If you believed your student loans might be forgiven in the future, would you reduce your payments now? What if they weren’t forgiven after all? This paper tackles such questions about how policy uncertainty shapes spending decisions and, in turn, welfare.

The authors study these questions through the lens of student loan forgiveness in the United States, which, following electoral promises, was announced in 2022 but never implemented due to judicial rulings. The August 2022 plan would have forgiven $10,000 for most borrowers, and $20,000 for those who had received Pell grants, subject to an income cap of $125,000 for single filers and $250,000 for married couples. Federal loan payments had also been paused since March 2020, with the pause extended seven times before ending in October 2023.

The authors surveyed borrowers on whether they believed loans would be forgiven, and by how much, and separately on whether they expected the pause to be extended again. They then linked the survey results to data on credit and spending to see how beliefs affect behavior. They find the following:

  • There is substantial variation in respondents’ beliefs concerning both the likelihood of loan forgiveness and the amount they predict will be forgiven.
  • These beliefs are sensitive to both news about policy events and experimental treatments regarding the likelihood of loan forgiveness.
  • Individuals who are more optimistic about forgiveness make fewer payments and are less likely to pay down their student loans. They also appear to shift that freed-up cash toward everyday spending — borrowers with balances under $10,000, whose beliefs turned optimistic after the August 2022 announcement, raised everyday spending by roughly $80 per month — while holding off on large, hard-to-reverse purchases like homes and cars until the uncertainty resolved. 
  • Borrowers who expected the COVID-era payment pause to be extended again made fewer payments once repayment restarted, reduced monthly payments by roughly $40, and were 7.5 percentage points more likely to be 90 days past due by May 2025. 
  • These actions can prove costly. In a model calibrated to the survey and credit data, if forgiveness ultimately occurs, pessimistic borrowers who repay too quickly can experience losses of 43.9 percent of the initial loan balance because they pay down balances that would otherwise have been forgiven. If forgiveness does not occur, optimistic borrowers who extend repayment in anticipation of relief incur smaller but still meaningful losses, up to 6.9 percent of the initial loan balance, due to additional interest and delayed repayment.

In a policy area with substantial uncertainty, differences in beliefs can affect actions. This finding has important implications for policy guidance, as statements from policymakers can lead to changes in behavior that have significant welfare consequences for borrowers. 

Written by Abby Hiller • Designed by Maia Rabenold