The Plan That Got Surrendered

Introduction

In January 2024, 3.6 million Americans had a federal student loan payment of $0 per month, legally, on paper, according to the GAO. Twenty-three months later, the Trump Department of Education signed a settlement with Missouri's attorney general killing the plan that produced those numbers, rather than continue the legal defense in a case headed back to district court. The contractors who will now move 7.5 million borrowers off the plan and onto something more expensive include MOHELA, the Missouri state instrumentality whose participation in the lawsuit established standing, sitting on $443.9 million in active federal servicing contracts.

The settlement nobody was forced to sign

The SAVE plan launched in August 2023 as the most affordable federal student loan repayment program in U.S. history. Undergraduate borrowers paid 5% of discretionary income (down from 10%), income protection rose to 225% of the federal poverty line from 150%, all unpaid interest was waived for on-time payments, and anyone with $12,000 or less in original principal qualified for forgiveness after 120 payments. By mid-2024, nearly 8 million people had enrolled, roughly a quarter of all federal borrowers in repayment.

Seven Republican-led states sued in April 2024 (Missouri, Arkansas, Florida, Georgia, North Dakota, Ohio, Oklahoma). The Eighth Circuit issued a nationwide injunction in July 2024, then in February 2025 affirmed and expanded it on "major questions" doctrine grounds. The case was then remanded to the Eastern District of Missouri for final judgment, which is where the policy choice (not the legal mandate) comes in.

On December 9, 2025, the Department of Education announced a settlement. Under Secretary of Education Nicholas Kent's statement: "For four years, the Biden Administration sought to unlawfully shift student loan debt onto American taxpayers." The settlement vacated the SAVE plan's final rule (88 Fed. Reg. 43,820) in its entirety, blocked new enrollments, denied pending applications, and required all 7.5 million current SAVE borrowers to transition to other plans. The Trump DOE was not legally compelled to do any of this. The district court had the remand, and the DOE could have defended the plan's non-forgiveness elements (interest subsidy, reduced payment formula) and let a federal judge rule on what survived. Instead, it signed.

In February 2026, the district court dismissed the case rather than approve the settlement, noting that the parties were now aligned on SAVE's demise. For roughly two weeks, no injunction was in force, and borrower advocacy groups argued the DOE could have revived SAVE and processed forgiveness during that window. It didn't. On March 10, 2026, the Eighth Circuit reversed the dismissal in a two-sentence ruling and ordered final judgment entered. Winston Berkman-Breen of Protect Borrowers put it bluntly: "President Trump and Republican Attorneys General worked together to ask a hand-picked, conservative federal court to kill the SAVE plan."

The 7.5 million account migration with no audits

The transition starts July 1, 2026. Loan servicers begin sending notices, borrowers get 90 days to choose a new plan, and anyone who doesn't choose gets auto-enrolled in the Standard or Tiered Standard plan. Two replacements launch the same day: the Repayment Assistance Plan (RAP) and the Tiered Standard Plan. RAP's forgiveness timeline is 30 years instead of SAVE's 10-25. For most borrowers, payments will rise; IBR, PAYE, and ICR are "almost universally more expensive than the SAVE plan" per Forbes' Adam Minsky.

The servicers running this migration are the five companies on the Unified Servicing and Data Solution (USDS) contract: Nelnet, MOHELA, Maximus/Aidvantage, EdFinancial, and Central Research Inc. Under the contract terms documented by the National Consumer Law Center, about 60% of servicer revenue comes from contact center and back-office processing work. Servicers face up to a 20% invoice penalty for failing service-level metrics, including a 95% accuracy floor.

GAO-26-108534, released March 2026, documents what happened to those accuracy metrics. After the April 2024 USDS contract launch, 4 of the 5 servicers failed accuracy standards in at least one of the two quarters Federal Student Aid managed to assess. Then FSA stopped doing the assessments. Accuracy and call quality reviews were suspended in February 2025 "due to lack of FSA staff capacity"; FSA staff fell from 1,433 to 777 between January and December 2025, a 46% reduction. GAO recommended the Department of Education resume the reviews. FSA Acting Chief Operating Officer Richard Lucas rejected the recommendation in writing in April 2026. The largest student loan account migration in the modern era starts July 1 with no independent federal check on whether servicers are calculating new payments correctly or tracking forgiveness credit.

For context on what that means: when the 2012-2013 ACS servicing contract ended, the result was "more than 5 million servicing errors affecting more than 1.3 million borrowers," per a CFPB review cited in the NCLC report. That transition happened with accuracy oversight running. This one moves 7.5 million accounts at once without it.

Who benefits

Nicholas Kent / Trump DOE: political framing and a $260 billion revenue line. The administration converts what would have been future loan forgiveness into future borrower payments. CBO estimated in 2023 that rescinding the SAVE rule would yield the federal government $260 billion over 10 years through increased payments. By settling rather than litigating, the DOE also vacated the entire rule, including the parts the courts hadn't ruled on, like the interest subsidy. Kent's framing ("if you take out a loan, you must pay it back") provides the public narrative; the budget benefit is the structural one. Wage garnishment notices started going out the week of January 7, 2026, authorizing seizure of up to 15% of take-home pay from defaulted borrowers.

MOHELA: $443.9 million in active USDS contracts and a Missouri-shaped reason to want SAVE gone. The Missouri Higher Education Loan Authority is, per the Supreme Court in Biden v. Nebraska, "an instrumentality of the state of Missouri." Missouri's standing to sue rested on MOHELA's financial relationship with the federal loan program, and MOHELA holds three active USDS task orders under USAspending: PIID 91003124F0320 at $223.6 million, PIID 91003125F0043 at $91.8 million, and PIID 91003126F0021 at $128.5 million. Total: $443.9 million. The servicer compensation structure pays on contact center and back-office volume, so 3.6 million $0-payment SAVE accounts generate maintenance work without payment-processing volume. Moving those borrowers to active monthly payments under standard plans generates the kind of servicing activity the contract is designed to pay for.

Nelnet: $392.4 million in active USDS contracts plus a $205.3 million follow-on order. Nelnet's current USDS task order (PIID 91003124F0321) runs through March 2027 at $392.4 million. A follow-on order (PIID 91003126F0022) at $205.3 million covers April through December 2026, directly through the SAVE migration window. The accuracy-review gap GAO documented applies to Nelnet just as it does to MOHELA, and the same servicing-fee mechanics turn the migration event itself into billable work.

What the default cliff looks like from here

The Debt Collection Lab at Princeton estimates approximately 5 million borrowers likely defaulted in October 2025 alone, on top of the 5 million already in default. If current delinquency trends hold, as many as 13 million borrowers may end up in default by the end of 2026. Student loan delinquency hit 9.6% in Q4 2025 and is climbing; subprime mortgage delinquency at its 2008 peak was just under 12%. Once a borrower defaults, the federal government can administratively seize up to 15% of paychecks and intercept tax refunds. Treasury Offset Program collections restarted in May 2025 and wage garnishment began in January 2026.

The legal argument against SAVE was not frivolous. The Eighth Circuit's "major questions" analysis read the Higher Education Act as not clearly authorizing mass forgiveness through regulatory action, and reasonable judges agreed. The policy choice was the decision to abandon the defense before the courts finished resolving which parts of the plan could survive. The interest subsidy and the reduced payment formula were never ruled on, because the DOE settled the case before a judge got there.

The bottom line

The thread version of this story stops at the settlement and the $443.9 million MOHELA number. What it can't fit: the two-week window in February 2026 when SAVE was technically alive and DOE could have processed forgiveness, the mechanics of how USDS servicer compensation rewards moving borrowers off $0-payment plans, and the GAO timeline showing the accuracy reviews didn't just lapse. The DOE was asked to restart them and put the refusal in writing.

If you were one of the 7.5 million SAVE borrowers, the question now is not whether your payment goes up. It's whether the servicer recalculating it has been independently audited for accuracy since February 2025. The answer is no, and the entity that decides whether to fix that has already said it isn't going to. What happens in August 2026, when the first round of borrower complaints lands (wrong payment amounts, lost forgiveness credit, plan transfers that never went through, account histories that don't add up), if it lands the way the 2012 and 2021 transitions did?