Americans owe $1.7 trillion in student loans — more than all credit card or auto debt. Here's how the student loan system works, who profits, and why the debt keeps growing despite decades of policy attention.
Educational content only — not personalized financial, legal, or tax advice.
Americans owe approximately $1.7 trillion in student loan debt — more than all credit card debt ($1.1 trillion) or auto loan debt ($1.6 trillion). Approximately 43 million borrowers carry federal student loan balances. The average balance for a bachelor's degree holder is approximately $30,000; for graduate and professional degree holders, $80,000-200,000+.
The system has grown dramatically: total student debt was approximately $260 billion in 2004 and $800 billion in 2010. The growth reflects both rising tuition (which grew 169% above inflation from 1980 to 2020) and expanding access to federal loans (which have no credit check or income verification, removing the market discipline that limits other forms of consumer lending).
This connects to Parametric Insurance and to Climate Adaptation Finance).
The federal government is now the dominant student lender, originating approximately 93% of all student loans through the Department of Education's Direct Loan program. Private student loans (Sallie Mae, Navient, Earnest, SoFi) account for approximately 7%.
Loan types: Direct Subsidized Loans (interest doesn't accrue during school, for undergrads with financial need), Direct Unsubsidized Loans (interest accrues during school, for undergrads and grads regardless of need), PLUS Loans (for graduate students and parents, no aggregate limit, higher interest rate).
Interest rates: Set annually by Congress based on 10-year Treasury yield plus a fixed add-on. For 2023-24: undergraduate subsidized/unsubsidized at 5.50%, graduate unsubsidized at 7.05%, PLUS at 8.05%. These are fixed for the life of the loan.
Income-Driven Repayment (IDR): The policy that has most changed student loan economics. IDR plans cap monthly payments at 5-10% of discretionary income, with balances forgiven after 10-25 years. The SAVE plan (2023) reduced undergraduate loan payments to 5% of discretionary income. IDR creates a de facto income-contingent grant for borrowers who never earn enough to fully repay — particularly relevant for social work, teaching, and other lower-earning fields.
The federal government (partially): The government earns interest on student loans. Pre-IDR expansion, student loans were modestly profitable for the federal government. Post-SAVE, the Congressional Budget Office estimates the federal student loan portfolio operates at a significant cost to taxpayers — approximately $400 billion in net subsidy over the 2023-2032 window.
Servicers: Companies that collect payments, process applications, and handle customer service for the Department of Education under contract. MOHELA, Aidvantage (Maximus), Nelnet, and Edfinancial are the primary servicers. Servicers earn approximately $2-3 per borrower per month — modest per-account revenue that adds up across tens of millions of accounts. Servicer quality has been a major point of criticism — widespread errors in income-driven repayment processing have been documented by the CFPB.
Universities: The absence of loan limits (particularly for graduate and professional programs) has allowed universities to raise prices without constraining enrollment. The university captures tuition revenue; the borrower assumes the debt. This dynamic has been most extreme in graduate programs: a master's degree in social work from a major private university costs $80,000-120,000 for a career that might start at $45,000/year.
President Biden's broad-based student loan cancellation ($10,000-20,000 per borrower) was struck down by the Supreme Court in June 2023. Targeted cancellation programs (Public Service Loan Forgiveness, borrower defense to repayment, IDR account adjustments) have proceeded and have cancelled approximately $150 billion in debt for approximately 4 million borrowers through early 2024.
The economic debate: broad cancellation is regressive (college graduates earn more than non-graduates; directing benefit to graduates over non-graduates favors relatively higher earners). The counterargument: graduate degree holders often have the highest balances and can least afford repayment in social service fields. Neither argument fully resolves the distributional complexity.