College on Credit: a multi-level analysis of student loan default

Size: px

Start display at page:

Download "College on Credit: a multi-level analysis of student loan default"

Cecilia Bishop
7 years ago
Views:

1 College on Credit: a multi-level analysis of student loan default Nick Hillman Assistant Professor University of Utah Educational Leadership & Policy The material is based upon work supported by the Association for Institutional Research, the National Science Foundation, the National Center for Education Statistics, and the National Postsecondary Education cooperative under Association for Institutional Research grant number RG Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the Association for Institutional research, the National Science Foundation, the National Center for Education Statistics, or the National Postsecondary Education Cooperative.

2 Overview Brief summary of loan trends and state of affairs of student loan default research Conceptualization of repayment success Data and analytical strategies Key findings relative to theory/policy Discussion

3 Research questions To what extent are: students socioeconomic, academic, or demographic characteristics associated with defaulting on student loans? institutional characteristics related to default rates?

4 Some qualifications What this study does Examines borrowers odds of defaulting relative to repaying on time Partitions the variance between student and institution levels Utilizes recent national survey data (BPS:09) What this study doesn t do Estimate a causal relationship between these factors Compare other repayment options (i.e. deferment, forbearance, etc.) Examine private loan repayment

5 How much indebtedness occurs? $25,250 (1) 66% (2) 8.7 million (3) Average debt of bachelors degree recipient $950 < debt > $55k Percent of undergraduate bachelors grads who borrow Number of borrowers (fed sub+unsub undergrads)

6 Share of undergraduates* borrowing federal loans *full-time, full-year only (1) 100.0% 92.8% % 63.0% 57.1% 50.0% 38.5% 44.3% 48.6% 31.3% 30.4% 25.0% 21.5% 11.7% 0.0% Proprietary two-year and above Private fouryear (doctoral) All institutions Public fouryear (doctoral) Public twoyear

7 Share of outstanding debt by source (NY Federal Reserve data)

8 2003 $ Billions Student Loan, $0.24 Other, $0.48 Credit Card, $0.69 Auto Loan, $0.64 HE Revolving, $0.24 Mortgage, $4.94

9 2004 $ Billions Student Loan, $0.26 Credit Card, $0.70 Other, $0.45 Auto Loan, $0.72 HE Revolving, $0.33 Mortgage, $5.84

10 2005 $ Billions Student Loan, $0.36 Credit Card, $0.71 Other, $0.39 Auto Loan, $0.73 HE Revolving, $0.50 Mortgage, $6.51

11 2006 $ Billions Student Loan, $0.43 Credit Card, $0.72 Other, $0.42 Auto Loan, $0.79 HE Revolving, $0.58 Mortgage, $7.44

12 Student Loan, $ $ Billions Credit Card, $0.76 Other, $0.40 Auto Loan, $0.79 HE Revolving, $0.61 Mortgage, $8.42

13 Student Loan, $ $ Billions Credit Card, $0.84 Other, $0.42 Auto Loan, $0.81 HE Revolving, $0.66 Mortgage, $9.23

14 2009 $ Billions Student Loan, $0.66 Credit Card, $0.84 Other, $0.41 Auto Loan, $0.77 HE Revolving, $0.71 Mortgage, $9.14

15 2010 $ Billions Student Loan, $0.76 Credit Card, $0.76 Other, $0.36 Auto Loan, $0.70 HE Revolving, $0.70 Mortgage, $8.83

16 2011 $ Billions Student Loan, $0.84 Other, $0.33 Credit Card, $0.70 Auto Loan, $0.71 HE Revolving, $0.64 Mortgage, $8.54

17 2012 $ Billions Student Loan, $0.90 Other, $0.32 Credit Card, $0.68 Auto Loan, $0.74 HE Revolving, $0.61 Mortgage, $8.19

18 Context on student loan default Policy HEA Title IV eligibility Gainful employment Research Do institutions matter? Or is default a pre-existing condition? (Monteverde, 2000) Practice Consequences of defaulting: Borrower: aid eligibility, wage garnishment, credit score, bankruptcy Institution: CDR, default management, 3-year window (30% CDR)

19 Total number in default Trends in loan default (2-year CDR) 350, , , , , , , , , , , , , , , , , ,000 50, Cohort year

20 Conceptual framework Student-level (level 1) Demographics: - Age - Gender - Race/ethnicity - Socio-economic status: - First-generation student - Family income level - Has dependents Academic experience: - Major choice - College GPA - Received Pell Grant - Cumulative federal loan - Took out private loans - Transfer - Degree completion Post-collegiate employment: - Employment status Primary loan repayment outcomes of interest: - In repayment - In default Alternative loan repayment outcomes: - Repaid loans - Emergency protection - Not yet in repayment Institution-level (level 2) Control: - Public - Private - Proprietary Level: - Two-year - Four-year

21 Data and analysis Data: BPS: Nationally-representative sample In repayment vs. In default (n=5,959) Panel weights Analytical technique: Multi-level model (ICC =.32) Binary outcome (1 = default, 0 = in repayment)

22 Key finding #1: Cumulative debt burden is not a strong predictor of default $20,050 $19,020 $18,150 Odds ratio Cumulative federal loans 0.919*** Cumulative federal loans (sq) 1.001** $8,270 $8,250 Not yet repaying In repayment Deferred Paid off In default

23 Key finding #2: Degree completion and employment are strong predictors of default 100% 90% 80% Share of borrowers in default by degree attainment Odds ratio No degree, not enrolled 2.314*** 100% 90% 80% Share of borrowers in default by employment status 70% 60% 65% Unemployed 1.705*** 70% 60% 50% 50% 40% 30% 20% 10% 0% 27% Degree recipient No degree, not enrolled 40% 30% 20% 10% 0% 13% Employed 30% Unemployed

24 Key finding #3: Student socio-economic characteristics are moderate predictors of default $60,000 Average family income level 160% Odds of default by race (compared to white) 145% 142% 138% $50,000 $48, % 120% $40, % $30,000 80% 70% $20,000 $21,409 60% 40% $10,000 20% $0 In repayment In default 0% African American Other Hispanic Asian/Pacific Islander Odds ratio Family income level (thousands).990***

25 Key finding #4: Net of student characteristics, institutional sector strongly predicts default Share of defaults, by sector Public fouryear, 11% Odds ratio of default, by sector Comparison group = public four-year sector 316% 259% For-profit two-year, 52% For-profit four-year, 9% Public twoyear, 19% Private fouryear, 8% Private twoyear, 2% 110% Private fouryear 132% Public twoyear 214% For-profit four-year Private twoyear For-profit two-year

26 Summary & Implications Not all default risk can be attributed to pre-existing conditions In fact, it may be a preventable illness Recall the grad rate and unemployment variables Federal policy implications Student-level Improve information regarding repayment options Consider alternative sanctions for defaulted loans Institution-level Program integrity: strengthen career college sanctions/access to Title IV aid Cohort Default Rate threshold could even lower for many forprofits (i.e. are we ok when 30% of borrowers default?)

27 Thank you! Nick Hillman Assistant Professor University of Utah Educational Leadership & Policy

The 5 W s of Student Loan Defaulters: Using Analytics to Form Your Default Prevention Strategy. Strategic Business Director

The 5 W s of Student Loan Defaulters: Using Analytics to Form Your Default Prevention Strategy. Strategic Business Director Using Analytics to Form Your Default Prevention Strategy Presented by: Ron Lee OASFAA Conference 2015 What are the 5 W s? Who is defaulting? Why are they defaulting? What tactics are you currently using