have lower completion rates than schools that may be less accessible to those students. That does not necessarily mean that disadvantaged students are better served in less-accessible schools, however. In this case, the graduation rate data should be considered alongside data that provide insight into how accessible and successful the school is for disadvantaged students. 26 Program Integrity: Gainful Employment; Final Rule. 34 Fed. Reg. Parts 600 and 668 (October 31, 2014): 64928. http://www.gpo.gov/fdsys/pkg/FR-2014-10-31/pdf/2014-25594.pdf. 17 The most commonly referenced completion rates are those reported to IPEDS and are included on the College Scorecard (measuring completion within 150 percent, or six years, for predominantly four-year colleges; and within four years for predominantly two- or less-than-two-year schools). However, they rely on a school’s population of full-time students who are enrolled in college for the first-time. This is increasingly divergent from the profile of the typical college student, particularly at many two-year institutions and some four-year schools. For instance, Marylhurst University in Oregon, a four-year institution that has been recognized for serving adult students, reportedly had a 23 percent, six-year completion rate – namely because a very small subset of its students (just one percent) fall in the firsttime, full-time cohort used to calculate completion rates. As with many schools that serve students who already have some college experience, this rate is, therefore, hardly representative of the school’s student body. The Department also considered, as described in the December 2014 Framework for the College Scorecard project, external efforts to improve the measurement of completion rates to include more students.27 While the Department is not including those data on the site, in part because the data are submitted voluntarily, these initiatives have helped to inform our work and other efforts in the field and should continue to be a part of the conversation on measuring institutional performance. The Department has previously announced plans to work with colleges and universities to improve the graduation rates measured by the IPEDS system. Beginning in 2016, colleges will begin reporting completion rates for the other subsets of their students: first-time, part-time students; non-first-time, full-time students; and non-first-time, part-time students. In the meantime, by using data on federal financial aid recipients that the Department maintains in the National Student Loan Data System (NSLDS) for the purposes of distributing federal grants and loans, we constructed completion rates of all students receiving Title IV aid at each institution. For many institutions, Title IV completion rates are likely more representative of the student body than IPEDS completion rates – about 70 percent of all graduating postsecondary students receive federal Pell Grants and/or federal loans. Given concerns about the quality of historical data, these NSLDS completion rates are provided on the technical page, rather than on the College Scorecard itself.28 We also produced and published, on the technical site, information on the completion rates of subgroups of Title IV students collected in NSLDS, 27 For instance, the Framework document said that “in a future iteration of college ratings, the Department may allow institutions to voluntarily submit their Student Achievement Measure (SAM) completion/transfer rate…” The Student Achievement Measure is a system of voluntary completion and transfer rate reporting that encompasses non-first-time, -full-time students as well as first-time, full-time students. 28 We identified some gaps in reporting from institutions of higher education that suggest those rates are not fully accurate. Prior to a 2012 Dear Colleague Letter from the Department of Education, institutions were not required to report enrollment and completion information for Pell Grant students or student loan recipients; therefore, the reporting for those students is often spotty, and the rates may be less representative. For instance, Harvard University has a reported six-year graduation rate for Pell Grant students of just 37 percent—implausibly low, particularly given an 80 percent completion rate for all of its Title IV students. The Department plans to rerun the data during the first year to permit institutions time to update their historical reporting. 18 including low-income students and first-generation students. These rates can help schools to identify gaps in their support of students, benchmark with other schools serving similar student populations, and improve the outcomes of populations within their schools that may need additional assistance to complete college. Transfer Rates The Administration also believes it is important that the College Scorecard address students who transfer to a higher degree program. Many students receive great value in attending a two-year institution first, and eventually transferring to a four-year college to obtain their bachelor’s degrees. In many cases, the transfer students do not formally complete the two-year program and so do not receive an associate degree prior to transferring. When done well, with articulation agreements that allow students to transfer their credits, this pathway can be an affordable and important way for students to receive four-year degrees. In particular, according to a recent report from the National Center of Education Statistics (NCES), students were best able to transfer credits when they moved from two-year to four-year institutions, compared with horizontal and reverse transfers.29 The Obama Administration has taken steps to ensure that strong pathways exist between two- and fouryear institutions. The First in the World program encourages the development of articulation and transfer agreements between two- and four-year institutions. The proposed America’s College Promise initiative would provide a matching grant to states that agree to strengthen articulation and transfer agreements to ensure that every student has access to a free community college education, and an opportunity to transfer credits to a four-year-school, among other things. To measure the successes of institutions that support students to transfer to higher degree programs, we calculated transfer rates for Title IV-receiving students who moved on from two-year to four-year colleges using a methodology similar to the NSLDS completion rates. Those data are included on the data and technical site of the College Scorecard.30 Labor Market Outcomes The Department has developed, and made available for the first time, data on the post-school earnings of federal student aid recipients. Included within this new set is data disaggregated by students’ precollege family income, which may provide additional information concerning economic mobility for students at the institution.31 29 “Transferability of Postsecondary Credit Following Student Transfer or Coenrollment.” National Center for Education Statistics, August 2014: http://nces.ed.gov/pubs2014/2014163.pdf. 30 We hope to be able to produce those figures for consumers after correcting for the same reporting limitations as exist for the completion rates. 31 While it is not feasible for the Department to provide an analysis of the direct impact an institution has on a former student’s earnings by looking at his individual earnings before and after attending the school, disaggregating post-enrollment earnings by a student’s family income, as reported on the FAFSA to determine the student’s eligibility for student aid, allows researchers, policymakers, institutions, and others to see if the institution was able to contribute to the student’s upward social mobility. It is important to note that considering upward social mobility may be an important factor in assessing how well a school serves its students, but it does 19 Post-school earnings provide an additional way to examine the contributions of postsecondary education, and the Administration collects and supports and partners with others in collecting information on students’ labor market outcomes. The Statewide Longitudinal Data System (SLDS) program funds state data systems that link student data from PreK-12 through higher education; and the Department has provided additional funds to support states in connecting those data with labor market outcomes. In addition, the Department has introduced regulations for career-oriented programs that measure the ratio of debt-to-earnings of a program’s graduates; with those data, the Department will be able to hold career-oriented programs accountable for preparing their graduates for gainful employment in a recognized occupation as required by law.32 Moreover, through the implementation of the Workforce Investment and Opportunity Act, the Departments of Education and Labor are working toward performance standards using labor market outcomes for federally funded workforce programs. Additionally, the Trade Adjustment Assistance Community College and Career Training program has helped schools develop partnerships to improve labor market outcomes, including coordinating labor market information with states. Earnings are not the only reason—or in some cases, even the primary reason—that most students enroll in college. Nevertheless, future employment is an important factor in many students’ decision to seek a college degree, and prospective students and their families currently do not have access to reliable information about future employment and earnings Figure 2 potential that each college can offer. It is important for students to be able to evaluate whether or not The typical share of students who the cost of an institution is worth the investment, earn more than a high school including possible student debt. We have provided graduate can vary substantially both short-term and long-term earnings data on across schools the College Scorecard, in order to help students 80% understand their career earnings prospects upon 60% leaving a particular institution. 40% 20% 0% Midwest Institute State Technical College of Missouri The short-term measure of employment and earnings is the share of former students at an institution who earn more than the average high school graduate in the U.S. – about $25,000 per year, six years after entering the school, regardless not account for all reasons why earnings are an important factor to consider. Other metrics, such as if a student earns enough to pay off the debt he accumulated to attend an institution, provide vital information to both consumers and those holding institutions accountable for their performance. 32 “Fact Sheet: Obama Administration Increases Accountability for Low-Performing For-Profit Institutions.” U.S. Department of Education, 1 July 2015: http://www.ed.gov/news/press-releases/fact-sheet-obama-administrationincreases-accountability-low-performing-profit-institutions. Under the Gainful Employment Rule, the estimated annual loan payment of a typical graduate may not exceed 20 percent of his discretionary income or 8 percent of his total earnings (over several years) to be eligible for federal student aid programs (Department of Education, 2015). 20 of whether the student completed the program. This measure was created of the belief that all colleges should be able to provide benefit to students beyond just a high school degree. The measure identifies schools with low performance—where few students earn above the threshold—but is less useful in distinguishing the labor market outcomes among higher-performing schools. However, it serves a particular purpose: to help prospective students identify whether they can realistically expect that attending this school will provide a minimum level of employment and afford them value above and beyond a high school diploma. At the same time, prospective students need to consider any projected income increases against the higher-education debt that they will be repaying. At some schools, the data suggest that students cannot expect that a college certificate or degree will create an earnings premium. For instance, at Midwest Institute in Fenton, Missouri, barely 30 percent of students earn more than a typical high school graduate within six years after entering the school. Those students fare similarly poorly on other measures of post-school outcomes; they have a fairly low student loan repayment rate of 36 percent, despite a median debt of completers of less than $10,000. However, at the nearby State Technical College of Missouri, fully 71 percent of former students earned more than a high school graduate within six years of enrolling; the median loan of graduating students is relatively low at $10,500; and 83 percent of former students are making progress on repaying their loans at the three-year mark after leaving school (see Figure 2).33 These data can provide valuable information on the immediate earnings potential of students at the Figure 3 school. Students must weigh the tradeoffs of costs and labor market outcomes Art Institute of Colorado Colorado State University--Fort Collins $0 $15,000 Net price for the lowest-income students 33 $30,000 $45,000 Median earnings We also produced a longterm earnings measure, which depicts the median earnings of students 10 years after entry into the school—regardless of whether the student completed a degree or certificate.34 This measure provides more detailed information about the earnings of former students as they expand This institution was formerly known as Linn State Technical College. This measure of median earnings 10 years after entry accompanies another, shorter-term (threshold) earnings measure. For programs not necessarily designed to lead directly to employment and/or a career—for example, liberal arts programs—a longer-term measure may provide a better sense of the eventual payoff for students. For programs meant to lead directly to gainful employment, it is critical to look at earnings closer to the time the student completed the program to ensure the student is making enough money to pay off the debts he accumulated to attend the program. 34 21 their job skills and build their careers. For instance, at Colorado State University (CSU)—Fort Collins, median earnings of alumni total about $44,300 ten years after entering. In contrast, at the nearby forprofit Art Institute of Colorado, students pay thousands more ($24,863 for the lowest-income students at Art Institute compared with $9,898 for the lowest-income students at CSU—which offers an art program), for earnings that are$10,000 lower ($32,400) than those of their CSU counterparts (see Figure 3). A median earnings measure also offers colleges and universities critical information with which they can begin to benchmark their own students’ results in the workforce and improve their academic standards and career services to align with their students’ goals and needs. While some institution may be focused on programs meant to immediately lead to gainful employment, in which case incorporating earnings as a measure of student success is an appropriate choice, other schools may emphasize public service, where students may go on to lower-paying but rewarding careers that offer significant non-monetary benefits, or could be religious institutions, where alumni go on to serve in temples, mosques, and churches. Some schools educate those in primarily science, technology, engineering, and/or mathematics (STEM) fields, who typically earn more after college.35As such, it is important to consider the purpose of the programs in which students enroll when determining how earnings data can help paint a fuller picture of student success. While the data do not currently allow the Department to examine earnings for all programs of study, it will soon be possible. Threshold and median earnings measures offer students an additional source of information to inform their college choices. Median earnings are included as a search criterion; students can select their target earnings range to filter out schools outside of that range. Threshold earnings, meanwhile, are used to sort institutions on the College Scorecard–a rough measure for consumers to identify the degree to which students are better off. Both measures are available on the institution’s page. Additionally, the median earnings metric is Figure 3 also provided on the technical page, Repayment rates span a wide range disaggregated by student across institutions demographic information, 100% including family income 14% and dependency status. 80% Loan Performance Metrics Loan performance metrics can help students assess their likely outcomes is the ability of prior students to repay their loans after 35 60% 63% 86% 40% 20% Not In Repayment In Repayment 37% 0% Argosy University--Washington, D.C. University of Maryland-Baltimore County Note that whatever their chosen missions or emphases, certain programs are also legally required to produce a certain outcome as a condition of their participation in title IV: They must prepare students for gainful employment. For such schools, debt- and earnings-related standards must also be satisfied. 22 leaving school. The loan repayment rate metric is also of keen interest to federal policymakers, given that taxpayer dollars fund federal student loans. Unlike the cohort default rate, which measures only the worst-case scenarios of former students’ inability to repay their loans, loan repayment rates measure what most student borrowers expect to know once they leave school: the ability to pay back their student loans. In this case, repayment rates are measured by the share of former students (both completers and non-completers) with federal loans who are able to pay down at least $1 of the principal balance within three years upon leaving the institution. The measure for loan repayment rates has been widely supported by many in the higher education community.36 In 2009, the Obama Administration released repayment rates for all institutions. Since 2009, advocates for greater transparency in higher education have issued calls for the data to be released on an ongoing basis for all institutions. The data published on each institution’s page on the College Scorecard respond to those calls for more transparency around loan performance outcomes. Indeed, a school’s repayment rate can provide a much clearer picture for students in making decisions pertaining to future student loan debt. On the whole, the data regarding loan repayment reveal significant differences across institutions. Fouryear for-profit colleges, for instance, average a repayment rate of 44 percent, compared with an average of 80 percent at public four-year institutions. This suggests students are more able to repay their loans at public schools. This also suggests that students at public schools borrow less, have better labor market outcomes, or both. Among particular institutions, the variations in repayment rate are even more striking, and illustrate the importance of looking at measures other than total student debt. At Argosy University, a four-year, for-profit chain of institutions, for example, the typical student (including both completers and non-completers) at the Washington, D.C. campus leaves school with about $9,500 in federal student loan debt. The average loan debt among graduates from the Argosy, Washington, D.C. campus, is $14,000. That campus has a repayment rate of about 37 percent – meaning that nearly six in 10 former students have not paid down a single dollar of their principal balance within three years after leaving school. In contrast, nearby University of Maryland—Baltimore County has an 86 percent repayment rate, despite its higher debt levels for graduates who, on average, borrow about $21,500 (see Figure 4). Composite Metrics While individual metrics can generate useful insights into colleges’ missions and performance, creating a composite metric may be useful in illustrating the multi-faceted aspects of institutional performance. For instance, examples of colleges with low costs or high enrollment of Pell Grant recipients alone may lack information about whether those colleges have a positive impact on student outcomes, such as 36 “Comments on Notice of Proposed Rulemaking on Gainful Employment.” New America, 27 May 2014: http://www.edcentral.org/wp-content/uploads/2014/05/05-27-14-New-America-Gainful-EmploymentComments.pdf; “Closing GE Loopholes: The Case for a Repayment Rate.” The Institute for College Access & Success, 5 June 2014: http://ticas.org/blog/closing-ge-loopholes-case-repayment-rate; and “Student Loan Repayment Rates Hold Ineffective Schools Accountable.” Young Invincibles, 17 June 2014: http://younginvincibles.org/student-loan-repayment-rates-hold-ineffective-schools-accountable/. 23 graduation rates or post-school employment. Therefore, metrics that incorporate more than one measure of institutional performance, though not included on the College Scorecard, can be valuable to students, families, and policymakers. In addition, no single metric can capture all of the purposes, missions, and outcomes of any single institution; knowing the earnings of a school without the debt students will owe after leaving the school, for instance, provides little context to evaluate the tradeoffs of a particular program and whether it actually prepares students for gainful employment. Some institutions specialize in certain programs that may produce more graduates in high-paying industries. Others commit to expanding college opportunity and pathways to the middle class by enrolling and graduating large proportions of students from all backgrounds. Others focus on preparing a high proportion of their students for graduate school. At programs designed to promote vocational education and help students enter a career, a composite measure of debt and earnings provides the clearest view of the payoff of the program. Composite metrics require subjective judgments over the weight of each criterion. Using available data, the Department of Education analyzed a range of composite options that could serve as a starting point for understanding colleges and universities and conveying information about the basic goals that students, families, and policymakers commonly share when assessing the value of an institution of higher education.37 As a result, the Department identified three approaches for developing composite metric approaches, aspect of an institution’s performance, and explored the strengths and weaknesses of each approach.    Engines of opportunity: Schools that are contributing to mobility into the middle class by offering an affordable education to many low-income ($0 to $48,000) students. Excellence in outcomes: Schools that boast the highest outcomes across a set of measures, including: completion rates of students, labor market outcomes, and loan performance metrics. Financial value: Schools that provide positive value in purely financial terms by producing positive labor market outcomes at a low cost. An approach for crafting the “engines of opportunity” measure is to combine, with equal weight, the earnings and completion rates of low-income students (e.g., Pell Grant recipients) and the net price paid by low- and moderate-income students (those with a family income of $0 to $48,000). The measure 37 While the data analysis cited in this section of the paper used regression adjustments to account for differing missions of institutions and students served, it is important to consider many factors (e.g., the homogeneity in the types of programs or students served, or the underlying purpose of the metric) to determine whether to proceed with using regressions or other, similar methods such as grouping institutions together for comparison purposes based on similar characteristics. For example, in the analysis the Department performed when constructing our Gainful Employment regulations, we determined that while student characteristics may “play a role in postsecondary outcomes,” they did not meaningfully affect the debt-to-earnings composite metric. Program Integrity: Gainful Employment; Final Rule. 34 Fed. Reg. Parts 600 and 668 (October 31, 2014): 65039-65057. http://www.gpo.gov/fdsys/pkg/FR-2014-10-31/pdf/2014-25594.pdf. 24 would also be conditional on serving an above-average share of Pell recipients for the type of school (e.g., above 42 percent of undergraduates for two-year schools or 36 percent for four-year colleges).38 This metric would only assess the performance of those schools serving high shares of low-income students. As such, it would be important to recognize only the high-performers, rather than create any incentives for schools to turn away low-income students to influence the measure. Based on the “engines of opportunity” measure, public colleges make up nearly nine in 10 of the topperforming four-year institutions, and the remaining are private nonprofit schools. Among the topperforming two-year colleges, more than 60 percent are public, and about one quarter are private forprofit institutions. Many of the highest-performing schools on the metric are large public universities or two-year technical colleges. Examples of top-performing schools using the “engines of opportunity” metric are the following (see Table 7): Table 7. Four-Year Colleges Aurora University (Private Nonprofit) Georgia Regents University (Public) Hamline University (Private Nonprofit) Elizabeth City State University (Public) SUNY at Albany (Public) University of California—Irvine (Public) University of North Carolina at Charlotte (Public) Two-Year Colleges Central Maine Community College (Public) Georgia Perimeter College (Public) Hibbing Community College (Public) Lake Area Technical Institute (Public) Minneapolis Business College (Private For-Profit) Mount Aloysius College (Private Nonprofit) Pearl River Community College (Public) The “engines of opportunity” composite measure could be constructed in many other ways, as well. One option is to take into consideration the proportion of low-income students that a school serves. For instance, consider a metric that combines a measure of access (e.g. the share of Pell-eligible students enrolled) with measures of affordability (net price) and success (completion rate and median earnings). The weighting for each component could be determined based on the importance ascribed to each part (or measure) in the formula; for instance, each part could be weighted evenly, or the outcomes measures could be weighted more heavily to recognize the primacy of serving all students well, including low-income students. This approach would then recognize the schools that enroll large numbers of low-income students and serve all of their students well, in addition to noting those schools that fail to do so. The development of an “excellence in outcomes” composite metric is based on evenly weighted 38 We compute the z-score (which measures the number of standard deviations a school differs from the mean institution in its level, top- and bottom-coded at 4 and -4 to reduce the influence of extreme outliers) for three measures: 10-year earnings of former students from families with $30,000 or less in family income; the completion rate for students who received both Pell and loans, restricted this way to avoid poor reporting for Pell-only students; and the net price paid by students who came from families earnings less than $48,000. Only schools with a share of Pell student enrollment above the average for their predominant degree level are included. The completion measure includes transfer outcomes at two-year schools. 25 completion rates and median earnings.39 Many of the top performers on the “excellence in outcomes” metric are high-cost, selective four-year institutions (76 percent are private nonprofit colleges); and, among two-year colleges, a number of the top-performing schools in completion rates and earnings are those with a nursing or other specialized focus (see Table 8). They include: Table 8. Four-Year Colleges Harvey Mudd College (Private Nonprofit) Middlebury College (Private Nonprofit) University of Maryland—College Park (Public) University of Notre Dame (Private Nonprofit) University of Texas at Austin (Public) University of Wisconsin—Madison (Public) Villanova University (Private Nonprofit) Virginia Polytechnic Institute and State University (Public) Two-Year Colleges Bard College at Simon’s Rock (Private Nonprofit) Bismarck State College (Public) Coffeyville Community College (Public) Good Samaritan College of Nursing and Health Science (Private Nonprofit) New Mexico Military Institute (Public) St. Paul’s School of Nursing—Queens (Private For-Profit) Valley Forge Military College (Private Nonprofit) Vermont Technical College (Public) Among the bottom 10 percent of four-year schools, on the other hand, about a quarter are public colleges, while two-thirds are private nonprofit colleges. At the bottom 10 percent of two-year colleges, students typically pay a lower cost, but also may struggle to complete their educations and/or find a well-paying career. The Department also considered a composite metric addressing “financial value.” This metric can offer students a window into the potential payoffs over time that students can expect to receive based on the typical earnings from attending a particular institution relative to the costs. For example, one variation of the “financial value” metric measures the median earnings, minus the average net price of the college amortized over 30 years.40 By this measure, two-year public colleges (nearly seven in 10) are consistently the most successful in offering a great value to students. Among four-year colleges, nearly 30 percent of high “financial value” schools are public colleges, while 70 percent are private nonprofit schools (see Table 9). For-profits rarely qualify as top-performers on either the two-year or four-year measure. Some of the top performers include: Table 9. Four-Year Colleges Barnard College (Private Nonprofit) James Madison University (Public) 39 Two-Year Colleges Cabarrus College of Health Sciences (Private Nonprofit) Colorado Northwestern Community College (Public) The “excellence in outcomes” composite is calculated in a similar way but for all schools with data rather than restricting by the fraction of Pell students, taking the average of z-scores for earnings and completion rates adjusted for student characteristics. The completion measure includes transfer outcomes at two-year schools. 40 We measure financial value by subtracting the annual amortized net price (over 30 years, assuming 4.3 percent interest and multiplying total net price by the number of years for predominant degree) from the student characteristic adjusted earnings for a school. 26 Muhlenberg College (Private Nonprofit) San Jose State University (Public) University of Miami (Private Nonprofit) University of Michigan—Ann Arbor (Public) University of Oklahoma—Norman Campus (Public) Xavier University of Louisiana (Private Nonprofit) Kauai Community College (Public) Lake Washington Institute of Technology (Public) Los Medanos College (Public) Northern Virginia Community College (Public) Redlands Community College (Public) St. Vincent’s College (Private Nonprofit) Meanwhile, among the bottom 10 percent of four-year colleges on the “financial value” list, nearly 80 percent are private nonprofits. Almost 65 percent of two-year low-performers on this metric are private for-profits, although they comprise less than one-fifth of all two-year institutions. In general, public institutions appear to offer a consistent financial value, while there is more variation among private institutions’ outcomes compared to costs. Producing these composite measures may provide a simple way to help students and families understand the tradeoffs in performance at different institutions. However, in some ways, composite metrics are a starting point for the public to evaluate more individualized performance metrics, such as those that appear on the College Scorecard. Given the varying missions of institutions of higher education, the diverse values of policymakers and other users of the data, and the individual students’ varying interests, abilities, and the financial costs of attending a specific school, these three composite metrics can inform the development of additional metrics that can contribute in developing a more comprehensive view of how well colleges serve their students. Finally, details matter greatly in creating these composite metrics. Small tweaks can have a significant effect on the outcomes; adding or removing the elements included in the composite measure, or revising the weighting applied to each of the metrics, can create notable differences in the makeup of top- and bottom-performing institutions of higher education. With these findings, the Department aims to encourage states, researchers, and other experts in the field to continue pursuing the most accurate and thoughtful ways of combining measures of success to provide a baseline of understanding an institution’s performance. We hope to continue the conversation with the higher education community on the definitions of value and success for institutions of higher education. Learning and Other Outcomes Students’ outcomes tied to learning—what a student knows and can do—is an important way to understand the results and quality of any educational experience. However, no learning outcomes currently appear on the College Scorecard. There are few recognized and comprehensive measures of learning across higher education, and no data sources exist that provide consistent, measurable descriptions across all schools or disciplines of the extent to which students are learning, even where frameworks for measuring skills are being developed.41 Moreover, institutions appropriately hold the authority to define and create measures of learning for their own students. Further work is necessary to 41 Adelman, Cliff, Peter Ewell, Paul Gaston, and Carol Geary Schneider. “The Degree Qualifications Profile.” Lumina Foundation. October 2014: http://www.luminafoundation.org/files/resources/dqp.pdf. 27 define and publicize student learning outcomes and the many other benefits of higher education that are not captured by the measurements included now. Other Metrics for Consumers A guiding principle to designing the College Scorecard was determining the metrics most frequently requested by, and most valuable for, students to use. We sought to identify the elements that may help students, parents, guidance counselors, and other stakeholders find and select high-performing colleges using a broad array of relevant, understandable measures that address the diversity and utility of metrics for different types of students. Some metrics included on the technical site of the College Scorecard may be more useful to researchers and policymakers than to students and families. Moreover, some additional information not mentioned above is included on the consumer page. These elements include the programs that each school offers—a critical question many students working to set themselves up for a particular career, particularly adult students, may be asking of colleges. The rate at which first-year students choose to return the following year; and details about the particular mission of the school (for instance, Historically Black Colleges and Universities and minority-serving institutions, as well as schools with a particular religious affiliation) are identified. We also explored the research and conducted testing with students and parents to identify how the data we provided should be presented on the College Scorecard. The institutional characteristics that students deem most relevant to their college search processes—proximity to their homes, for instance—were added to the search features as a result of this valuable input. We look forward to continually updating the site as we learn more about which additional information and functionalities are most useful to students, families, guidance counselors, and other users. Ultimately, our goal is to provide the information, detailed or high-level, about the characteristics of the schools in a manner that will help students to narrow their searches down from more than 7,000 schools nationwide. Finally, we recognize that the information on each college’s individual page is of little utility if students do not have the context to understand how that school compares to other, similar schools, or to identify higher-performing alternatives. Our website is designed to help students survey the options within their own specifications, and weight the options available to them. Steps Forward Improving the Data The Department of Education conducted significant research, validation, and analysis on the data available both through the Integrated Postsecondary Education Data System (IPEDS) and the National Student Loan Data System (NSLDS). Throughout two years of exploration of the data, it became clear that many of the improvements already in the works will allow the Department to improve the quality of information it provides to students, families, and the public. Already, IPEDS is planning critical improvements to its completion rate data. As noted earlier, the current graduation rates in IPEDS assess the outcomes only of first-time, full-time students, a shrinking 28 proportion of the students enrolled in colleges and universities nationally and especially in some sectors of postsecondary education. Beginning in 2016, colleges will begin reporting completion rates for the other subsets of their students: first-time, part-time students; non-first-time, full-time students; and non-first-time, part-time students. Those data will give students and families critical information about the school’s comprehensive performance in helping students get to their commencement day. NSLDS data will begin publishing student outcomes by the program of study for the 2012 cohort. This information will expand the field of research and strengthen the relevance of information for the higher education community. For instance, researchers have found that earnings data can vary as much within a school, across the many majors and programs they offer, as between schools.42 Furthermore, moving forward, reporting for NSLDS completions is likely to improve. In particular, the only source of completion data for Pell Grant recipients by institution, which is available to the public, will benefit from a reporting improvement that began for the 2012 cohort. With these data, we will be able to calculate accurate completion and transfer rates for Pell Grant students. The Department will work with institutions to ensure they are providing accurate information for these students, in accordance with NSLDS reporting requirements. By putting the data in the hands of all members of the higher education community—from students to researchers—along with the documentation about what analyses are possible with these data, the field can now hold institutions accountable for better disclosure and reporting about their performance. As we update the data moving forward, we expect to continue working with schools to account for those improvements. We also hope these data—including their limitations—will help to ensure continued conversations with researchers, states, and colleges considering and grappling with similar questions. Improving the College Scorecard The launch of this new College Scorecard marks a transformed approach to improving the resources available to inform research and decisions regarding postsecondary institutions, and will rely on continued partnership and engagement with the public and the higher education community to continually test, and design measures and data available on the Scorecard. The participation of students, families, faculty, administrators, advocates, researchers, and others is paramount to ensuring that the tool optimizes effectiveness and impact for its users. One approach, for example, would allow the Department of Education to provide data disaggregated by student subgroups like income, age, and gender. With that information, students would be able to enter information about themselves – and see the results for students with a similar profile at any school. This information is provided on the technical site of the College Scorecard so that others can begin experimenting with the best ways to construct a “Students Like Me” feature. As we develop this and other potential upgrades to the College Scorecard, we hope to work with counseling and college pipeline programs, policy and data experts, and researchers in consumer choice and other areas to design the most urgent and important improvements. 42 Carnevale, Anthony, Ban Cheah, and Andrew Hanson. “The Economic Value of College Majors.” Georgetown Center on Education and the Workforce, 2015: https://cew.georgetown.edu/cew-reports/valueofcollegemajors/. 29 Changing the Public Discourse Most importantly, we want to make sure college is more accessible, affordable, and valuable for students. With the College Scorecard resources, all members of the higher education community and the general public can become more informed and involved in ensuring that college opportunities improve for all. The College Scorecard can serve as a tool to inform ongoing and new state efforts to implement accountability systems for higher education, in parallel with federal efforts to ensure colleges and universities spend taxpayer dollars intentionally, on schools serving students well, members from all corners of the public can participate in holding schools accountable to all populations they serve. The College Scorecard data also provide schools with stronger tools to benchmark against peer institutions, address their weaknesses, and adopt better practices. High-quality, comparable data across all institutions in the U.S. can provide an unprecedented level of information to help schools measure student success and grow. To that end, we hope accreditors and institutions will recognize the importance of these metrics and focus on providing a good value for students and to help more students graduate and find well-paying jobs. In today’s 21st-century economy, a college education is no longer just a privilege for some, but rather a prerequisite for all. College remains the greatest driver of socioeconomic mobility in America. With new information and resources, we—students, parents, advisors, school leaders, teachers, researchers, accreditors, leaders and faculty of institutions, policy advisors, and the public in general—become partners in exchanging better knowledge and practices about how to promote the highest standards and most thoughtful practices in the American higher education system.