Tom Heintjes: Hello, and welcome to another episode of the Economy Matters podcast. I'm Tom Heintjes, managing editor of the Atlanta Fed's digital magazine Economy Matters, and today we're sitting down with Claire Greene, a payments risk expert in the Bank's Retail Payments Risk Forum. Today we're going to discuss recent findings from the Survey and Diary of Consumer Payment Choice, which are tools that help us to understand how the payments system is evolving. And as anyone who has studied the Federal Reserve knows, we are deeply involved in the US payments system, so a thorough understanding of changes in how people pay for goods or services—or anything, really—is important. So without further ado, let's say "hi" to Claire Greene. Claire, thanks for being with us today. It's great to have you back on the podcast. It's been a while!
Claire Greene: Yes, it has, and it's a pleasure to be here, Tom—thanks very much. As a regular reader of Economy Matters, I'm really glad to participate.
The Atlanta Fed's Claire Greene. Photo by Peter Davis of the Boston Fed
Heintjes: Well, that's always very kind of you to say. Claire, let me start off by asking you: What are the biggest changes you noted from the last Survey and Diary of Consumer Payment Choice to this one?
Greene: So, for that we're talking about from 2021 to 2022—and the most notable thing is that things didn't change that much. Payments are a very strongly ingrained habit, and here are some of the things that didn't change—these facts are the same in both years. Two-thirds of people paid electronically from a bank account, 85 percent, which might surprise you, used cash at least once in the past 30 days. The share of purchases made remotely was the same as in 2021, and also the same as in 2020—about 20 percent. And even crypto asset ownership didn't change between 2021 and 2022. Probably, as you found, your personal circumstances didn't change a whole lot from 2021 compared to '22, compared to maybe from 2019 to 2020, and that's reflected in this data.
Heintjes: Right. I'm going to take a deeper dive into some of what you just talked about, but before we begin talking about payments and the changes you've seen, let's talk a minute about the Survey and Diary of Consumer Payment Choice. What are they, and what do they measure specifically?
Greene: These are quantitative, nationally representative surveys of US consumers age 18 and older, and I can organize our inquiry around three concepts. We're going to be talking about a three-by-three matrix here—three things we want to know, and three attributes of each of those three things. The first is we look at the kinds of payment instruments—and by that, we mean three types: paper, cards, and electronic. So three kinds of cards: credit, debt, and prepaid. Three types of paper payment instruments: cash, paper checks, and also money orders. And then the electronic ways to pay directly from a bank account, which payments experts would call ACH payments. And for our consumers, we refer to as either using online banking bill pay, or maybe putting your account number in at a website as a way to pay. So that's the first thing we look at: the kinds of payment instruments. Then, the second thing is what we think and do, related to those payment instruments. And there we also have three broad questions regarding what we think about these payment instruments: how we assess them for characteristics such as convenience, cost and security, whether or not we have adopted them—that is, do we own them or have we got them set up, ready to use. In other words, is there cash in our wallet? Do we have a debit card? And then, how we use them in particular situations? So that's the three things related to what we think and do. And then, the third point is transaction types, of which there also are three: bills, purchases, and person-to-person payments.
Heintjes: I've been referring to the Survey and Diary of Consumer Payment Choice as one thing—but they're actually two things, aren't they? And what are the differences between them?
Greene: They really are one thing, and the reason for the two names has a lot to do with the history of how they were set up. Here's basically how these survey instruments work. First of all, they're both surveys. A diary is a particular kind of survey that involves daily recording of what you do. People take what we call the survey on the first day of their survey taking, and there they answer questions that are not about recording—they're more recall: Do you have a credit card? What do you think about ways to pay with cash? Are you the person in your household most responsible for making financial choices? Snapshot-style questions, we can call those, about your financial situation and the payment instruments you own and what you think about them. Then, we have a three-day diary that follows immediately upon those survey questions. A diary is different because it involves very specific recording of exactly what you do. In contrast, for a survey I might say, "Tell me about how many purchases you made with cash in the last month." Whereas for the diary, at the end of your first day I'll say to you, "Did you make any payments today?" And then if you tell me "yes," I will ask you to record the specific details of that exact payment: the day, the date, the time of day, the payment instrument you used, how much you paid, the merchant or payee, whether or not you made that payment in person, and also whether or not you used some sort of device to make the payment. So you can see—as opposed to me learning, "I make 30 payments per month for purchases," to have that very specific data about one particular payment gives a lot of insights into the way we make our choices and why.
Heintjes: Yes, I'm going to follow up on the diary. How reliant is the diary on people not forgetting that they stopped to fill up the car, stopped on the way home for a gallon of milk, etc.? I don't know that I could keep all that in mind, day by day.
Greene: You're right. It's extremely reliant on memory, and it certainly is the case that the reason the diary was introduced, starting in around 2015, was to benchmark the survey data, where we were asking people about what they do in a particular month. One thing we found, just for example: when we were testing the diary with friends and colleagues, I asked a friend of mine what he had learned from taking the diary. What he said was, "Well, I thought I bought a candy bar once in a while for a treat, but it turns out I buy a candy bar every day." That is the advantage of the recording method of the diary over the recall method of the survey. Now, despite that, you still have to remember what you did on the particular day in order to record it. We have recently, as a result of some survey experiments, added a question to the diary. So you'll be entering your payments for three days in a row in the month of October. When you log on the second day, you will be offered a list of the payments you entered on the first day and then offered the question, "Would you like to enter anything else?" And what we found in our first attempt at using this method is that it did add approximately two payments per day to the number of payments that people were recording. So survey research is a constant tug of war between adding more questions—which means then, of course, the cost of the survey increases because we pay our vendor by minutes of question time—and then secondly, we've added potentially cognitive burden to the respondent. We want to keep that to a minimum, because our primary goal is to get a good count of the number and dollar value of payments. So this appears to be a follow-up question that is definitely worth the time, and asking the respondent to do that.
Heintjes: So, you're not only learning more about payments, but about human nature and the nature of memory itself, interesting. Claire, how long have you been conducting the survey, and asking people to keep a payments diary? The results from the first one must look like they came from another planet compared to what we see today.
Greene: You're right, there have been really a lot of changes—big changes—over this time period. The survey has been around since 2009, with a pilot in 2008, and the diary since 2015, with a pilot in 2012. I did look back at some of the 2009 data for comparison, so here are a couple of fun facts. In 2009, 30 percent of US consumers reported that they had some sort of online payment account—which, basically back then was pretty much limited to PayPal. In 2022, 66 percent of consumers have that kind of an account, so that's really a big change—as a percentage share, more than double. Then in 2009, about half of all the payments that people recorded were paid with cards. In 2022, that's up to almost two-thirds, so another big change: the continuing growth of payment cards.
Heintjes: How much was the pandemic behind any of these changes, as far as you can infer from these results? Did COVID accelerate—or, conversely, slow down—any of these trends? And if we compare the most recent data to right before the pandemic, are there any big changes—are you able to tease out the effects of COVID on payment trends?
Greene: We definitely saw enormous change during COVID—keeping in mind that the word "enormous" in the context of payments doesn't necessarily mean "enormous" in other contexts, because payments is a habit and very slow to change. For payments, during COVID we're talking about a trickle-down effect of more people working from home more often, and also our general reluctance to engage with people in person—whether that's our reluctance to hand a friend a $20 bill, or our reluctance to go into a store and pass our credit card over to a clerk. So just picking out a couple of examples, going back to the online payment accounts, 66 percent adoption of those accounts in 2022—that's up from 54 percent in 2019. And the really big jump in adoption there occurred in 2020, and also 2021. From 2021 to 2022, the share of consumers with those accounts is stable. So, a lot of change happening during COVID and now sort of consolidating—people not necessarily falling back into their old habits, so there's definitely some stickiness here. We all learned a lot, right? We learned how to tap and pay, we learned how to put six payment apps on our phones, and that learning has stuck with us postpandemic.
Heintjes: You referred to proliferating payment methods—how many payments a month do people make on average, and how has this number changed? I imagine it's gone up with the greater number of ways to pay for things nowadays, right?
Greene: I think the number of payments we make per month is not so much a function of the ways, it's more about the whys. So just for example, thinking about the bill payments you pay—that's going to be pretty constant from one month to the next, and even from one year to the next, in terms of the number. But our number of purchases could vary quite a bit with our personal situation—or even, for example, during COVID, just how likely we were to leave the house. So the total number of payments in 2022 is back to its pre-COVID, 2019 level—which is 39, and that number did drop some in 2020 and 2021. So it dropped in those peak COVID years.
Heintjes: I want to touch on payment cards, including debit, credit, or even prepaid cards. They seem to be everywhere these days. How does ownership of prepaid cards compare with the rate of having a bank account? After all, you don't really need a bank account to own a prepaid card.
Greene: This is a really good question, because you're right—a prepaid card can be a substitute for a bank account. We can use it to pay, sometimes you can use it to get your wages onto a prepaid card, you can store money there. One thing you can't do is earn interest, as far as I know. So 94 percent of US consumers have a bank account, and 60 percent have some sort of a prepaid card. What's important to remember about prepaid cards is that they come in all sorts of different flavors, with many different uses. So you might have a coffee card with money stored on it for which you get a free coffee now and then, as opposed to a more general-purpose prepaid card that you can use anywhere. It would have a logo on it from one of the card networks, and you might be able to receive income on that card, and if you manage it carefully, you could register it online to obtain pass-through deposit insurance. Now, another thing that is happening with prepaid cards is that the definition of "card" itself is getting kind of confusing, because now we can store money in places and make payments without necessarily having a physical piece of plastic.
Heintjes: Well, I'm going to turn to credit cards, as opposed to prepaid cards. Of course, credit cards have been popular for years. What trends do you see in the diary and survey about credit card trends, and have trends on people's unpaid balances changed in recent years? I know you've done other research on unpaid balances on credit cards. What do you see there?
Greene: You're right—that was another big impact during COVID. We saw the result of the Economic Impact Payments and also the increased unemployment benefits under the CARES Act. So like a lot of research into consumers, these consumer surveys do find a drop in the shares of consumers revolving on their credit cards during COVID—and the people during COVID who did revolve had lower unpaid balances. One thing that I think is interesting about that is that the shares of people carrying an unpaid balance has stayed pretty constant since 2020: about 41 percent, 42 percent. That's down from 47 percent in 2019, but the difference is that revolvers in 2022 have a much higher balance than in 2020—almost three times higher—$2,700 is the median balance of a credit card revolver for 2022.
Heintjes: Claire, before we go any further let's pause to listen to this important message from the Atlanta Fed.
[advertisement]
Heintjes: Welcome back to this episode of the Economy Matters podcast, where we're talking with Claire Greene about recent results from the Fed's new payments study. Let's talk about cold, hard cash. Is cash still king? I see people in line in front of me using a card to pay for, you know, a pack of gum. What do you see are trends in cash?
Greene: I'm going to say that cash is still king.
Heintjes: OK.
Greene: Eighty-three percent of consumers used cash at least once in the prior 30 days in 2022. That's a bigger share than any other payment instrument—bigger than credit cards, bigger than debit cards, bigger than some electronic way to pay from your bank account, certainly bigger than a paper check. And when we look at purchases only, even including online purchases, one in five purchases is made in cash. I think one thing that's good to remember when you're standing in line and watching everybody pay for that piece of gum with a card, is that payments behavior really varies a lot among people at different income levels and different demographic groups, and dependent upon if people live in a more urban or rural environment. And so, a primary rule—especially when we're talking about something that affects all of us, every day—is don't use yourself for survey research. We might not be normal.
Heintjes: Well, I try not to be judgmental, but I find myself thinking, "Do you not have a $1 bill?"
Greene: That's funny.
Heintjes: So, you mentioned checks, and I want to ask you about the poor, beleaguered check. Has checks' decline leveled off, or are they still trending down?
Greene: In our data they appear to be continuing to decline, but not any more rapidly. In the Federal Reserve Payments Study
, there were an estimated 11 billion checks written in 2021. So that's not a small number—the total dollar value is about $27 trillion. In our data, we see about 46 percent of people who reported using a check in the prior 30 days. And it does occur sometimes that when you're considering what the payee wants to receive and how the payer wants to pay, that a check can, in the end, be a convenient payment method of last resort.
Heintjes: Let me ask you about the payment approach called "buy now, pay later." I've never used it, but I've heard of it. How is this payment method—or program, or whatever you want to call it—different from using credit? Is it becoming a mainstream trend?
Greene: Well, that's a good way of phrasing the question: trying to figure out just what to call this. In our survey, buy now, pay later is not a payment instrument. When you sign up for buy now, pay later, you commit to a sequence of withdrawals either from your bank account via an ACH payment or via a debit or credit card payment. So those are the payments. The marketing premise is that when you use buy now, pay later, a credit check is not required and choosing this method won't affect your credit score. So that makes it attractive perhaps for people who are maxed out on credit cards, or who for other reasons don't want to use their credit card, or who don't have a credit card. I guess you could call it a mainstream payment method, because three quarters of consumers have heard of it. We do find, however, that like you, not that many of our respondents reported using it in the immediate past, in the prior 30 days. Just about 6 percent of consumers said they use it, and of those consumers it was most common for them to use it only once. I don't know about you, but I get offered it all the time—even for groceries sometimes.
Heintjes: Yes. Well, you know, when I first heard of it, I thought it was kind of like the old layaway—you know?—but I guess it's distinct from what we used to call layaway because you get it. You don't have to wait to obtain the good.
Greene: Exactly. It's layaway in reverse.
Heintjes: Let's turn our attention to mobile payments. I know they're not new anymore, but the method is newer than credit and debit cards, to be sure. What did the survey and diary tell us about trends in mobile payments?
Greene: So just like buy now, pay later, mobile pay is not a payment instrument. Let's call it a "user experience wrapper" that goes around usually your credit card, your debit card, perhaps an ACH payment. And in mobile, we see another really big pandemic effect. In 2019, 38 percent of consumers said they'd made a mobile payment at least once in the prior 12 months. In 2022, that 38 percent had jumped up to 63 percent. So again, we all learned a lot during COVID. We consumers learned how to make mobile payments, but also merchants learned how to offer mobile payments, how to activate that capability in their payment terminals, and how to train their employees to accept mobile payments. So, it's a bit of a complicated ecosystem—people on both sides of the transaction have to be prepared to either use or accept that payment type.
Heintjes: Claire, when it comes to the payment method people use, are there differences between paying bills and making purchases that you see?
Greene: Yes, that's huge. There are two aspects, I think, affecting these differences. The first is paying remotely versus in person, and then the second is the relative power of the payer and the payee. So first, most purchases are in person—about 80 percent—and most bills are paid remotely, including, of course, by mailing a paper check—85 percent. So that's pretty much reversed: 80 percent in person for the purchases, 85 percent remotely for the bills. And then for purchases, the payee is still in the process of angling for your money—the sale is not closed. So while it's not that usual to have constraints on payment choice for purchases—especially when you're talking about in-person as opposed to online—for bills, however, the payee has more control over requiring payment in certain ways. Just for example, you can't pay your mortgage with a credit card, no matter how much all of us would love to get points for paying our mortgages. And maybe if you're paying your taxes to your local government and you want to use a debit card, there might be an added fee. But if you set it up with your bank account routing number, so it can be executed via an ACH transaction, that payment would become free. So there are all different ways for payees to either encourage us with discounts or fees to use a particular payment method or just to require that we not use some way to pay that's more expensive for the receiving payee, typically. So, that's kind of the framework for purchases versus bills. The way that plays out in payment choice is that for purchases, cards rule: 71 percent of purchases are made with cards, considering all purchases, in person and remote, and the remaining about 30 percent via other methods, including 21 percent in cash. For bills, it's just the reverse: 30 percent of bill payments are made with cards, and the remaining 70 percent via other methods, including half via electronic payment method, straight from your bank account.
Heintjes: Interesting. Claire, earlier in our conversation you referred to crypto, and I wonder how much our conversation today needs to touch on crypto assets. Crypto falls into that large category of things I'm aware of but have never used. I know some people use them to pay for things, but do the survey and diary capture the use of crypto?
Greene: Well, first of all—I'm glad to hear you use the term "crypto assets," because I myself am not convinced that these are truly payment instruments. Just because someone back in 2009 called crypto a currency doesn't really make it one. I consider them to be speculative investments. We do ask our consumers whether or not they own crypto, and when they do we ask them why and also whether or not they use them to make a payment. So the first thing is it is very unusual for people to report that they've used crypto to make a payment. Second, we tend to see that changes in the market value of cryptocurrencies affects the shares of consumers that we see reporting that they own some. So, for example, in 2020 about 4 percent of consumers reported they own some crypto, and then there was a lot of price appreciation, particularly in Bitcoin. In 2021, we had more than double that—9 percent—of consumers saying they own some crypto. Then, right around the time that our survey was in the field in fall 2022, there was a kind of crypto price crash, so after that big jump from 2020 to 2021, the shares of consumers saying they own crypto was flat—it stayed at 9 percent. And another thing related, I think, to that price change is in '21, and some earlier years, it was most common for people to report that they own crypto for investment. I think in 2022, perhaps you start seeing a little bit of buyers' remorse, maybe, in their responses The most common answer in 2022 was "I'm curious about technology." So, "I'm not doing this maybe to experience a big return, but I'm doing it as an investment in myself, in my own learning."
Heintjes: Do you and your Fed colleagues tweak the survey to improve its accuracy, and do those tweaks affect the resulting data? In other words, comparing over years, does that get difficult as you change the survey itself?
Greene: We have no choice but to tweak the survey to improve its accuracy, just looking at the sorts of changes that have happened over the last 10 or more years. And these tweaks certainly do affect comparability—this is very challenging. Our key goal, again, is to get the overall number and value of payments right, and then the percentage shares of payments made by the different payment instruments—because that changing mix of the way we consumers pay is a primary interest of ours. So, you're right: sometimes you'll see an asterisk in the survey tables, or you might see two rows related to a particular variable where we report how we've asked a question, say, for the first seven years, and then how it's different more recently, and perhaps some commentary in the footnotes on how you can associate those different data points.
Heintjes: Right. Claire, this is all very interesting, but I have to ask you: what do we do with the results? As I noted at the top of the episode, the Fed is a big player in the overall payments ecosystem, so how does this information affect what we do?
Greene: This data is important to support the Fed's mission and operations, because policymakers and payments operators need to understand the impact of this kind of rapid innovation and change. So first of all, just talking about the Fed's mission: as part of the mission to ensure full employment and price stability, the Fed is committed to encouraging an economy that works for all. And therefore, we need to understand the impact of innovation and change for public policy, for financial education, and for consumer protection. Now, a traditional view of the payments system is that it's the plumbing that connects financial institutions. The main actors are financial institutions, card networks, and payment services providers. The newer view of the payments system has changed so that the end users—consumers, businesses, state and local governments, nonprofits, et cetera—also should have a voice. In our survey, for all the participants there is an accompanying set of income, demographics, and household data. So, it's possible to examine the types of people who are more or less affected—either for good or for ill—as the payments system changes. And then, having all this data leads to informed choices and informed decision-making.
Then the second aspect is Fed operations, and as you mentioned, the Fed is a very important provider in the payments system. So I'll just look at two ends of the method spectrum right now: the very oldest, and the very newest. Cash, of course, is the OG fast payment method: I just hand it to you, and you've got it. And Federal Reserve Cash Services is just like any other kind of factory: the cash comes in in little trucks, and there's a receiving dock, and then the cash gets itself onto conveyor belts, there's sensors to examine the bill, there's robotic assists, because the cash gets heavy. So it's important for that factory to keep up to date and to anticipate demand in order to plan for labor and capital investment. So that's one end of the Fed operations. All the way at the other end, we have the new payment rails like the clearinghouses, real-time payments [RTP] and FedNow. These new ways to pay could lead us to substitute away from the existing ways to pay, and perhaps influence, just for example, the Fed's investment in check image processing. So RTP and FedNow are instant, bank-to-bank payment methods. You can think of them as platforms that products can be built on top of. We consumers aren't going to have a direct interaction with these instant payment rails, just like we don't have a direct interaction with the ACH system. And these rails, you can think of like your phone operating system. Just like apps can be built on your phone operating system, new products can be built on the rails—and that is potentially going to create, from everything I'm reading, a huge surge in innovation over the next few years. So, understanding that adjustment as it occurs will be important, again, for two things: public policy, and also planning for the Fed's business units.
Heintjes: I'm glad you mentioned FedNow, which debuted in July. So, there's a little plug for our new service, and we'll have a link to FedNow on our website—so thank you for mentioning that. Claire, this has been a really fascinating conversation. I want to thank you for spending some time with us. I know the Retail Payments Risk Forum is always keeping up with changes in payments, as well as other important related matters, so we'll have to have you back on soon.
Greene: That would be fun—I'd love it. Thanks, Tom.
Heintjes: And that brings us to the end of another episode of the Economy Matters podcast. I would like to thank you for spending time with us today, and I hope you'll join us next month when I'll sit down with Atlanta Fed economist Julie Hotchkiss to discuss her recent research into how the rising popularity of electric vehicles has affected the collection of gasoline taxes devoted to road building and maintenance, and the various possible ways to address the matter. It's a timely topic, and I hope you'll be here for it. Thanks again, and let's meet again next month.