Macro Roundup

Frederick M. Hess, "PISA results offer more bad news for US schools," American Enterprise Institute, December 5, 2019, https://www.aei.org/education/pisa-results-offer-more-bad-news-for-us-schools/

PISA results offer more bad news for US schools

On Tuesday, the Program for International Student Assessment (PISA) reported its new results for how 15-year-olds are faring across about 80 nations. In what’s become a familiar theme, US math and reading scores were flat since the last test in 2015, with the US ranking 30th in the world in math and 8th in reading. As a New York Times headline stated, “‘It Just Isn’t Working’: PISA Test Scores Cast Doubt on U.S. Education Efforts.” Indeed, the results showed that about one-fifth of American 15-year-olds haven’t yet mastered the reading skills expected of a 10-year-old.

These results are disappointing, but not surprising. As we noted last spring, US student performance on PISA has been stagnant — for two decades, since 2000. While scholars have raised legitimate questions about how much faith we should put in PISA, the results mirror what we see on the gold standard domestic assessment, the National Assessment of Educational Progress (NAEP). Indeed, the new NAEP results, released just last month, reported a continued, decade-long stagnation in US student achievement.

The remarkable thing, as we observed in the spring, is that these grim results have been accompanied by close to twenty years of impressive gains on annual state tests — the kinds of outcomes eagerly touted by state leaders, and required first by the federal No Child Left Behind Act and now by its successor, the Every Student Succeeds Act. The ongoing disparity between measures such as PISA and NAEP, on one hand, and state tests, on the other, should leave us leery when it comes to these ubiquitous assessments, which get used for everything from accountability systems to school improvement plans to research into which reforms “work.”

As we wrote last spring, “The politicos and state education officials claiming credit for these gains are the same ones who choose state tests, define what qualifies as ‘proficient,’ and monitor graduation rates to guard against funny business. The results are tied into state accountability systems, where lousy results can produce practical and political headaches. Thus, policymakers have both the means and the incentive to inflate the numbers any way they can.” And there’s good cause to believe that such shenanigans are all too common: Researchers have reported that gains on state assessments tend to outpace gains on the far more reliable NAEP by an astounding 50 to 100 percent.

The PISA and NAEP results are disconcerting. They should prompt plenty of hard questions, including about how much faith we should have in state tests which too often provide a fundamentally misleading picture of how American students are faring.

“…US student performance on PISA has been stagnant — for two decades, since 2000….these grim results have been accompanied by close to twenty years of impressive gains on annual state tests — the kinds of outcomes eagerly touted by state leaders…. Researchers have reported that gains on state assessments tend to outpace gains on the far more reliable NAEP by an astounding 50 to 100 percent….”

Ernie's updated his demographically-adjusted LFPR on the new jobs numbers. Age-adjusted EPOP and LFPR both rose a few bps. Adjusted EPOP remains above, and is widening its lead over its 1999-2000 average. Adjusted LFPR is within 20bp of its 1999-2000 average.

CEP research looks at the disconnect between state testing and NAEP results and finds that while NAEP generally aligns with state trends, they move in the same direction the magnitudes of the state trends are significantly overstated versus the NAEP results. For example the median average annual gain in the percentage proficient on state tests was larger then on NAEP - for grade 4 math, state test showed a median average annual gain of 1.3, where NAEP showed a.5. For 8th grade math those numbers were 1.8, and 1 respectively. (Table 6).

They speculate that several factors might be causing the spread, 1) state test are more aligned with state educational standards. 2) score inflation due to narrow test preparation. 3) Motivation, neither students or parents get individual NAEP results so the test are lower stakes for teachers and students alike versus state exams.4) Y/Y changes in test difficultly.

"...Since 2005, test scores have increased in most states with sufficient data. States with test score gains between 2005 and 2009 far outnumbered those with declines on two different assessments (state tests and NAEP) and two different indicators (percentages scoring proficient/basic and mean scores)..... Within the same state, trends on NAEP usually moved in the same direction as trends on state tests.States with positive trends between 2005 and 2009 on their own tests tended to show positive trends on NAEP. This pattern was apparent in percentages proficient/basic and, to an even greater extent, in mean scores.....Gains on state tests tended to be larger in size than gains on NAEP.This was not always the case, however. In a limited number of states, gains on NAEP were larger than gains on state tests, especially in grade 4 reading...."

Naomi Chudowsky and Victor Chudowsky, "Rising Scores on State Tests and NAEP," Center of Education Policy, September 2010, https://www.cep-dc.org/displayDocument.cfm

Higher income families are increasingly taking on debt to pay for undergraduate education:

"...Some of the biggest and most consistent changes in student borrowing patterns over the past two decades have occurred among students from higher-income families. In the 2015-16 academic year, students from higher-income families were just as likely to take on debt for an initial year of an undergraduate education as were students from low-income families.Undergraduates from higher-income families take on the largest debts, while those from the lowest-income families take on the smallest debts. The share of low-income students earning bachelor’s degrees with debt has been remarkably consistent (75 percent) for 20 years, while the share of high-income students borrowing has doubled to 60 percent...."

Jason Delisle, "The surprising role of high-income families in student debt trends: Examining undergraduate borrowing by income, 1995-96 to 2015-16," American Enterprise Institute, December 4, 2019, https://www.aei.org/research-products/report/the-surprising-role-of-high-income-families-in-student-debt-trends-examining-undergraduate-borrowing-by-income-1995-96-to-2015-16/

The surprising role of high-income families in student debt trends: Examining undergraduate borrowing by income, 1995-96 to 2015-16

Introduction

Observers from across the ideological spectrum argue that the US is in the midst of a student debt crisis. This view is largely motivated by the fact that student debt now totals $1.5 trillion after rising rapidly in the past decade, particularly during the last economic recession.1 In 2003, outstanding debt was just $311 billion after adjusting for inflation.2

These trends have prompted several Democratic presidential candidates to propose that the federal government forgive most or all outstanding student debt (the vast majority of which was issued through federal programs).3 They argue that the debt is unaffordable for many and is the result of severe inequities in our higher education system. Concern about inequitable student debt extends beyond public policy. A growing number of private employers now repay a share of their employees’ student loans in part out of concern that students from disadvantaged backgrounds are the ones most likely burdened by debt.4

In light of these loan forgiveness proposals, employer-sponsored benefits, and the broader concerns about rising student debt, understanding who takes on student loans is important. A clearer picture of who borrows will help identify the beneficiaries of broad loan forgiveness proposals and the more limited employer repayment benefits, or any other policy aimed at reducing student debt and repayment obligations. It can also help gauge the extent to which student debt burdens reflect inequities in the US higher education system.

Prior analyses on this topic focus on the demographics of borrowers who currently hold the $1.5 trillion in outstanding debt. For example, researchers at the Urban Institute show that higher-income households hold a disproportionately large share of all student debt.5 Using different data, the Federal Reserve Bank of New York reports similar findings but also concludes that relative to household earnings, debt is higher among lower-income households.6

This report takes a different approach to understanding who holds student debt. It examines borrowers’ characteristics when the loans were originated, whereas the studies mentioned above capture borrowers at any point during repayment. Both perspectives are useful, but the former is less common in policy discussions. Moreover, previous research on income characteristics at loan origination appears at odds with data on borrowers in repayment. Some analyses focused on demographics at loan origination conclude that low- and middle-income students are “more than twice as likely as other students to have student loans” or that “high student debt goes hand in hand with low income.”7

To help fill the void in the research, this analysis focuses on borrowing patterns among students who enrolled in an institution of higher education in the 1995-96 and 2015-16 academic years. (It includes data points for the intervening years in an appendix.) The analysis is limited to two main statistics for undergraduates by family income: the share of students who took on debt and the amount they borrowed. These statistics are reported for two distinct groups of students at different points in their enrollment: first-year undergraduates and students who earned a bachelor’s degree in the years covered in this analysis 8 Data for the analysis come from the US Department of Education’s National Postsecondary Student Aid Study (NPSAS), which provides a representative sample of the undergraduate population for the 1995-96, 1999-2000, 2003-04, 2007-08, 2011-12, and 2015-16 academic years.9

While debt from graduate and professional students makes up a large share of all outstanding student debt (approximately 40 percent), this analysis excludes these students.10 The income information included in the NPSAS is difficult to interpret for these students because it reflects the student’s own income (and income from a spouse) while enrolled as a graduate student. The undergraduate data include parental income for dependent students or the student’s own income if he or she is an independent. This analysis includes both dependent and independent undergraduate students despite this difference because the overall findings are similar when only dependent students are included in the analysis.

Note that independent students, who tend to have low incomes and make up about half of undergraduates, skews the income distribution of the
undergraduate population in this analysis (shown in Appendix C). Debt figures for this analysis include the amount of principal borrowed for all types of student debt (federal student and parent loans, private, state, etc.). They do not show unpaid interest. The 1995-96 and 1999-2000 data sets include loans that a student received from a family member in aggregated borrowing figures, but later data sets exclude these loans. To be consistent, this analysis excludes family loans from earlier data sets.11 Borrowers are grouped into five income categories that approximate the US household income quintiles for the last year in the analysis, 2015-16.12 All figures hereafter are in 2015 dollars unless otherwise noted.13

Notes

Federal Reserve Bank of New York, Research and Statistics Group, Quarterly Report on Household Debt and Credit, August 2019, https://www.newyorkfed.org/medialibrary/interactives/householdcredit/data/pdf/hhdc_2019q2.pdf.
This was adjusted for inflation using the personal consumption expenditures price index. Federal Reserve Bank of New York, “Quarterly Report on Household Debt and Credit,” May 2019, https://www.newyorkfed.org/medialibrary/interactives/
householdcredit/data/xls/hhd_c_report_2019q1.xlsx.

Jeff Stein, “Sanders Proposes Canceling Entire $1.6 Trillion in U.S. Student Loan Debt, Escalating Democratic Policy Battle,” Washington Post, https://www.washingtonpost.com/business/economy/sanders-to-propose-canceling-entire-16-trillion-in-us-student-loan-debt-escalating-democratic-policy-battle/2019/06/23/1eed053a-9561-11e9-aadb-74e6b2b46f6a_story.html; Elizabeth Warren, “I’m Calling for Something Truly Transformational: Universal Free Public College and Cancellation of Student Loan Debt,” Medium, April 22, 2019, https://medium.com/@teamwarren/im-calling-for-something-truly-transformational-universal-free-public-college-and-cancellation-of-a246cd0f910f; and US Department of Education, Office of Federal Student Aid, “Federal Student Aid Portfolio Summary,” https://studentaid.ed.gov/sa/sites/default/files/fsawg/datacenter/library/Portfolio Summary.xls.

Anne Tergesen, “Employers Try a New Perk: Matching Student Loan Payments with 401(k) Contributions,” Wall Street Journal, October 10, 2019, https://www.wsj.com/articles/employers-try-a-new-perk-matching-student-loan-payments-with-401-k-contributions-11570708801.

Sandy Baum and Alexandra Tilsley, “Which Households Hold the Most Student Debt?,” Urban Institute, May 2, 2019, https://www.urban.org/urban-wire/which-households-hold-most-student-debt.

Andrew F. Haughwout et al., “Who Borrows for College—and Who Repays?,” Federal Reserve Bank of New York, October 9, 2019, https://libertystreeteconomics.newyorkfed.org/2019/10/who-borrows-for-collegeand-who-repays.html.

Brandon Debot and David Reich, “House Budget Committee Plan Cuts Pell Grants Deeply, Reducing Access to Higher Education,” Center on Budget and Policy Priorities, October 21, 2015, https://www.cbpp.org/research/house-budget-committee-plan-cuts-pell-grants-deeply-reducing-access-to-higher-education; Institute for College and Success, “Pell Grants Help Keep College Affordable for Millions of Americans,” September 18, 2019, https://ticas.org/wp-content/uploads/2018/07/Overall-Pell-one-pager.pdf; Jon Marcus, “Billions in Pell Grants Go to Students Who Aren’t Graduating, New Data Shows,”

Public Broadcasting Service, October 30, 2018, https://www.pbs.org/newshour/education/billions-in-pell-grants-go-to-students-who-arent-graduating-new-data-shows; and Rebecca Safier, “Study: Student Loans Weight the Heaviest on Black and Hispanic Students,” Student Loan Hero, September 17, 2018, https://studentloanhero.com/featured/study-student-loans-weigh-heaviest-black-hispanic/.

Foreign students are excluded from this analysis. Only US citizens and resident aliens are included.

National Center for Education Statistics, “National Postsecondary Student Aid Study,” https://nces.ed.gov/surveys/npsas/.

The Congressional Budget Office estimates that 42 percent of annual loan disbursals measured in dollars are for students attending graduate and professional school. Federal budget tables do not include figures for the share of outstanding debt used to finance graduate and professional educations. An Urban Institute study finds that 48 percent of outstanding student debt is held by households that include someone with a master’s or professional degree. This debt, however, includes loans that financed an undergraduate education. For more information, see Congressional Budget Office, “Student Loan Programs—CBO’s May 2019 Baseline,” May 2019, https://www.cbo.gov/system/files/2019-05/51310-2019-05-studentloan.pdf; and Sandy Baum, Victoria Lee, and Alexandra Tilsley, “Which Households Hold the Most Student Debt?,” Urban Institute, May 20, 2019, https://www.urban.org/urban-wire/which-households-hold-most-student-debt.

About 15 percent of undergraduates reported these family loans at an average of about $3,700 in 1995 dollars. See variables FAMLOAN and FAMOWE in the 1995-96 NPSAS at National Center for Education Statistics PowerStats, “National Postsecondary Student Aid Study: 1996 Undergraduates,” November 2009, https://nces.ed.gov/datalab/powerstats/pdf/npsas1996ug_subject.pdf.

The income quintiles are for 2014 but are adjusted for inflation and converted to 2015 dollars using the personal consumption expenditures price index. This means that the income groups do not match household income quintiles for years before 2014. Instead, income groups are constant throughout the analysis after adjusting for inflation. The income information for the 2015-16 NPSAS reflects earnings in 2014 as it generally relies on the respondents’ most recently filed federal income tax return for this information. The income groups therefore reflect the household income quintiles in 2014 but are shown in 2015 dollars throughout this analysis. See National Center for Education Statistics PowerStats, “National Postsecondary Student Aid Study: 2016 Undergraduates,” January 2018, https://nces.ed.gov/datalab/powerstats/pdf/npsas2016ug_subject.pdf.

Figures are adjusted for inflation using the personal consumption expenditures price index.