Credit Cards

Credit Cards

Richard H. Thaler and Cass R. Sunstein discuss credit cards here using the figures, regulations, and services described in Nudge: The Final Edition. The statistics and product examples belong to that account’s period.

Credit cards serve two functions. First, they provide a mode of payment in lieu of cash. The second purpose of a credit card is to provide a ready source of liquidity if you want to spend more than you currently have in cash. Debit cards, which look just like credit cards, serve only the first function, because they are linked to a bank account and do not allow for borrowing unless linked also to a line of credit. (Warning: Some debit cards offer lines of credit at high fees. If you use a debit card to borrow, you should make sure that the fees you pay are lower than they would be with a credit card.)

Credit cards are blessedly convenient. Paying with a credit card is often faster than paying with cash and lets you avoid struggling with change; digging into your pocket to find the correct change and managing the large jar of pennies at home are vexations from which you are liberated. Not to mention the frequent-flier miles! These facts have not been lost on American consumers. In 2018, the average card user had four different cards.[6] But if you are not careful, credit cards can be addictive, and Humans often misuse them. Consider the United States figures reported in the book:

• Forty-three percent of balances are not fully paid off each month (revolvers), while 31 percent are entirely paid off (the rest have no balance or are inactive).
• Total credit card debt was $1.1 trillion in February 2020.
• The average outstanding debt balance of a U.S. household was around $6,000 in 2019 across 3.1 credit cards. Total interest payments were $121 billion with most interest rates hovering between 14 and 18 percent.
• By 2018, around 9 percent of general-purpose cardholders and about 4.5 percent of private-label cardholders had at least one severe delinquency in the preceding 12 months.
• On top of interest, users spend a lot of money on fees of some kind. Fees are about 5.5 percent of cycle-ending balances for the year, with just under half of that being late fees.

The authors connect these figures to the problems of self-control discussed in Chapter 2 of the book. Credit cards allow people to spend before they have the cash to pay. In the pre–credit card era, households were pretty much forced to use a pay-as-you-go accounting system. That is why people used Christmas clubs and jars of money labeled according to purpose or payee. Now if you don’t have the cash to fill your car up with gas, there is always your credit card. Credit cards inhibit self-control in other ways. One study by marketing professors Drazen Prelec and Duncan Simester found that people were willing to pay twice as much to bid on tickets to a basketball game if they could pay with their credit card rather than cash.[7] And there is no telling how much money people pay with the cards in order to get those precious frequent-flier miles. When the spending limit on one card is reached, there is always another card to use, or a new account can be opened using one of the solicitations that arrive announcing that you have been “preapproved.”

The traditional approach to dealing with issues like this is thorough regulation. For example, in 2009, the U.S. Congress enacted the Credit Card Accountability Responsibility and Disclosure (CARD) Act, which was designed to protect Humans from various risks and costs, including over-the-limit fees and late fees. Informed by behavioral science, the law included several nudges in the form of disclosure requirements, intended to ensure that shrouded attributes were no longer so shrouded; for example, credit card statements have to offer clear information about the consequences of making only minimum payments for lengthy periods. Certain kinds of fees were also forbidden, and in that sense, the law goes far beyond nudging. Even when it does, it is designed to protect Humans from their mistakes, on the theory that financial institutions have been exploiting behavioral biases (above all, limited attention and unrealistic optimism). The book cites research estimating that the law saved consumers about $11.9 billion annually.[8] Again, the distributional effects appear to be good: the savings are concentrated among people with poor credit ratings.

However, as with any regulation of complex products, sellers can always think of new ways to deceive buyers. We can and should use the tools of libertarian paternalism to offer consumers more help in making good decisions. Serious consideration should be given to more and better disclosure requirements, focused on protecting those at the bottom of the economic ladder. For debit cards, further steps might be taken to protect consumers from overdraft protection programs, through which they take out high-interest loans. The account describes behaviorally informed regulations from the Federal Reserve Board that prohibited banks from making overdraft protection the default option when opening a new account. That is a reasonable initiative, but as we have seen, default options are not always sticky.[9] When opening their accounts, customers are asked whether they would like the exciting option of being able to borrow money if they run short of cash. Few will inquire about the details of this seemingly generous offer—from which banks stand to make a lot of money.

At the same time, overdraft protection can be useful for some consumers. It’s convenient not to be turned away at the ATM, and the fees for bouncing checks are high. There is a hard balance to be struck here. At the very least, the Federal Reserve Board ought to consider further nudges to protect customers from being lured into enrolling in supposedly protective programs from which they lose on balance.

We think credit cards are a good domain in which to use Smart Disclosure. We suggest that credit card companies be required to post all their rules and fees in an online database much like the Mortgage File. This would allow choice engines to help people make better decisions about which card(s) to use.

Here is an example: One way credit card companies have slyly raised prices is by reducing the number of days you have between the time you get your bill and the day your payment is due. If you miss that payment, you pay not only a penalty but also interest on all the purchases you make the next month, even if you normally pay off your bill in full. For a heavy credit card user, such as a frequent business traveler, missing a single large payment by just one day can result in an extra payment of hundreds of dollars.

The bigger advantage from Smart Disclosure, however, would come from helping to change behavior. The way a consumer uses a credit card can matter more than which card is chosen. In the figures cited here, the average American household that uses credit cards has four cards and more than $6,000 in outstanding balances. That means they are paying a lot of money on both interest and fees—in many cases, much more than they need to be paying.

One illustration of the suboptimal way in which families deal with their credit card debt is how they choose to allocate the money they pay off across the various cards they have. Take the following simple example. Dan owes $2,000 on Credit Card A, which charges 18 percent interest, and $1,000 on Credit Card B, which charges 23 percent interest. He has decided that he can afford to pay $600 this month to reduce his credit card debt. The minimum payments required by the two cards are $40 and $20. For consumers facing a problem like Dan’s, how much do they pay to each card?

A team of economists has studied this question using data from the United Kingdom and later replicated the results using American data.[10] Before we get to what most people do, let’s first ask what Dan should do? For sure he should pay at least the minimum on each card, since penalties for failing to pay that tiny amount are high. After allocating amounts for those minimum payments, the best strategy for Dan is to pay all the rest of the money to the high-interest-rate card. That is a simple enough rule: pay the minimum on each card, then pay down the card with the highest interest rate. Yet the researchers find that only about 10 percent of their sample follow something approximating this rule.

What do they do instead? There are many possible heuristics people might follow, including pay the same amount on each card, but the strategy that is most common is one the researchers call “balance matching.” In our example, Dan would pay $400 to Card A and $200 to Card B. People appear to be nearly unresponsive to the differences in interest rates, which can be 6 percentage points or more. The more cards a household has, the more this mistake costs. Of course, it is also the case that the larger the balances, the greater the costs.

We should emphasize that this particular mistake is just the tip of the iceberg. There are many other forms of arbitrage families could use, the most obvious being not to run up so much debt. Other money-saving possibilities include drawing down savings to pay off more. It is not uncommon for households to have money in a checking or savings account earning essentially nothing and have large credit card debts.[11] Whether or not this is a poor choice depends on complicated mental accounting and self-control questions. Having a rainy-day mental account can be sensible, and some households use the balance limits on their cards as a self-control device to limit spending. The same kinds of issues arise in the case of borrowing elsewhere at a lower rate, such as with a home equity loan or against 401(k) assets.

Let’s just concede that if people can’t solve the easy problem of allocating their payments optimally across their various cards, they are unlikely to be getting the other stuff right, the most important of which is paying at least the minimum payment on time. This problem can be solved by setting up an auto-pay from your bank to the credit card, but only about 15 percent of cardholders use this service to pay their bills.[12] Of course, given the high fees that checking accounts charge for writing checks with insufficient funds, an auto-pay setup may not be smart for consumers who do not keep much money in their bank accounts.

We hope it is clear why we emphasized that the biggest problem with credit cards is how they are used. Being absentminded and innumerate is very costly in this domain. And while auto-pay can solve the memory problem, an even better strategy would be to turn over the management of credit card debt to someone who will both do the math and not be forgetful. As they say, there is an app for everything, and this is no exception. The book uses an app called Tally as an example.[13] (Full disclosure: Tally was started by Jason Brown, who, when he was a student at the University of Chicago Booth School of Business, took Thaler’s class. We have no financial stake in his company.)

Here is what would happen if Dan, the guy with the $3,000 owed to two credit cards, signed up for Tally: After doing a soft credit check, Tally would automatically pay off all $3,000 of Dan’s credit card debt and take responsibility for managing his accounts. Tally would then monitor all activity on those cards and Dan’s checking account, and make sure that all bills were paid on time. Most important, every month Tally would nudge Dan to pay off his debt faster by recommending a higher default payment based on his available cash and upcoming expenses. For using the service, Tally would charge Dan interest on the money it lends him, but at a lower rate than charged by the credit card.

Tally would also save Dan a lot of money by making sure that he never gets hit with a late fee or has an unpaid balance at the end of the month, because that is when using a credit card gets really expensive. If you have a zero balance and buy something for $1,000, you will not have to pay any interest to the credit card until the due date on your next bill, which can be as many as fifty-five days away. But if you paid just a penny less than the full amount some month, interest would start accruing on that $1,000 purchase starting the day you bought it. Ouch.

How can Tally pull this off? Jason is a nice enough guy, but Tally is supposed to be a business, not a charity. The answer is that credit cards charge really high interest rates even to people with pretty good ratings. This means that Tally can borrow at lower rates directly from banks on a person’s behalf and then get paid a small fee by the bank.

Even if you have never missed a credit card payment and have plenty of money in the bank, if you slip up and miss a payment, you will be charged an interest rate close to 20 percent, even in the low-rate environment described in this account. Remarkably, 20 percent of all late fee revenue comes from people with “super prime” credit,[14] because even people making good salaries can be absentminded. This is one reason that credit cards are such a profitable business. If you can afford to keep a reasonable balance in your checking account but tend to be absentminded like us, make sure you are on auto-pay.

The distinction is between choosing a product and using it. The authors argue that choosing is especially important for mortgages, while usage is especially important for credit cards. For mortgages we are hoping that Smart Disclosure and EZ loans lead to better choices, and we urge people to use auto-pay to make sure they don’t miss a mortgage payment. For credit cards, Smart Disclosure would allow choice engines to emerge as well, especially if user data were included. If you pay off your card each month, then you might care most about airline points, but if you run a balance, you should care most about interest rates and fees.

But Tally is not a choice engine. It is a user engine. Wouldn’t it be nice if there were more of those in life? Although we like Jason and wish him well, we hope his business model spawns lots of imitators and competition. Anyone who can help people reduce the costs of their credit card habit gets a thumbs-up from us. And though Make It Easy remains our mantra, one of the best ways to make it easy is to make it automatic.