The snowball approach to debt repayment involves paying down the smallest debt first, followed by the next smallest, and so forth.

If one were instead able to and follow through with paying down debt in order of highest to lowest interest rate, you'd pay the total debt down more quickly.

One proponent of the snowball method says:

people trying to reduce debt need "quick wins" (i.e., no more bills from creditors paid off in full) in order to remain motivated toward debt reduction

Is this really the case? Has it been demonstrated that people need quick wins? Do people following the snowball method have more success than those following the mathematically optimal highest-to-lowest-interest-rate strategy?

  • 2
    Wikipedia refers to "a Kellogg School of Management research with over 6,000 participants". The effect can only be a psychological one and the claim denies the usual (though wrong) assumption of "rational market participants". Instead we have to assume that those with more open bills pending are frustrated and spend more money in chocolate instead of paying bills ... Of course the effect depends on the spread between interest rates; I still suggest paying the 10% per day Mafia credit first. Personally, I'd rather decide rationally (as that would allow me to buy even more chocolate) – Hagen von Eitzen Dec 4 '14 at 9:56
  • 1
    ... On the other hand, if you have one debt of 100$ at 0% interest and 10,000,000$ at 5% and right now you have only 100$ available, the difference between fully cancelling the small debt vs. paying part of the big debt is relatively small: 5$ after one year. This is so negligible compared to the 500,000$ of interest that accumulate anyway that it doesn't make a lot of difference to spend these 5$ on the psycholigical effect suggested ... – Hagen von Eitzen Dec 4 '14 at 10:02
  • 1
    This sounds quite a bit like a Skinner box to me. Small wins to keep motivation up and train/enforce behavior has definitely been studied in other areas. I don't know if it's been rigorously studied for debt repayment, though. – Is Begot Dec 4 '14 at 14:53
  • 1
    What if I consider doing something optimally a small win, and paying more interest than necessary a loss? The snowball method wouldn't work for me then. Am I typical? Or is the person that Dave Ramsey imagines typical. – spoderman Dec 5 '14 at 16:17
  • 1
    @Spoderman: My understanding of Dave Ramsey's take on this is that if getting out of debt were a math (logical) problem, those same people wouldn't be in debt the first place. – poke Dec 6 '14 at 18:17

This is an interesting question. I've seen this idea bandied about personal finance blogs and forums before, and always thought it was a bit nutty, since, as noted in the comments to this question, paying off the debt with the highest interest rate first will minimize both total payments and time until you are debt free. Despite this, it is a notable claim, and one that is widely repeated (at least in my experience).

It turns out this strategy has been studied very recently by the US "National Bureau of Economic Research" (NBER), which appears to be a think-tank devoted to economic policy issues. They claim to be "non-partisan", though I have no way of verifying this.

The introduction of their working paper (page 1) provides a good synopsis of the claim and reasons why it might be true. The claim was popularized by radio host Dave Ramsey.

This paper also cites a 2012 analysis of data gathered from a debt collections agency, which showed that the snowball method was actually an extremely effective strategy for people in arrears. The number of accounts closed so far turns out to be a strong predictor of closing more accounts, regardless of the dollar amounts on the individual accounts. The dollar amounts paid off so far, in contrast, were not predictive.

The 2012 study already answers the question empirically (yes, at least for people sent to collections agencies, the snowball is effective), but the 2014 study investigates why this might be in much more detail, and proposes a full mathimatical model for the actions of human agents in paying off the debt (it looks like a great read, and doesn't seem to be paywalled). They also offer experimental results on a different set of tasks, showing that humans both i) perform tasks faster (voluntarily) when the tasks are ordered from fastest to slowest instead of slowest to fastest or randomly ii) humans are more likely to complete the entire set of tasks when they are ordered from shortest to longest than the other way round (71% completion rate vs. 48%!), or randomly (71% vs. 58%!). Interestingly, although past research found that, given the choice, humans prefer to pay off debt by ordering it from smallest to largest, in this study humans who were allowed to order their tasks actually preferred to do that largest ones first, even though this was less efficient!

  • 2
    It may just be that people that can’t manage depts, so become the target of a debt collections agency benefit from a different method then “logical” people that don’t let the depts get out of control in the first place. – Ian Ringrose Dec 6 '14 at 17:36
  • 1
    @IanRingrose Yes, that's consistent with the 2012 study's finding. The 2014 study however, shows that people in general (not just those sent to debt collectors) have a reward mechanism that makes this strategy effective. – John Doucette Dec 6 '14 at 21:52
  • I understand. In a video game, they make the first level easiest, the second level slightly harder, and so on. If they gave you the hardest level to start with, you would just give up. – GEdgar May 30 '15 at 21:35
  • I'm not sure the video game analogy fits; I think the general need to match difficulty to player learning and the narrative tension are more important. … With these real world task experiments I doubt the people are learning how to do things on top of the rest of the testing… – StarWeaver May 31 '15 at 18:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .