RAP or IBR? The Decision Framework for Clients Who Still Have a Choice
Since July 1, 2026, every federal borrower who does nothing lands on the Repayment Assistance Plan. But a client who took their last federal loan before that date still has a real choice: they can elect Income-Based Repayment instead. That choice is worth hundreds of dollars a month in either direction, and the direction is not obvious from the plan names.
One plan is now closed to almost everyone, so clear it off the table first. PAYE stopped accepting new enrollment on June 30, 2026, for every loan type. A client already enrolled in PAYE can stay on it for now, but nobody can newly elect it. For existing borrowers choosing a plan today, the live comparison is RAP versus IBR.
Here is the framework we use, with the actual crossover numbers, computed from the same engine that runs our client projections.
First question: which IBR does your client get?
There are two IBR cohorts, and the cohort changes the answer more than almost anything else.
- Borrowers with no federal loans before July 1, 2014 get the newer IBR: payments are 10% of discretionary income, with forgiveness after 240 payments.
- Borrowers who had a balance before July 1, 2014 get the original IBR: payments are 15% of discretionary income, with forgiveness after 300 payments.
Both versions define discretionary income as AGI above 150% of the federal poverty line for the household’s family size, and both cap the payment at what a 10-year Standard plan would charge.
RAP works on an entirely different chassis. There is no poverty-line offset. The payment is a tiered percentage of total AGI, from 1% up to 10% for AGI above $100,000, minus $50 per month per dependent, with a $10 monthly floor and no Standard cap.
The income bands, run through the engine
We scanned AGI from $20,000 to $250,000 in the calculation engine, using 2026 poverty guidelines. Here is where the cheaper monthly payment flips, by household shape, for the 10% IBR cohort:
| Household | IBR wins below | Contested middle band | IBR wins above |
|---|---|---|---|
| Single, no dependents | ~$30,000 | $30,000 to $70,000, RAP usually ahead | ~$71,000 |
| Single, 1 dependent | ~$38,000 | $38,000 to $70,000, winner alternates | ~$71,000 |
| Single, 2 dependents | ~$49,000 | $49,000 to $60,000, winner alternates | ~$61,000 |
| Married filing jointly, one income | n/a | RAP almost never wins | essentially all incomes |
Inside the contested band the winner flips at RAP’s tier boundaries: IBR tends to win right after each $10,000 line (where RAP’s percentage just jumped) and RAP tends to win the back half of each band. Which is the practical point of the table: outside the band you can call it from income alone, inside the band you have to run it.
Three more patterns fall out.
Dependents shrink RAP’s winning zone. IBR’s poverty-line offset grows with every family member. RAP’s $50-per-dependent deduction is flat and small by comparison. Each added dependent pulls the “IBR wins” threshold down by roughly $10,000 of AGI.
Marriage nearly erases RAP’s winning zone. A joint filer’s family size raises the IBR offset by a full poverty-line step, and a spouse adds nothing to RAP’s deduction. For a married couple filing jointly where one spouse earns most of the income, IBR beat RAP at essentially every AGI we scanned. A couple at $100,000 AGI with two kids pays about $650 on RAP and about $492 on IBR, and the gap widens from there.
At the very bottom, IBR’s floor is $0 and RAP’s is $10. Below roughly 150% of the poverty line, IBR charges nothing, and those $0 months still count toward forgiveness and PSLF.
For the 15% cohort, the table almost inverts. RAP’s tiers beat a 15% discretionary calculation at nearly every income above roughly $28,000 for a single borrower, and above roughly $45,000 with two dependents. If your client had loans before mid-2014 and is single, RAP is probably their cheaper payment. Married is the exception again: with a non-earning spouse and kids, the winner flips back and forth between $70,000 and $125,000, so it has to be run, not guessed.
Watch the $10,000 cliffs
RAP’s tier percentage steps up at every $10,000 of AGI, and the step applies to the whole AGI, not the marginal dollar. A single borrower at $80,000 pays $466.67 a month on RAP. At $81,000, the tier ticks from 7% to 8% and the payment jumps to $540. One raise, $73.33 more per month, $880 more per year.
This makes AGI management a live planning lever in a way the old plans never were. Retirement plan deferrals, HSA contributions, and bonus timing can hold a client under a tier boundary. It also means any RAP-versus-IBR answer near a $10,000 boundary is fragile, so model next year’s income, not just this year’s.
Monthly payment is not the decision. Four modifiers.
The bands above answer “which payment is lower today.” Four things can override that answer.
1. The Standard cap protects high earners on IBR. An IBR payment can never exceed the 10-year Standard amount for the balance. RAP has no cap. A client earning $250,000 with a $60,000 balance pays a capped, Standard-sized payment on IBR while RAP would take a flat 10% of AGI. High income plus a modest balance points hard at IBR, or simply at paying the loan off. We covered the high-earner default problem in depth in RAP is the default, and for your best clients it is the wrong plan.
2. RAP’s balance can never grow. RAP waives any interest the payment does not cover, and adds a $50-per-month principal credit when the payment is too small to reduce principal on its own. Negative amortization is impossible on RAP. On IBR, a payment below accrued interest means the balance climbs. For a low payment against a large balance, RAP’s subsidy can be worth more than IBR’s lower payment, especially for a client who may earn much more later.
3. The forgiveness clocks are very different, and prior time carries. IBR forgives at 240 payments (10% cohort) or 300 payments (15% cohort). RAP forgives at 360. Months already spent on SAVE, PAYE, IBR, or ICR carry into whichever plan the client moves to. A client 15 years into an IDR clock is 5 years from forgiveness on the newer IBR and 15 years from forgiveness on RAP. For anyone deep into a clock, the shorter horizon can dominate every payment difference in the table above.
4. Forgiveness at the end of the term is taxable; PSLF is not. Under current law, a balance forgiven at the end of an IBR or RAP term is ordinary income in the forgiveness year. A client projected to have a large balance forgiven needs that tax modeled now, not discovered at year 20. PSLF forgiveness remains tax-free, which leads to the shortcut below.
The PSLF shortcut
If your client is on a PSLF track, the framework collapses to one line: both RAP and IBR are qualifying plans, the 120-payment clock does not care which one the client is on, and forgiveness is tax-free either way. Pick the lower payment from the bands above and move on.
One warning before any of this
This menu only exists for clients whose last federal disbursement was before July 1, 2026. A single new federal loan after that date, including one semester of grad school borrowing, collapses the entire portfolio’s options to RAP and the new Standard plan. If a client is considering more school, that decision and the plan decision have to be made together.
And remember which way the default cuts: RAP happens automatically, IBR only happens if someone files for it. Every client for whom IBR wins is a client who must take action, and most will not know that.
Run it on real numbers
The bands in this post are honest, but they are bands. Your client is a specific AGI, family size, filing status, cohort, balance, and forgiveness clock, and the four modifiers move real money. Finology Software models RAP against IBR on a client’s actual loan data, side by side, monthly payment and lifetime cost, with the forgiveness tax accounted for.
Run any client under both plans at finology.tech. Already have an account? Log in.
Run a client-ready repayment plan in minutes
Every number sourced, every path compared. Model RAP, the new Standard, IBR, PAYE, ICR and PSLF side by side.
