View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

Student Loan Borrowing and Repayment Trends, 2015
April 16, 2015

Andrew Haughwout, Donghoon Lee, Joelle Scally, Wilbert van der Klaauw
The views presented here are those of the authors and do not necessarily reflect those of the
Federal Reserve Bank of New York, or the Federal Reserve System.

Some higher education background
 College remains a worthwhile investment for most of those who
pursue it
 Median annual earnings were $23,000 higher for bachelor’s
degree holders compared to high school graduates in 2014
 Unemployment rates for bachelor’s degree holders much lower
than high school graduates (6% vs 3.5% in 2014)

 Student loans are an important tool for financing college for
many students
 Speech by William C. Dudley, March 4 on student loans:
 “ . . .[T]here is much uncertainty and heterogeneity in outcomes,
with net returns to these human capital investments being
negative for some. Understanding causes and consequences of
this heterogeneity is important.”
 “How we finance post-secondary education has significant effects
on a variety of critical economic outcomes.”
2

Researching student borrowing
 We try to better understand student loans for two reasons:
 Student debt is large, and appears to have significant
effects on macroeconomic outcomes (household formation,
homeownership, consumption)
 Good information on borrowing and repayment is necessary
to design best policies for financing this most important kind
of investment

3

Student loans defy business cycle
Billions of Dollars
1200

Billions of Dollars
1200

Non-mortgage balances
HELOC

Auto Loan

Student Loan

Credit Card

1100

1100

1000

1000

900

900

800

800

700

700

600

600

500

500

400

400

300

300

200

200

100

100

0

0

Source: New York Fed Consumer Credit Panel / Equifax

4

Student loans increased as other debts declined
 During and after the Great Recession, households reduced
their other debts, but student loan balances continued to
increase
 Because the majority of student loans are federal, tight bank
lending standards did not affect student loans
 The increase was a result of both increasing numbers of
borrowers and increasing average balances per borrower
 Between 2004 and 2014, there was an 89% increase in the
number of borrowers and a 77% increase in the average balance
size

 College enrollment grew by 20% between 2005 and 2010 –
faster than any period since the 1970’s – and has declined
slightly since

5

Total student loan balances by age group
under 30

Billions of Dollars
1,200

30-39

40-49

50-59

60+
5%
12%

1,000

18%

800

600

400

33%

2%
9%
14%

200

32%

33%
42%

0
2004

2005

2006

2007

Source: New York Fed Consumer Credit Panel / Equifax

2008

2009

2010

2011

2012

2013

2014
6

Increases among borrowers of all ages
 Student loan balances grew for borrowers of all ages between
2004 and 2014:
 The number of borrowers over 40 increased at twice the pace of
the number of borrowers under 40
▫ Especially fast growth in balances held by for borrowers age 60+, up
nearly nine-fold

 As a consequence, share of balances held by borrowers age 40+
increased from 25% to 35%

 2/3 of student loan balances are held by borrowers not in their
20s

7

Distribution of borrowers by 2014Q4 balance
1.0%
0.8%
2.4%
2.9%

Balance in 2014Q4:

7.2%

20.8%

≤$5k
>$5k and ≤$10k
>$10k and ≤$25k
>$25k and ≤$50k
>$50k and ≤$75k
>$75k and ≤$100k
>$100k and ≤$150k
>$150k and ≤$200k
>$200k

18.5%

18.0%

28.5%

Source: FRBNY Consumer Credit Panel/Equifax

43.3 million borrowers
Mean balance: $26,700
Median balance: $14,400
8

Significant heterogeneity in amounts owed
 Most borrowers have a current outstanding balance below $25k
 About 40% owe less than $10K

 Mean outstanding balance is $26k; median balance is $15k
 Significant numbers of borrowers with large balances (over $25K)

 Borrowers in their 30’s and 40’s have the highest mean and
median balances, at about $31k and $17k respectively

9

Who is borrowing now?
Number of originating borrowers by age*

Millions
14

12

10

8

6

4

2

0
2004

2005

2006

2007

2008

less than 25

2009
25-29

2010
30-39

2011

2012

2013

2014

40+
10

Source: FRBNY Consumer Credit Panel/Equifax; * excludes small number of borrowers with missing age

Number of active borrowers peaked in 2010
 The number of active student loan borrowers peaked in 2010,
at about 12 million; now down to about 9 million
 Half of active borrowers are under age 25
 After the recession, there was a slight increase in the relative
share of new borrowers in their thirties, and a decline in the share
of 18-29 year olds

 The Department of Education estimates suggest that there was
a modest drop in enrollment between 2010 and 2013, of about
500,000 students
 More importantly, a smaller share of enrolled students has
taken out loans since 2010
 Although the number of active borrowers is declining, the
aggregate outstanding balance of borrowers overall has
continued to grow as repayment rates are very low
11

Increase, then decline in borrowers from lower
and middle income areas
Number of originating borrowers by ZIP code income*
Millions
14

12

10

8

6

4

2

0
2004

2005

2006

2007
2008
2009
2010
up to 40k
40k-60k
60k-80k

2011
80k+

2012

2013

2014

12
Source: FRBNY Consumer Credit Panel/Equifax; Internal Revenue Service; * excludes small number of borrowers with missing ZIP code incomes

Larger rise in level of lower income borrowers
following recession
Active borrowers by ZIP code income

Index, 2004=100

1.5

1.4

1.3

1.2

1.1

1

0.9
2004

2005

2006

2007
up to 40k

2008
40k-60k

2009

2010
60k-80k

2011

2012

2013

2014

80k-100k
13

Source: FRBNY Consumer Credit Panel/Equifax; Internal Revenue Service; * excludes small number of borrowers with missing ZIP code incomes

Increase, then decline in borrowers from
lower and middle income areas
 We don’t have borrower income, so we assign borrowers to
groups based on the average income in the ZIP code they lived
in when they originated their first student loan
 Income data are from IRS and refer to 2010
 Much of the change in borrowing has come from an increase,
then decrease, in the number of active borrowers from lower
income areas following the recession
 More borrowers from lower and middle-income areas is
consistent with the intended purpose of student loan program:
to facilitate upward income mobility

14

Wrapping up, part 1
 Student debt continues to increase, especially for older
borrowers
 Increase reflects new borrowers, higher balances and slow
repayment
 There is significant heterogeneity in amounts owed
 Highest balances are owed by borrowers in their 30s
 The number of active borrowers peaked in 2010
 Significant decline since then
 Increase and subsequent decline driven by borrowers from
relatively lower-income areas

15

Student loan default & repayment

for internal use only

Distribution of payment history
Snapshot of Borrowers in 2014:Q4

11%

Percent of Borrowers
6%

29%
always current, balance decreasing
always current, balance increasing
current with previous blemish
now delinquent

20%

now in default

34%

Source: New York Fed Consumer Credit Panel / Equifax

17

Understanding repayment outcomes
 We will compare several groups to better understand the
determinants of repayment outcomes:
 Those who left school before, during, and after the Great
Recession
 Younger borrowers and older borrowers
 Borrowers from higher income and lower income areas

18

Defaults and default rate
Annual rate

Thousands per year

4.0%

1,400

3.5%

1,200

3.0%
1,000
2.5%
800
2.0%
600
1.5%
400
1.0%
200

0.5%

0

0.0%
2003-4

2005-6

2007-8
defaulted borrowers

2009-10

2011-12

2013-14

rate of default

19
Source: FRBNY Consumer Credit Panel/Equifax

Defaults and default rate
 The number of borrowers who default each year (defined as at
least 9 months past due) increased from about half a million ten
years ago to 1.2 million annually in 2011 and 2012
 The number of borrowers who default has declined a bit in the
last two years
 The rapid increase in the defaults is partly due to the increase
in the number of borrowers, but even after accounting for the
increase in the borrowers, the rate of default increased over
time from 2.4% in 2004 to 3.6% in 2012
 The default rate declined somewhat after 2012 to 3.2%
20

Default rate by school leaving cohort
Share ever defaulted by years after leaving school
2009 Cohort
26%
5 years later

25%

2007 Cohort
24%
7 years later
2005 Cohort
25%
9 years later

20%

15%

10%

5%

0%
1 year later

2 years

3 years

4 years
5 years
6 years
(after entering repayment)

7 years

8 years

9 years

Source: New York Fed Consumer Credit Panel / Equifax

21

Cohort default rates deteriorate


Default rates for the 2005, 2007, and 2009 school-leaving cohorts
have all now reached roughly 25%.
− Cohort default rates continue to increase even beyond the
third year of loan repayment



There is a pronounced worsening of the cohort default rate
schedules over time
− Default rates are higher for later cohorts at virtually all
durations

22

5-year cohort default and delinquency rates
by ZIP income
2005 school
leaving cohort

2007 school
leaving cohort

2009 school
leaving cohort

60%

60%

60%

50%

50%

50%

40%

40%

40%

30%

30%

30%

20%

20%

20%

10%

10%

10%

0%

0%
less than 40k-60k 60k-80k
40k

80k+

0%
less than 40k-60k 60k-80k
40k
default

80k+

less than 40k-60k 60k-80k
40k

80k+

120+ dpd

23
Source: FRBNY Consumer Credit Panel/Equifax

Cohort default, delinquency rates by income


Default and delinquency rates higher in lower-income areas



Default and delinquency rates for borrowers from higher income
areas (with mean income over $60K) have remained remarkably
stable across the three cohorts

24

5-yr Cohort default and delinquency rates by age
2005 school
leaving cohort

2007 school
leaving cohort

2009 school
leaving cohort

60%

60%

60%

50%

50%

50%

40%

40%

40%

30%

30%

30%

20%

20%

20%

10%

10%

10%

0%

0%

0%

less 25-29 30-39
than 25

40+

less 25-29 30-39
than 25
default

40+

less 25-29 30-39
than 25

40+

120+ dpd

25
Source: FRBNY Consumer Credit Panel/Equifax

Cohort default, delinquency rates
by school-leaving age
 The 5-year-default rate increased for all age groups, and it is
somewhat elevated for those who left school between age 3039
 About half of the 2009 cohort borrowers in this age range had
gone through defaults or severe delinquency in the subsequent
5 periods

26

Student loan repayment behavior
Student Loan Repayment Status in 2014

delinquent or in
default
17%
current, balance
increasing
33%

current and paying
down
37%
current, same
balance
13%

Source: New York Fed Consumer Credit Panel / Equifax

27

Nearly half of borrowers are not yet repaying
 17% of the borrowers are in default or in delinquency
 Only 37% of borrowers are current on their loan and actively
paying down
 46% of borrowers are current on their loans but are not in
repayment

28

5-year cohort repayment difficulties,
by ZIP income
2005 school
leaving cohort

2007 school
leaving cohort

2009 school
leaving cohort

80%

80%

80%

70%

70%

70%

60%

60%

60%

50%

50%

50%

40%

40%

40%

30%

30%

30%

20%

20%

20%

10%

10%

10%

0%

0%
less than 40k-60k 60k-80k
40k

80k+

less than 40k-60k 60k-80k
40k
default

120+ dpd

80k+

0%
less than 40k-60k 60k-80k
40k

80k+

debt increase

29
Source: FRBNY Consumer Credit Panel/Equifax

Proportion of borrowers with payment problems
by cohort and income
 While default and delinquency rates were stable for higher
income area borrowers across the three cohorts, more recent
cohorts from higher income areas had a larger proportion who
had not paid down
 About half of the 2009 cohort are having payment difficulties on
their loans (having either defaulted, gone delinquent or made
no progress paying off their balance)
 Repayment difficulties are particularly prevalent in
borrowers from lower-income areas, where about 2/3 of
borrowers have had some sort of repayment difficulty
 In contrast, just over 1/3 of the borrowers from the highest
income areas have had repayment difficulties

30

2009 cohort repayment rates by income
Proportion of balance remaining by ZIP income
100%
95%
90%
85%
80%
75%
70%
65%
60%
1 year

2 years

3 years

4 years

5 years

(years after leaving school)

less than 40k

40k-60k

60k-80k

80k+
31

Source: FRBNY Consumer Credit Panel/Equifax

Little repayment progress by borrowers from
lower-income areas
 We look at the net progress that borrowers from each income
area have made in paying down their student loan balances
 The borrowers from the lowest income areas have made
practically no progress on paying down their loans
 The aggregate balance, five years after leaving school, is
still at 97% of where it was when they left school
 This is a sharp contrast with the borrowers from higherincome areas, who have paid down nearly 30%

32

Wrapping up
 These results are further evidence of the important
heterogeneity we see in the outcomes of higher education
investments financed with student debt
 Borrowers who left school in the Great Recession had particular
difficulty with their student loan repayment, with many defaulting,
becoming seriously delinquent, or not being able to reduce their
balance
 Borrowers from lower and middle-income areas as well as
borrowers who originated loans in their 30s are also at greater risk
of default and delinquency

 The low overall repayment rate helps explain the steady growth
in aggregate student debt, now at nearly 1.2 trillion dollars

33