Four debts. One spreadsheet. The same question every month: which one gets the extra money?
There are two well-known answers. One says to send it to the debt with the highest interest rate. The other says to send it to the debt with the smallest balance. The first is usually called the avalanche. The second is usually called the snowball.
The two are not really arguing about arithmetic. On the arithmetic, rate-first ordering wins, and nobody disputes it. They are arguing about something harder to measure. Which plan does a real person keep going with for two or three years?
Here is the arithmetic, worked all the way through. Then the research on the other half.
The two orderings, briefly
Both methods start the same way. You keep paying the minimum on every debt, every month. Nothing gets skipped. Then you take whatever you can spare on top and send all of it to one debt.
When that debt is gone, its old payment joins the pile and rolls onto the next one. The total going out each month never drops. That rolling total does the real work in both methods.
The only difference is which debt you point it at. Avalanche ordering targets the highest interest rate, whatever the balance. Snowball ordering targets the smallest balance, whatever the rate. That is the entire disagreement.
The Consumer Financial Protection Bureau's own guide to reducing debt lays out both and does not pick one.
A worked example
Say the debts look like this.
- Store card: $900 at 28.99 percent
- Credit card: $8,400 at 22.15 percent
- Personal loan: $3,600 at 11.86 percent
- Car loan: $12,800 at 7.14 percent
Those rates are close to the real market in mid-2026. The Federal Reserve's G.19 consumer credit release put the average rate on credit card accounts assessed interest at 22.15 percent in the second quarter of 2026. It put 24-month personal loans at 11.86 percent and 60-month new car loans at 7.14 percent.
Store cards sit higher. The CFPB's 2025 report on the credit card market found private label retail cards averaged 31.3 percent in 2024, against 25.2 percent for general purpose cards.
Total debt here is $25,700. The minimums add up to $695 a month. Say there is $250 spare on top of that. So $945 goes out every month, and it stays at $945 until the last debt clears.
Month one
Interest is charged first. On the store card that is $900 times 28.99 percent, divided by twelve. About $21.74. The credit card charges $155.05. The personal loan charges $35.58. The car loan charges $76.16.
Added up, $288.53 of the $945 is gone before a single balance moves. Roughly three dollars in every ten are rent on money already spent.
In this example the store card is both the smallest balance and the highest rate. So both methods start in the same spot. The spare $250 joins the store card's $30 minimum, and $280 lands on it. It clears in month four either way.
After that the paths split.
Highest rate first
The order runs store card, credit card, personal loan, car loan.
The freed-up $280 rolls onto the credit card. That is a large balance at a high rate, so it takes a while. The credit card clears in month 24. The personal loan follows in month 27. The car loan is last, in month 32.
Total interest paid over the run: $4,472.
Smallest balance first
The order runs store card, personal loan, credit card, car loan.
The freed-up $280 rolls onto the personal loan instead. That is a $3,600 balance, so it moves fast. It clears in month 13. The whole pile then shifts to the credit card, which clears in month 27. The car loan is last again, in month 33.
Total interest paid over the run: $4,867.
The gap
Ordering by rate costs $394.69 less and finishes one month sooner. Call it $395 and a month.
Two of the four debts land in the same month under both plans. The store card is first either way, and the car loan is last either way. The whole difference comes from swapping the middle two.
The felt experience is not the same at all, though. Under balance-first ordering, half the debts are gone by month 13. Under rate-first ordering, month 13 arrives with one debt gone and the credit card still eleven months out.
Same money. Same monthly payment. Very different-looking first year.
Why the gap is smaller than it sounds
$395 over 32 months is real money. It is also about 1.5 percent of the starting balance, and less than half of one monthly payment.
Three things keep the gap narrow.
The first is that both methods pay every minimum. Nothing is deferred. The expensive debt keeps shrinking the whole time. It just shrinks slower.
The second is that the rolling payment catches up. In either order, the final debt is being hit by the full $945 near the end. The two plans converge.
The third is the one people miss. The highest rate and the smallest balance often sit on the same debt. Small balances tend to be cards, and cards tend to be the expensive end of the ledger. When the two rules agree, there is no trade-off to have at all.
The gap widens when the rules disagree loudly. A large balance at a punishing rate, sitting next to a tiny balance at a cheap one. It shrinks toward nothing when they nearly agree.
What the research on small wins found
The case for balance-first ordering is not that it is cheaper. It is that people finish it.
David Gal and Blakeley McShane, both then at Northwestern's Kellogg School, studied client records from a debt settlement firm. Their 2012 paper in the Journal of Marketing Research reported something odd. The share of accounts a client had closed predicted whether they cleared all their debt. The share of total dollars paid off did not, once account closures were accounted for.
Their estimate: "one year from enrollment, a client who has consistently paid down the smallest balances is 14% more likely to complete the program than one who paid down random balances."
That is observational data, and the authors treat it as such. People who close accounts may simply be more committed to begin with.
Later work tested the idea more directly. Keri Kettle, Remi Trudel, Simon Blanchard and Gerald Häubl ran a field study and three experiments, published in the Journal of Consumer Research in 2016. Concentrating repayments into one account, rather than spreading them across all of them, raised how aggressively people paid down debt. The effect was strongest when the money went into the smallest account.
Their explanation is neat. People judge their overall progress by the largest proportional drop in any one account. Paying $280 against $3,600 looks like progress. Paying $280 against $8,400 does not. It is the same $280.
Alexander Brown and Joanna Lahey ran a version of this in a lab, using a dull task instead of debt. Their working paper for the National Bureau of Economic Research found people worked through the parts faster when the parts ran smallest to largest. They also found that when people picked the order themselves, they chose that one least often.
None of this makes balance-first ordering cheaper. It is not cheaper. What it suggests is that finishing is something you can affect by design, and that the ordering is one of the levers.
Where the cheaper ordering loses
A plan that saves $395 saves nothing if it gets abandoned in month nine.
That is the honest shape of the trade-off. Rate-first ordering has a known arithmetic edge. Balance-first ordering has a claim on persistence. The first edge is certain and small. The second is uncertain and potentially much larger, because the alternative to finishing is not finishing.
Someone who has started and stalled twice already holds real information about themselves. So does someone who has run the same spreadsheet for eight years without missing a month. Those two people are not choosing between the same pair of options.
It is worth noticing what the example quietly assumes. Neither plan failed. Both cleared $25,700 in under three years. That only happens if $945 goes out every month for 32 straight months. The interest comparison takes for granted the exact thing that is hard.
When the ordering question is not the question
A few situations push this whole debate into second place.
A promotional 0 percent balance has no rate to attack until the promotion ends. What matters there is the end date and the rate that follows it.
A debt drifting toward default is not an ordering problem. Late fees, collections and credit damage tend to cost far more than any rate gap. The New York Fed's household debt report for the second quarter of 2026 put 4.7 percent of outstanding household debt in some stage of delinquency, with credit cards the sorest spot.
Secured debt behaves differently again, since falling behind can cost the car or the house.
And a very cheap fixed-rate loan raises a different question entirely: whether the spare $250 belongs in debt payoff at all.
Holding both facts at once
Two things are true here, and they do not cancel each other out.
Ordering by interest rate costs less. In this example, $395 less and one month sooner. The arithmetic behind that is not in dispute.
Ordering by balance clears accounts sooner, and the research links clearing accounts to staying with the plan.
The cheaper method is only cheaper if it gets finished. Which one gets finished is not a fact about the arithmetic. It is a fact about the person doing it, and it differs from person to person.
There is one thing both orderings share that may matter more than the choice between them. The monthly total has to hold steady when a debt disappears. In the example, $945 stayed $945 for 32 months. Let it fall back to $695 once the store card clears, and the payoff stretches well past either plan.
References and sources
- Board of Governors of the Federal Reserve System, G.19 Consumer Credit, released 7 August 2026. Source of the second-quarter 2026 rates used in the example: 22.15 percent on credit card accounts assessed interest, 11.86 percent on 24-month personal loans, 7.14 percent on 60-month new car loans.
- Consumer Financial Protection Bureau, The Consumer Credit Card Market 2025, December 2025. Source of the 31.3 percent private label and 25.2 percent general purpose average APRs for 2024.
- Consumer Financial Protection Bureau, How to reduce your debt. Describes both the highest-rate and smallest-balance approaches side by side.
- David Gal and Blakeley B. McShane, Can Small Victories Help Win the War? Evidence from Consumer Debt Management, Journal of Marketing Research, Vol. XLIX (August 2012), 487–501. Source of the account-closure finding and the 14 percent figure quoted above.
- Keri L. Kettle, Remi Trudel, Simon J. Blanchard and Gerald Häubl, Repayment Concentration and Consumer Motivation to Get Out of Debt, Journal of Consumer Research, Vol. 43, No. 3 (2016), 460–477. Field study and three experiments on concentrated versus dispersed repayment.
- Alexander L. Brown and Joanna N. Lahey, Small Victories: Creating Intrinsic Motivation in Savings and Debt Reduction, NBER Working Paper 20125, May 2014. Lab experiments on task ordering; later published in the Journal of Marketing Research in 2015.
- Federal Reserve Bank of New York, Quarterly Report on Household Debt and Credit, Q2 2026, August 2026. Source of the 4.7 percent delinquency figure.
This article is educational. It does not represent financial, tax, legal, accounting or investment advice. Consult the appropriate qualified professional advisors before acting on its contents.