Video: How debt repayment strategies affect consumer motivation to get out of debt

Keri Kettle: Thank you Dilip for the wonderful introduction. I’d like to thank Dilip for inviting me to speak here and the organizers from the FCAC. It’s a real privilege. I’d also like to thank Sam Hirschman for giving you all the background information on what a big problem credit card debt is. My research and this research was published in the Journal of Consumer Research in 2016 so you don’t need to take pictures of the slides.

The paper exists. Please email me if you’d like a copy. We look at how debt repayment strategy affects motivation. I want to give you a stylized example here. We know that most consumers who have problematic debt have more than one debt. Here which debt should you pay first?

We have different interest rates. The smaller account $500 has a higher interest rate. This probably replicates what a lot of consumers experience. They might have a credit line, a larger amount than a credit card. I’d say most people would agree that you probably want to pay off account B first, higher credit rating, smaller account.

What if they are the same rate? Perhaps the econs are now panicking because they don’t know what to do. Perhaps we have a behavioural that says well, you know, that $500 I’d like to get rid of that. Scott has a paper that says people are averse to accounts. We all want to get rid of B so most of us would probably go for B.

What if it was close? What would we do here? We’re interested in the behavioural side of this not the rational optimal decision. We began our research by looking at debt repayment as a goal. Here’s a stylized example. This person has $5,000 in debt over five credit cards. How is she going to repay that?

This person we assume the most indebted consumers have a goal to get out of debt. We might consider this to be an overarching or super ordinate goal but within that they have individual or sub-goals to repay each debt individually. How do sub-goals work? We know there’s a lot of research on them.

As an easy example imagine you’re running a marathon or a 10K and here we’re running a marathon. You have the long term goal, super ordinate goal of finishing the race or if you’re a better runner than I am you might have a certain time that you’re desiring but that’s a long way off. You don’t think okay I’ve only got 40 km to go. Instead you might create sub-goals.

I have a goal to finish the first mile or the first mile in five minutes. Sub-goals work because they’re closer to our proximate. By definition they’re easier to attain. We use sub-goals and we know they motivate us to achieve the super ordinate goal. One difference between debt and running is you have to run in a certain order. You can’t complete mile 31 before mile 1.

But with debt you can repay any debt in any order you want. We’re going to delineate between two types of strategies here. One we call a dispersed strategy. Dispersed strategy management is you have $500 to repay and you spread it more or less equally over different accounts. This is what a lot of research suggests that consumers tend to do.

Or we could engage in what’s called a concentrated strategy. We’re going to focus on repaying one debt as much as we can. So we have conducted a few experiments. We have some field data. We shared the field data with you first in which we looked at whether or how a concentrated strategy might motivate people to repay their debt faster than a dispersed strategy.

We got this credit card data from a wonderful organization called Hello Wallet. They’re a personal finances app. You could set goals kind of like mid.com, all these apps. What’s good about Hello Wallet and this is not a pitch for Hello Wallet but what was good about this data was they’re sold to a Fortune 500 company.

Their clients are not actually consumers. Consumers use them but they’re funded or paid by Fortune 500 companies as an employee benefit. They don’t use advertising. Our data is fairly cleaner that way. What we wanted to do with this data was first see if we can measure repayment strategy in the real world and also then to see if a concentrated strategy predicts subsequent debt repayment.

For those of you who like the Greek alphabet you’re very excited. For the rest of you I’m going to explain this in English. This is a measure of very similar of what was presented earlier. It’s a measure between 0 to 1 of the distribution or dispersion of payments. In English basically we’re going to get a value between zero and one and it’s the variance in the payments divided by the average payment.

What does that mean? Here’s a person who has two debts or sorry two debt repayments. They were paying debt A and B and they give 5 to each. That has a variance of zero because the numbers are exactly the same. We have a measure of concentration which is called zero. This is a perfectly dispersed strategy.

You can use this measure as well for those of you who work for banks to measure how concentrated or dispersed their debts are. It doesn’t have to be a debt repayment amount. It can be an actual debt number. You get the same. If we double our debt repayments to 10 we still have a perfectly dispersed strategy. We get a measure of zero.

Conversely if we have all of our repayment into one account as in example 3 then we get a value of 1 and again we can increase the amount. I have another slide which I deleted which I’m sure you can do for 3, 4, 5, 6 accounts. We can measure the extent to which individual consumers use a dispersed or concentrated strategy and of course there’s numbers between 0 and 1 that you get depending on how much you allocate to each account.

Our data we intentionally excluded people who only had one credit card. We only looked at people with multiple credit cards. We had nothing against people with one credit card. They’re fine people I’m sure but we can’t measure debt repayment strategy for a person with one credit card.

We have to have people with multiple credit cards. Here’s a dispersion of how concentrated their debts are. This is not the repayments. This is their debt. We see it as on the left hand side are people who have multiple credit cards. Our average was three and a half credit cards. These are people on the left hand side. Most people have three credit cards with reasonably equal amounts of debt on each.

There’s a bit of a bimodal distribution. People on the right hand side have one larger credit card and a few smaller ones. We’re going to look at how concentrated people’s debt repayments are. We see a real bimodal distribution. There’s actually smiling when Dilip put up his slide. I thought well, that sees the same thing I am, not exactly but as we can see there’s a large group on the left hand side who use basically a perfectly dispersed strategy.

They take their money every month and allocate it across credit cards. On the right hand side there’s a small group that do have a concentrated repayment strategy. They seem to tend to want to pay down one account at a time. We analyzed. We had thousands of customers, tens of thousands of months of data.

We took a person’s debt repayments over the month and put them all together. I want to highlight that our effects are what we call within subject for those years. Stats, people we used a fixed effects model. What that means is we’re not going to compare – so let’s say Nicole is a concentrated strategy person. Well done Nicole.

Scott is a dispersed strategy person. We need to talk. We’re not comparing whether Nicole repays her debt faster than Scott. What we’re looking at is whether Nicole repays her debt faster if she uses a slightly more concentrated versus slightly more dispersed strategy. We're looking within a person.

So to pre-empt your question, what else is predicting this, this is within person. Here’s our exciting model. I’ll just guide you through them. Basically the more concentrated a person’s debt repayment strategy in a one month, month T, the more they repaid of their debt in the following month, very exciting.

We have a bunch of controls. This is invariant to how much they’ve repaid in the past. It’s invariant to how much debt they have, how much they spend each month, the number of cards, closing an account. We have all kinds of controls here. It doesn’t matter whether their accounts were more or less concentrated, if their debt was spread out. It’s a debt repayment strategy that’s motivating greater debt repayment.

The field data and we have several experiments that show the same pattern of results. Using a more concentrated strategy within the same person led them to a greater debt repayment in the subsequent period. What explains this effect? It’s really about feelings of progress.

Imagine you’re running this marathon and every step feels pretty small in the context of 40 km or 26 miles. However for the first mile those steps feel like you’re getting closer to that first mile as compared to the 26th mile. To test this we had a bunch of scenarios. These are some stylized examples. This is going back to the example I gave you at the beginning.

Here’s a situation. You have two debts, $4,500 and $500 and imagine that you repaid a certain amount to each. How much progress do you feel you made towards repaying your debt? Here we’ve got the person who’s got a concentrated repayment of $300 into their biggest account, you know 3.5 out of 7 on a scale from 1 to 7. That’s pretty low actually.

Or you could concentrate it into the smaller repayment or smaller amount and they feel like they’ve made much more progress. That progress is motivating. If I feel like I’m getting closer to repaying my debts, again the total amount of the debt, total amount of the repayment is exactly the same. You might say Keri being a good researcher I see that in the bottom one they finish off their $200. They’re close to repaying their debt.

So we had other scenarios where they finished off at the same amount. It doesn’t matter where they end up. It’s the perception that they’re getting closer to repaying their debt. They’re making progress even though all of the other numbers are exactly the same.

It explains the effect of a concentrated debt repayment strategy on motivation to repay debt. People feel like they’re making greater progress if they’re allocating as much money as they can to one account, ideally their smallest debt account. Back to our question, which debt should you repay first?

I wish I could give you an answer. I can’t. Unfortunately our data doesn’t tell us they pay it 5% faster or 10% faster. I will say this though. The answer to this question is not obvious and it depends on how big you believe or expect the role of motivation to play in debt repayment. Thank you very much.

(Applause)