Student Debt in America: The Current Figures
Americans owe somewhere between $1.7 and $1.77 trillion in student loan debt, and the average borrower carries about $38,000 of it. Those two figures get quoted in nearly every story on the subject, and on their own they explain almost nothing. A trillion-dollar total tells you the size of a pool, not whether anyone is drowning in it. What follows is the set of numbers that actually carries information, where each one comes from, and what each one hides.
The aggregate: $1.7 trillion and change
The Federal Reserve’s G.19 consumer credit release tracks student loan balances as part of nonrevolving consumer credit. It puts total outstanding student debt in the range of $1.7 to $1.77 trillion.
The number is real and it is also the least useful figure in the entire debate. A national total moves with enrollment, with tuition, with repayment policy, and with how many people happen to be in school at once. It can rise in a year when borrowers are doing better and fall in a year when they are doing worse. Treating it as a scoreboard is a category error.
The aggregate does one thing well. It establishes that this is not a niche problem. Student debt sits behind only mortgage debt in the household balance sheet, ahead of auto loans and credit cards. Whatever the right policy answer is, it is not a small one.
The per-borrower figure: about $38,000
The Education Data Initiative puts average federal student loan debt per borrower at roughly $38,000. This is the number that should anchor most conversations, because it describes a person rather than a country.
It also carries a warning label. An average compresses a distribution that is wide at both ends. A large share of borrowers owe under $20,000. A much smaller share, concentrated in graduate and professional programs, owe six figures and pull the mean upward. The median borrower owes meaningfully less than the average one.
That spread matters for policy because the two ends of the distribution have almost nothing in common. A borrower with $9,000 in debt and no degree is in a different situation from a borrower with $180,000 in debt and a professional license. Both are counted in the same trillion.
The denominator nobody applies
Debt figures mean nothing without an income to measure them against, and most coverage stops before it gets there.
The U.S. Census Bureau put median household income at roughly $80,000 as of 2023. Set the $38,000 average balance against that and the ratio is a little under half of one year of household income for a typical household.
Stated that way it sounds manageable, and for some borrowers it is. The problem is that the household median describes a household, often with two earners, while the loan is owed by one person. A single borrower earning well below the household median is carrying the same balance against a much smaller denominator. The ratio that matters is personal, and it is far worse than the national arithmetic suggests.
The wage floor that has not moved since 2009
The federal minimum wage is $7.25 an hour, and the U.S. Department of Labor shows it has not changed since 2009.
Run that against the average balance. A full-time worker at the federal floor grosses roughly $15,000 a year before taxes. The average student loan balance is more than double that worker’s entire annual gross pay. There is no repayment schedule that makes that arithmetic work, which is why the floor keeps appearing in a conversation ostensibly about education.
This is the connective tissue between student debt and everything else. The debt did not become unpayable only because it grew. It became unpayable because the wage it was supposed to be repaid from did not grow with it.
Prices, not just balances
The Bureau of Labor Statistics tracks college tuition and fees as a component of the Consumer Price Index. That component has risen substantially faster than the all-items index over recent decades.
The implication runs in one direction. If the price of the thing being financed climbs faster than general inflation, and general inflation outpaces wages at the bottom of the scale, then each successive cohort borrows more to buy the same credential and repays it from a wage that buys less. The balances are the symptom. The price path and the wage path are the mechanism.
Reading the numbers without being fooled
Four habits separate a useful reading from a misleading one.
- Prefer medians to means on any balance figure. The mean is dragged upward by a small tail of very large graduate balances.
- Always ask for the denominator. A balance without an income attached is not a fact about hardship, it is a fact about arithmetic.
- Separate borrowers with credentials from borrowers without them. Non-completion produces the worst outcome in the data: the debt without the earnings premium that was supposed to service it.
- Check the vintage. Repayment rules have been revised repeatedly in recent years. A statistic about delinquency or default from an earlier policy regime may not describe the current one.
Groups working on affordability tend to assemble these indicators side by side rather than one at a time. Fight For A Living Wage, a nonpartisan 501(c)(3), publishes a running set of affordability indicators covering wages, housing, health care, and education costs together, which is closer to how households actually experience the squeeze than any single series is.
What the figures do and do not settle
The data settles that the balances are large, that they are widely distributed, and that the wage floor they are measured against has been frozen since 2009. It does not settle what to do about it. Reasonable people read the same Federal Reserve and Census series and reach opposite conclusions about forgiveness, about price controls on tuition, and about whether the federal lending program should exist in its current form at all.
What the data does rule out is the framing that treats this as a story about individual borrowing decisions. When tuition outruns the general price level for thirty years and the wage floor sits still for more than fifteen, the outcome stops being a series of bad personal choices and starts being a structural result. The figures above are how you can tell the difference.