Research

Student Loan Payments

Evidence from 4 Million Families

July 1, 2019

Findings

Executive Summary

Student loan debt is the fastest growing household debt category, having more than doubled over the last ten years to $1.5 trillion in 2018, second only to mortgage debt, and affecting 45 million borrowers. Although the financial returns from a higher education degree over a lifetime typically exceed the costs, roughly 22 percent of student loan borrowers are in default. As a result, some have framed the “student loan crisis” as a crisis of student loan repayment rather than student loan debt. Since 2009 a range of income-driven repayment options has emerged to mitigate the financial burden for families by better aligning repayment obligations with their ability to pay.

A major complication in policymakers’ ability to propose promising solutions is the lack of data on how families—not just individual borrowers—are shouldering the burden of student loan repayment and the impact of student loan debt on other financial outcomes. The central challenge is that student loan payments and debt information are difficult to observe in conjunction with other financial outcomes, such as income, spending, and other debt payments, and certainly not on a high-frequency basis for large samples.

With this report, the JPMorgan Chase Institute aims to describe how student loan payments fit into the context of families’ larger financial lives. We offer the debate insight into a new, high-frequency cash flow perspective on student loan payments and how they relate to a family’s income, liquid assets, spending, and other debt payments. This perspective, based on student loan payment transactions observed out of a universe of 39 million Chase checking accounts between October 2012 and July 2018, is novel not just for its large sample size, but also its visibility into private and federal student loan payments (including any fees and fines), alongside income, spending, liquid assets, and other debt payments. In addition, this data asset is distinct in terms of its family perspective, which allows us to take into consideration the potential for a family to be making payments on multiple student loans and on behalf of other borrowers. This is an important, but often overlooked or hidden piece of the student loan repayment picture, given that roughly 19 percent of individuals report receiving help from others to pay off their student loans.

With this new data asset, we aim to answer five key questions:

  • What share of take-home income are families spending on student loan payments?
  • How does the financial burden of student loan payments differ across demographic groups?
  • How consistently do families repay student loans, and how volatile are repayment amounts?
  • In what ways do student loan payments differ from other types of loan payments, notably auto loan and mortgage payments?
  • How do student loan payments fluctuate with income, liquid assets, and expenditures?
01

The typical family’s median student loan payment is $179 per month or 5.5 percent of take-home income in months with positive payments. One in four families spend more than 11 percent of their take-home income on student loans.

Infographic describes about the Rolling Window Sample of accounts with at least one student loan payment within six months, March 2013 through July 2018.

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Finding One

Dumbbell dot plots showing the student loan payment amount and student loan payment burden for families in our sample during all months in our study, as well as only months with a positive payment. A typical family’s median student loan payment is $179 per month, or 5.5 percent of take-home income in months with positive payments. Additionally, one in four families spend more than 11 percent of their take-home income on student loans.

01

Younger and low-income families are most burdened by student loan payments, but there is no material difference in burden by male versus female account holders.

Infographic describes about Student loan payment burden by subsample

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Finding Two

Dumbbell dot plot showing student loan payment burden by subsamples, including age, gross income, and gender. Younger and low-income families are most burdened by student loan payments. There is no material difference in burden by male versus female account holders.

01

While overall 54 percent of families make consistent student loan payments, low-income families are less likely to make consistent loan payments (44 percent) compared to high-income families (63 percent).

Bar graph describes about Distribution of fraction of months with positive student loan payments, by gross income

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Finding Three

Histograms showing the distribution of fraction of months with positive student loan payments, by gross income. Overall, 54 percent of families make consistent student loan payments. Low-income families are less likely to make consistent loan payments compared to high-income families.

01

Among families actively paying multiple loans, the proportion making consistent payments is lower for student loans than auto loans (10 percentage point difference) and mortgages (6 percentage point difference).

Bar graph describes about Sample of accounts with at least two student loan payments and at least two other debt payments (auto loan on the left, mortgage on the right).

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Finding Four

Histograms showing the distribution of fraction of months with positive payments of student loans compared to auto loans and compared to mortgages. Among families actively paying multiples loans, the proportion making consistent payments is lower for student loans than auto loans by 10 percentage points and lower for student loans than mortgages by 6 percentage points.

01

Income, liquid assets, and expenditures increase sharply prior to starting student loan payments and decrease after stopping student loan payments.

Graph describes about Median credit & debit card spend, liquid assets, and labor income

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Finding Five

Line graphs showing labor income, liquid assets, and expenditures (median credit and debit card spend) in the window around the first student loan payment and window around the last student loan payment (in months). All three increase sharply prior to starting student loan payments and decrease after stopping student loan payments.

Data

For this study, we assembled several distinct data assets from an overall sample of JPMorgan Chase families that made student loan payments from their Chase checking accounts.

We began with a universe of 39 million families with Chase checking accounts between October 2012 and July 2018. From this universe, we constructed a subset of 30 million “core” accounts for which we observe sufficient activity to consider the account a primary financial vehicle for the family. From these core accounts, we identified 4.6 million families who have made at least one student loan payment out of their Chase checking account.

The data assets used for analysis were created from this base of 4.6 million families. Each sample uses different inclusion criteria and serves a different analytical purpose, described in the below graphic. For additional details, see the Data Asset and Methodology section.

 

Infographic describes about the data assets used for analysis were created from this base of 4.6 million families. Each sample uses different inclusion criteria and serves a different analytical purpose

Source: JPMorgan Chase Institute

© 2019 JPMorgan Chase & Co.

Data Asset

Financial outcomes studied: amount and frequency of student loan payments, payment burden as percent of account inflows, account inflows over time (labor income and other inflow sources), and account outflows over time (credit and debit spend, and auto and mortgage payments). Demographic views of the above financial outcomes are also studied with segmentation by age and gender of the primary account holder, and by gross income (annual).

We began with a universe of 39 million Chase checking accounts, which we filtered to 30 million “core” Chase checking accounts that have at least 5 transactions per month over a period of at least 6 consecutive months. From that sample, we identified 4.6 million customers with at least 1 student loan payment in their history. WE derived analytical samples from this set of 4.6 million customers as follows:

To analyze student loan payment levels and burden, we constructed a Rolling Window Sample of 4.1 million customers. Each month includes customers with student loan payment in the current or 5 preceding months. To assess student loan payment consistency and volatility, we constructed a Payment History Sample of 2.3 million customers with 2 or more student loan payments. We follow each customer in this sample from the month of first observed payment through the month of last. Finally, to study how student loan payments fluctuate with income, liquid asset, and expenditures, we constructed to event studies: the Payment Start Event Study (625,000 customers), and the Payment Stop Event Study (505, 000 customers). Each study includes customers with 2 or more student loan payments, who we follow for a 6-month lead-in window (or trailing window) of $0 student loan payments, and observe financial outcomes for an additional 24 months following the first (or preceding the last) payment.

Conclusion

Taken together, our insights from this new, high-frequency lens into student loan repayment behavior have important implications for policymakers, financial institutions, higher education institutions, and employers. There are important segments of the population who are still significantly burdened by student loan payments, especially younger and lower-income account-holders, despite the availability of income-driven repayment programs. In particular, student loan payments are sensitive to large income changes, and may lack sufficient mechanisms to adjust payments to accommodate income fluctuations. Insofar as student loan payments are less consistent and more volatile than auto loan and mortgage payments, families may be benefiting from the greater leniency that exists with student loan repayment compared to other loan types. Still, it remains to be seen whether the negative consequences of this lower consistency will outweigh the benefits of greater leniency. Overall, there may be better ways to structure or implement student loan repayment plans that would ensure that families are not over-burdened and are able to make consistent payments. Revisiting underwriting and federal student aid criteria and considerations might help address the root of the student loan repayment problem. More broadly, colleges and universities, employers, and financial institutions have a role to play in helping borrowers manage their student loan debt.

Authors

Diana Farrell

Diana Farrell

Founding and Former President & CEO

Erica Deadman

Erica Deadman

Consumer Research Lead

Fiona Greig

Fiona Greig

Former Co-President