This report sheds light on the possible spending adjustments these households will need to make in the face of their restarted student loan payments. We use de-identified administrative banking data from Chase checking and savings accounts to measure households’ income and expenses, and we project how much a household’s student loan payments will increase in coming months based on their student loan payments right before the pandemic.
Our analysis suggests that budget reallocations will be modest for most households, but many may need to decrease their other expenses to cover their student debt payments. Our key findings are:
- 15 percent of mortgage holders in our sample had been making regular student debt payments before the pandemic.
- Among these joint mortgage–student debt payers, the median household will need to use roughly 3 percent of their take-home income for student debt payments once they resume.
- While this 3 percent increase is relatively modest, we project that, given recent levels of spending and other expenses, the median household with both student debt and mortgage obligations will have a budget shortfall of 0.8 percent of take-home income once they start making full student debt payments. This means that many households will need to decrease spending or use savings to cover student debt payments.
- Lower-income mortgage holders will need to make larger budget reallocations: in the lowest income quartile, median reallocation is projected to be 4.6 percent of take-home income versus 2.2 percent in the highest income quartile.
Overall, we find that the student debt payment restart should have modest effects. We estimate that a relatively small portion (15 percent) of mortgage holders in our sample are also expected to restart student debt payments, and among those households, the median payment amount will be roughly 3 percent of take-home income. However, these seemingly small figures mask crucial nuances. Many households will need to reduce their other expenses in some way to make their restarted student debt payments without drawing on savings or incurring additional debt. It’s important to note that we would expect a natural decrease in discretionary spending to occur—independently of the payment restart—as households finish spending down the excess savings they accumulated in 2020 and 2021. Nevertheless, we project budget adjustments will be slightly larger for lower-income households, and these households may have more difficulty making these adjustments. And while the Biden administration has taken some steps to allow flexibility in how people resume payments, additional monitoring of households’ response to the end of the pause, as well as their financial security, is warranted.
Data and Methodology
Our analysis uses the bank transaction data of 455,000 deidentified households’ personal checking and savings accounts. Households are included in our sample if they meet two criteria. First, they must have held a Chase-serviced mortgage during the six months spanning February through July 2023. Second, they must have had a Chase checking account that meets minimum usage requirements to ensure that the household conducts most of its financial activity through these accounts. We calculate each household’s average monthly expenses and take-home income from the transactions in these accounts.
We flag households as being involved in student debt repayment if they made student debt payments in at least five of the six months before the start of the pandemic (September 2019 through February 2020)., We use the monthly average of the household’s student payments in these pre-pandemic months as their baseline student debt payment, and we assume that households will resume paying this amount once the payment pause ends. We calculate the projected additional amount households will need to pay after the resumption of student debt payments as the difference between this pre-pandemic baseline amount and the monthly average of payments the household made from February through July 2023.
We use February through July 2023 to establish the steady state spending before the end of the payment pause for several reasons. Because our question of interest is how well households will fit the re-started payments into their month-to-month budget, we want to estimate households steady-state payment habits. As such, we do not want to include one-time, lump-sum payments on loans in our estimate of households’ routine monthly behavior, and work by other researchers found a surge in such payments beginning in August 2023. Similarly, we do not want to include debt payment data later in 2023 because of borrower confusion about when payments were due, inconsistency in servicer communication to borrowers, and the Biden administration’s decision not to report borrowers as delinquent on their payments for the first year after the payment restart. As a result, only 60 percent of borrowers with payments due in October had made those payments by mid-November (Kvall 2023).
Results
To assess how much capacity household budgets will have for restarted student debt payments, we calculate the fraction of a household’s income was used for expenses before the pandemic and right before the restart of payments and present the results in Table 1. Of our sample of homeowners, 15 percent made regular student debt payments before the pandemic.
Table 1: Income and spending among homeowners
|
No student debt payments
|
Regular debt payments before pandemic
|
Pre-pandemic |
End of pause |
Pre-pandemic |
End of pause |
(Sept '19 - Feb '20)
|
(Feb - July '23)
|
(Sept '19 - Feb '20)
|
(Feb - July '23)
|
Fraction of take-home income left after
|
|
|
|
|
all expenses except student debt
|
|
|
5.9% |
3.1% |
all expenses
|
6.9% |
3.0% |
1.4% |
2.2% |
all expenses plus projected increase
|
|
|
|
-0.8% |
in student debt after end of pause
|
|
|
|
|
Annualized take-home income |
$83,637 |
$106,052 |
$111,479 |
$142,723 |
Notes: “Pre-pandemic” period is September 2019 through February 2020. “End of pause” period is February 2023 through July 2023. “Regular student debt payments before pandemic” means that a household made student debt payments in at least 5 of the 6 months in the “pre-pandemic” period. All dollar values are nominal. Projected increase in student debt payments is the difference between average student debt payments made during the pre-pandemic sample period and average student debt payments made before the end of the payment pause. Sample of homeowners is Chase checking account customers who also held a Chase mortgage in the 6 months before payments resume and whose checking account activity meet minimum activity thresholds.
The first two rows of Table 1 describe what fraction of their income homeowners used to cover expenses right before the pandemic and right before the end of the payment pause. These expenses include all outflows from the household’s accounts except for account-to-account transfers, including mortgage and other debt payments, credit card payments, and debit card spending such as food, clothing, and transportation. In the six months before the pandemic, the median homeowner not making student debt payments had 6.9 percent of their take-home income remaining after all expenses. For homeowners who made regular student debt payments before the pandemic, the median was lower at 1.4 percent. When looking only at non–student debt expenses, student debt payers still use a higher share of their income on expenses, with only 5.9 percent of take-income left over.
Over the pandemic, both groups used more of their take-home income. The median non–student debt homeowner had 3 percent of their take-home income left after expenses in the six months ending July 2023 and the median student debt–homeowner had 3.1 percent of their income left after all non-student debt expenses. Some households continued to make student debt payments despite the payment pause: adding these payments decreases median income remaining to 2.2 percent, a difference of 0.9 percentage points.
We next consider the hypothetical scenario that households spent the same amount of money on student debt payments in 2023 as they did right before the pandemic. In this scenario, the median household would have a shortfall of 0.8 percent of take-home income assuming they made student payments at the same level as before the pandemic and did not alter any of their spending or other expenses. That is, they would have to spend 100.8 percent of their take-home income, a difference of 3.0 percentage points relative to actual expenses. This suggests that the median household will need to reallocate roughly 3 percent of their take-home income, either from spending or saving, to make their restarted student debt payments.
Households with a projected income shortfall will either need to use savings to cover their spending and full student debt payments or they will need to decrease their other consumption slightly to avoid drawing down savings. It is important to note that even without the restart in payments, we anticipate that households will likely decrease their spending during the period of the payment restart as households finish spending down the excess savings they accumulated in 2020 and 2021. As Wheat and Eckerd (2023a) show, households tend to maintain a target cash buffer on hand. When their cash buffer rises above this target, households increase their spending. By construction, for a household to spend down their savings, they must have expenses exceeding their income. As a household’s cash buffer approaches their target, they decrease their spending back to baseline levels and resume their usual saving rate. As Wheat, Deadman, and Sullivan (2024) show, households spent out of their excess savings (and transferred to other accounts) throughout 2023. And while many lower-income households have resumed their usual spending and saving patterns, higher-income households—which are disproportionately represented in our sample—are not done dissaving.
Finally, we consider how much households of different incomes will need to budget for their restarted payments. Figure 1 shows that the fraction of take-home income needed to cover restarted payments decreases with income, ranging from 4.6 percent for our lowest income quartile and decreasing to 2.2 percent for our highest income quartile. This is consistent with past Institute findings that student debt payment burdens (i.e., monthly payments divided by monthly income) are negatively correlated with income (Farrell, Greig, and Sullivan 2020).