In August 2022, the Biden administration announced that most federal student loan borrowers would see up to $10,000 wiped from their balances, with Pell Grant recipients getting up to $20,000. The plan never happened. Nearly a year of legal battles ended with the Supreme Court striking it down in June 2023. But between those two events, millions of borrowers had to make real decisions with real money — pay down loans now, or hold off in hopes the debt would vanish?
A new NBER working paper tracks what borrowers actually believed during that window of uncertainty, and how those beliefs shaped their financial behavior. The short version: optimism about forgiveness led borrowers to pay less on their loans and spend more on other things. And for some, that optimism turned expensive once the courts made their final ruling.
A natural experiment in policy limbo
Dmitri Koustas of the University of Chicago and his co-authors Michael Weber (Purdue) and Constantine Yannelis (University of Cambridge) saw the extended legal battle as an unusual chance to study a question economists have long puzzled over: when the government promises something but that promise might not materialize, how do households respond?
Governments regularly signal future policy changes — tax cuts, benefits, tariffs, stimulus payments — but many of these announcements come with substantial uncertainty about whether they’ll actually happen. The researchers wanted to link what people believed about these signals to what they actually did with their money.
Student loans offered a rare setup. Forgiveness had been proposed, then formally announced, then challenged, then blocked. The payment pause that began during the pandemic was extended seven separate times. Beliefs about all of this shifted week to week, and the researchers could measure both the beliefs and the resulting behavior.
Building the survey and linking it to real financial data
The team designed a survey asking borrowers detailed questions about their expectations. What percent chance did they think there was that at least some of their loans would be forgiven in the next year? The next 2-5 years? Ten years? What dollar amount did they expect? Would the payment pause be extended again?
They ran the survey in waves from mid-2022 through 2025, capturing beliefs before and after each major event: the initial forgiveness announcement, the court injunctions, the Supreme Court decision, and the eventual restart of payments. They also embedded a randomized experiment. Roughly one-third of respondents saw an optimistic news article about forgiveness, one-third saw a pessimistic article, and the rest saw nothing. This let the researchers test whether nudging beliefs actually changed behavior.
The survey was linked to two rich datasets: anonymized credit bureau records covering loan balances, payments, delinquencies, mortgages, auto loans, and credit scores; and Numerator, a consumer panel that captures spending by scanning receipts. Employment and earnings records were available for about a third of the sample. In total, the researchers collected 8,922 completed responses from 4,802 unique borrowers through the credit bureau sample, plus 3,028 through Numerator.
Beliefs were all over the map
The first finding was how much borrowers disagreed with each other. In mid-2022, the average belief about the chance of at least some forgiveness in the next year was 25 percent, with a standard deviation of 30 percentage points. Some borrowers thought forgiveness was essentially guaranteed. Others thought there was zero chance.
Beliefs moved with the news. The August 2022 announcement pushed average expectations up by roughly 22 percentage points. Reading an optimistic news article in the experiment nudged beliefs up by about 4 percentage points. After the Supreme Court decision, beliefs fell below where they had been before the whole saga began. By May 2025, average expectations for forgiveness in the next year had dropped to around 9 percent.
Even at moments of peak optimism, most borrowers remained skeptical. Roughly half cited concerns about the legality of the plan; about a third weren’t sure whether they’d personally qualify.
Optimism translated into fewer payments and more spending
The behavioral patterns lined up with what the beliefs would predict. Borrowers in the top half of the optimism distribution were about 30 percentage points less likely to make any student loan payment during the pause and paid roughly $100 less per month on their loans on average.
The randomized information experiment confirmed that this wasn’t just correlation. Borrowers who happened to read the optimistic news article subsequently made fewer loan payments in the following months. A 10 percentage point increase in belief about forgiveness was linked to a 3.7 percentage point drop in the likelihood of making a payment.
Spending patterns shifted too. Around the August 2022 announcement, borrowers who had been pessimistic about forgiveness — and who had balances small enough ($10,000 or less) to be fully wiped out — reduced their student loan payments by about $40 per month and increased non-durable spending by roughly $100 per month. Much of that extra spending went to discretionary categories rather than food.
Interestingly, more optimistic borrowers appeared to hold off on big-ticket purchases. Their mortgage balances grew less than pessimists’ during the uncertain period, consistent with them waiting to see how things shook out before making major commitments.
The bill came due
Being wrong about forgiveness carried consequences. Borrowers who had been most confident that the payment pause would be extended past October 2023 were less likely to resume payments when it actually ended. By May 2025, after the “on-ramp” grace period ended and missed payments started hitting credit reports, these optimists were 7.5 percentage points more likely to be 90 days past due on their loans compared to borrowers who had expected the pause to end.
To gauge the size of the damage, the researchers built a simple model of a borrower who chooses how much to pay each month and whether to take a 10-year or 20-year repayment plan, all while forming beliefs about future forgiveness. Within a fixed repayment plan, the welfare costs of holding wrong beliefs were modest — under 2 percent of the initial loan balance.
But once the choice of repayment term entered the picture, the numbers grew dramatically. A pessimistic borrower who chose a 10-year plan and paid down aggressively — only to have forgiveness actually arrive — could lose up to 43.9 percent of the initial loan balance in equivalent consumption. An optimist who extended repayment expecting relief that never came would lose up to 6.9 percent.
What it means for policy communication
The authors argue that their findings have implications beyond student loans. When policymakers announce initiatives that later face legal or political obstacles, households don’t just wait passively for resolution. They adjust spending, borrowing, and repayment decisions in ways that can be costly if the policy never arrives, or if the announcement is reversed.
The researchers are careful to note that this doesn’t mean governments should stay quiet about policy plans. Rather, they interpret the results as pointing to the value of clarity and credibility in policy communication. Ambiguous signals about future benefits, in their reading, can lead households to make decisions that turn out to be privately costly.
One caveat worth flagging: the paper’s sample of survey respondents had higher-than-average student loan balances, which may reflect that borrowers with more at stake were more motivated to participate. The behavioral responses documented here may be strongest for borrowers with meaningful skin in the game, and may look different among those with smaller balances or looser attachment to the news cycle.
