Cash in the 21st Century: Data from Federal Reserve Surveys


Jean Roark: Hello, and welcome to the Federal Reserve Bank of Atlanta's Talk About Payments webinar. Today we'll discuss cash in the 21st century. I'm Jean Roark from the Federal Reserve, and I'll be your moderator. Before turning our call over to the speakers, I'd like to run through our call logistics on slide two.

If you haven't joined us through the webinar yet, click the link you received after registering. For the best webinar experience, use the FAQ document—which can be found using the Materials button in the webinar player page—but I'll highlight a few important notes for you.

You can listen to the audio through your PC speakers, or through your phone. If you use the phone option, slides will not sync with audio unless you change your settings. You can do this by selecting the gray gear located on the upper right corner of the slide window just above the presentation. From there, you should see a few options in the media chooser and you can select the phone option. You'll only want to do this if you're consuming the audio through your phone.

You can also expand the size of the slide, and to do this you use the Maximize button in the upper right corner of the slide window located on the webinar player page. And of course, if you'd like a PDF version of today's presentation, you can access it using the Materials button.

We're also taking your questions today, and we're going to answer them at the end of our presentation, but you can certainly submit them at any time during our event. If you've joined us in the webinar, you can just use the Ask Question button on the player page and we'll get your questions queued up for our presenters today. We'll also have a polling question in the midst of our presentation, so be sure to grab your mouse when we launch that polling question.

It's now my pleasure to turn our call over to Claire Greene.

Claire Greene: Hi, everyone. Thanks for joining us for this Talk About Payments webinar today. I'm here with Oz Shy and Shaun O'Brien to talk about cash in the 21st century. Our presentation has two parts. First, we will set the stage with recent data from the Diary of Consumer Payment Choice, a nationally representative survey conducted by the Federal Reserve Banks of Atlanta, Boston, and San Francisco. Second, we will describe some research using that data, addressing the questions of how consumers get cash, how they use cash, and the potential impact of not being able to use cash.

Turning to slide 3: I'm Claire Greene, payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed. Oz Shy is a senior policy adviser and economist at the Federal Reserve Bank of Atlanta. And Shaun O'Brien is a senior policy consultant in the Cash Product Office at the Federal Reserve Bank of San Francisco. We all—together along with a team of economists, statisticians, and survey experts—work on the Diary of Consumer Payment Choice. That's the data we'll be sharing with you today.

Looking at slide 5: like me, you probably see signs like this, here, or slide 6, there—that's in Atlanta—or slide 7, seemingly everywhere (this is in San Francisco). And this anecdotal evidence makes us ask, turning to slide 8, at some time soon, will cash be just for art projects? And that's the question we're looking at today through the lens of recent consumer data.

By the way, these views are our own and may not necessarily reflect the views of any Federal Reserve Bank, or the Federal Reserve System.

We're going to look at this question via the five Ws and the how of cash—so that's who, what, when, where, why, and how of cash. Turning to slide 9, we'll look at our first W: What is happening with cash in the short and long term?

Oz Shy: So it's good to look at the origins of cash, where it all came from. The earliest massive use of cash goes back to Lydia, which today is Western Turkey on the border with Greece. In the seventh century BC, they minted a lot of coins made from electrum, which is a composition of gold, silver, and some copper. But this is all coins. We had to wait hundreds of years until paper cash appeared, and it first appears in China because coins were too heavy to carry for traders so they started issuing paper bills.

Anecdotally, the Chinese were the first to introduce paper cash, but they may be also the first one to abolish cash, as they're trying to become a cashless society. Regarding Europe, we had to wait another 800 years, and cash first was observed in Europe in 1661—and you have this picture on the right, this is the first paper cash note from the bank of Sweden.

Greene: Thanks, Oz. That was Oz Shy, economist at the Federal Reserve Bank of Atlanta. Now moving to slide 10 for a more short-term view of what's happening with cash, Shaun O'Brien from the Cash Product Office will comment on recent demand for cash.

Shaun O'Brien: So as you can see on the chart, we've seen an increase in the demand for U.S. notes, and that demand has reached nearly $1.7 trillion as of April 2019, which is about a 5 percent increase over April of 2018. One of the things that I find interesting about the increase in demand for U.S. notes is how relatively consistent the growth has been—not just in total value or in the total number of notes, but if we break out the notes by their denominations.

So here we can see the number of notes in circulation for the ones through fives, the tens through twenties, and the fifties through hundreds. And you'll really notice a relatively consistent growth of the "transactional" notes, which are those we consider to be the ones through twenties, from 1980 up until today.

The growth for what we think of as the "store-value" notes, the fifties and mostly the hundreds, had been consistent up until about 2009. And then after the financial crisis, the number of notes that the public demanded both domestically and internationally continued to increase as cash was used as a way to store assets. And thinking about the financial crisis: the increase in the growth rate really hasn't slowed down over the last decade.

Greene: So we're looking at cash for two purposes: to make a payment, which is what we're talking about today, and also as a store of value. Before we dive into the consumer data about making payments, we'd like to give you a chance to see how "like" or "not like" you are compared to other consumers. So I'll turn it over to Jean for a polling question.

Roark: All right. Thank you so much, Claire. I really appreciate that. And the participants should grab their mouse, and there should be a polling question that pops up on your screen—and it does take a few seconds for that to happen, but that question should pop up for you, and I'll read it out loud as well. The question says: Do you have any cash right now in your wallet, pocket, or your purse?

So I'm going to just stall for a second. The options are just "yes" or "no," so hopefully you see that question up on your screen, and I'm just going to pause for one second.

All right. Checking in with a couple of people, making sure that they can see that question. All right. I‘ll give you a couple of extra seconds here.

All right, at this point I am going to go ahead and stop the poll and show the results to our participants. All right. And so, Claire, just so you know—boy, we had a whole bunch of responses—84 percent of our audience said "yes" and 16 percent said "no." So I'll turn it back over to you.

Greene: Thanks. Well, I would say our audience today is quite similar to the U.S. adult population as a whole, because in our nationally representative Diary of Consumer Payment Choice, on the first night of the diary, consumers report to us whether or not they have any cash on their person, and 79 percent of respondents say "yes." So that compares to 84 percent today.

Turning to slide 11, you might keep in mind how similar or dissimilar you think you may be from the majority of U.S. consumers as we look at this data. So slide 11 is our first slide that shows the percentage share—in this case, of all payments—and today we'll be talking typically about the percentage share of all payments, or the percentage shares of bills, or the percentage shares of a person of a certain income level, just for example. This slide shows the percentage shares of all payments—that includes purchases, bills, person-to-person payments, whether made online or in person or by mail, by phone, everything—reported in the month of October in 2016, 2017 and 2018.

And moving from the left, you can see that we study the eight core—what we call the "traditional" payment instruments—and that includes paper payment instruments, so cash, checks, and money orders. It also includes cards—debit cards, credit cards, and prepaid cards—as well as electronic ways to make a payment, which the ACH experts on this call would call an ACH payment, [and] we call, for a consumer-facing view, an online banking bill payment made at your bank's website or a bank account number payment where you put in your routing number and your account number at a third-party biller (for example, your cell phone company).

So just going back to the data that's shown in this chart, you can see on the far left the depiction of the percentage shares of all the consumer payments that are made in cash, and that was 26 percent of all consumer payments by number in October 2018. That share has fallen 5 percentage points from 2016, and if you move over two sets of bars to the blue bars—credit cards—you can see that a lot of the increase in the percentage share of payments has occurred in credit cards over that time. So basically, just showing that in recent years the percentage share that all we consumers make in cash has declined.

Now we're going to move on to slide 12, and Shaun will give us some more information on what's happening with cash.

O'Brien: So as Claire has just described, between 2016 and 2018, we saw a 5 percentage point decline in cash usage. One of the important things about cash is it's used most often for small-value payments. So this chart right here is divided up really into two sections: on the left-hand side, you have 2017 and 2018 data for payments that are under $10, and then on the right-hand side, you have 2017 and 2018 data for payments $10 to less than $25.

And you'll notice that between the two years, the largest drop took place for payments that were under $10, where cash usage went from about 55 percent to 49 percent. When you contrast that with payments that were a little bit higher, you'll notice that the share of payments that took place using cash was relatively consistent. So when we think about that 5 percentage point drop, it's really centered around this declining cash usage for the payments that are under $10.

Greene: Thanks. Moving to slide 13, the takeaway for what's happening with cash—again, my own view—is that the use of cash is drifting slowly down. It is definitely not vanishing, and it could be subject to various gusts of economic winds or political or social change, and that trajectory could change. What we're seeing is that despite the many, many innovations in payments of the past decade, there is no killer app that is killing cash.

Moving on to our second W: who uses cash (slide 14)? The most important point is that everyone uses cash. We define "using cash" as carrying some, as we asked you just before, or storing some in your home, office, or car (for example), or using cash to make a payment at least once in the past year. We find that fewer than 1 percent of U.S. consumers 18 and older fail to carry or hold some cash at least once in the year, or to make a payment with cash.

Now that we've talked about everyone, we're going to break this down into different groups of people and compare how some groups are different from others. Moving to slide 15, we're first going to look at people who state a preference for using cash compared to other payment instruments.

O'Brien: So one of the questions that we asked individuals before their first day of the diary is: What type of payment instrument do you prefer to use for non-bill payments? And about 95 percent of the responses are for cash, debit cards, or credit cards. And so when we look at those individuals and then look at their actual payment behavior, what we find—really, not surprisingly—is that an individual's preference is correlated with the payment instrument that they use most often. But what is really interesting about those individuals who prefer cards, either debit or credit cards, is that they still use cash for approximately one in five transactions. So it's kind of going back to Claire's point of everybody using cash.

Greene: We've shown now that preferences are important, so in slide 16 we asked the question, are preferences changing?

O'Brien: So one of the benefits of the diary, particularly with respect to the participants, is there are almost 2,000 individuals who have taken the diary in 2016, ‘17 and '18, and so we can follow how those individuals have changed their preference over time. In 2016, we saw that about a quarter of individuals stated that they had a cash preference, and then that dropped to about 22 percent in both 2017 and 2018. In aggregate, that corresponds to an increase in those who are preferring credit cards, and this shift in preference corresponds well with the chart that Claire showed earlier, where the share of cash transactions has been declining at the same time the number of credit cards has been increasing as well.

Shy: I actually just want to add that we are mixing here demand and supply, the issues. Because as we are moving away from the recession, more people get credit cards approved, and once they get the credit card they begin to like it and then they begin to prefer it. So this slide cannot distinguish between supply-side and demand-side preference effects.

Greene: That's an important point to be making, is that we consumers aren't making this choice in a vacuum. We don't have complete control over what we do because there are the providers of credit, and also of course what the payee happens to want, which we'll be talking about as we continue on.

Looking at slide 17, say a person states a preference for cash—what happens when he or she switches?

O'Brien: So on the previous slide we saw that those preferring credit cards had increased, while those preferring cash had decreased. That chart in aggregate makes one think that individuals are switching away from cash and then transferring or stating their preference as credit, but when we look at individuals' answers over time, that's not what's taking place. Instead, there seems to be a flow of individuals from one payment instrument into another, and for this particular example we're looking at cash. So in 2016, all of these individuals had stated that they had a cash preference. And then when we look at those same individuals for 2017, we see that about 60 percent remained as a cash preference and about 21 percent had switched to debit—far more than credit cards. And we see, again, a similar pattern in 2018.

And so the shift away from preferences, from cash to credit, seems to be far more complex than individuals simply going from cash and then instantly switching the next year to credit cards.

Greene: So that's the end of our first subgroup—that is, people who prefer cash. And this is a different kind of a subgroup than the others we're going to be presenting because these are people who have stated a preference—and of course, what they say may not actually be what they do in the instance. Our next set of ways of categorizing people are going to be more objective, starting with slide 18, where we look at people by their banking status and their cash use by banking status.

So again, the same kind of a column chart where each column represents 100 percent of the payments made by the defined type of person underneath, reading from the left to the right: unbanked people, underbanked people (as defined by the FDIC, which would include taking out a payday loan, using a check cashing service), and fully banked people (also as defined by the FDIC). The first thing you can see here is that unbanked people, all the way on the left, make the majority of their payments in cash. They make 60 percent of their payments in cash. That's the big, green block.

Moving to the right, people with a bank account—both the middle column and the right-hand column—make at least half of their payments with some bank account-linked payment method. So again, that would be debit card, paper check, online banking bill payment, or bank account number payment. These people with a bank account, both underbanked and fully banked, have pretty similar shares of cash payments. The underbanked people make one-quarter of their payments in cash, and the fully banked people make one-fifth of their payments in cash.

In slide 19, we look at the use of cash by household income: from the left, the lowest household income group, earning less than $25,000 per year; to the right, the highest-income group is earning more than $125,000 household income per year. The first thing you can notice is that these green bars for cash use decline steadily as we move from the lower income group to the higher income group. And it is the case that the lowest income consumers, earning less than $25,000 a household, use cash for four in 10 of their payments.

Looking all the way over at the high-income group, you can see that the higher-income group is using credit for four in 10 of their payments. Oz has an interesting comment on the types of payments these two different groups are making.

Shy: Yes. This slide shows the volume, so let's look at high-income people. The one on the right is the $125K and more. If we dig into their transactions, we see that most of their use of cash is for very low-value transactions—between zero to $5 or between zero to $10—whereas if you move to the left to the lower-income groups, you see that people start using cash for all value transactions. So there's a difference in the composition of how cash is being used between the two income groups, the high and the low.

Greene: Exactly. So if we were to re-present this bar chart as percentage shares by value, there would be a much greater difference—the high-income group's share would be very, very small.

Moving to slide 20, many people are interested to know how a person's age might affect his or her use of cash. Millennials do use cash relatively less often than others in October 2018—these are the two bars that have the rectangle around them. An important point to remember is that millennials were ages 24 to 38 in 2018, so they are adults, many of them with student loan debt, houses, cars, children. They use cash for about 20 percent of their payments, compared, for example, to people older than 65, who use cash for one third (33 percent) of their payments.

And Shaun has done some additional analysis on the consumers younger than 25.

O'Brien: So one of the interesting things about looking at cash usage by one's age is we typically hear, as Claire had pointed out, that younger individuals tend not to use cash—or at least, that's the common misperception. And one of the interesting things when you look at individuals under 25 is the share of cash usage for them is similar to those who are 65 and older. So while we tend to think of those individuals as never using cash, that simply isn't the case. And part of that speaks to what Oz had mentioned before, that there are supply-side effects going on here where they may not have a debit card, they may not have a credit card, and so at least some portion of their cash usage has to do with just not having access to some of those payment instruments.

Greene: Moving to slide 21: I like this slide because it shows that we have new data available in the Diary of Consumer Payment Choice, which by the way is all free and online. We have categorized our respondents according to census designations of whether their location—where they live—is rural, urban, or a mixed location. Because many people do wonder, does the type of community you live in affect your payment instrument use?

A couple of interesting findings: first, that everyone—in the urban locations, rural, the mixed locations—makes about the same number of cash payments per month, 11 or 12 cash payments per month in October 2018. Where the difference in the percentage shares plays out is that people in the urban areas make more payments in the month: people in the urban areas average 45 payments per month while those in the rural areas average 38 payments per month. And that difference in the number of payments overall per month translates into rural people making about 30 percent of their payments per month in cash, and the more urban group making 25 percent of their payments in cash.

That finishes up our second W: who, who uses cash? First of all, everyone uses cash. People with a stated preference for cash use it more. Unbanked people use cash more than others. Low-income people use cash more than others. And older people use cash more than others.

Now we're going to take a break from the Ws and move to slide 22. How do we get cash? It turns out that by number of instances of getting cash, we mostly get cash from family and friends. By the total dollar value of the amount of cash we all get, we mostly get our cash from a bank account, checking or savings.

Shy: So we were really surprised when we first saw that there's a lot of cash exchange within the family, from hand to hand and within family and friends—in volume more than what people get from an ATM. We should just mention that they are all correlated because somebody has to go to an ATM to get some money to transfer within the family (or just get cash from an employer). But there's a lot of movement of money, of cash, within the family, and that's their major discovery here in this slide.

Greene: Moving to slide 23—which I guess you could say combines the idea of having fun and what is a frequent payday—we find that by value we get 40 percent of our cash on Fridays.

Shy: Yes. And the opposite, the most boring days, are Sundays, Wednesdays, and Thursdays, that people don't run to the ATM to get cash that often. And by value, Sunday is the lowest day of the week, in terms of getting cash.

Greene: Slide 24 shows 659 instances of getting cash reported in the Diary of Consumer Payment Choice in 2017, and we've chopped it at the high end to show you the three-quarters of these instances that were for amounts of $100 or less.

Shy: So this slide is extremely important because this shows us what denominations people get when they get cash, and this statement ["Mostly, we get cash in multiples of $20"]. Most ATMs give you only $20 bills, and you can see on this chart that most of the cash receipts are in multiples of $20: 20, 40, 60, 80, and 100. Now, some ATMs give you $10, and you can see that on the blue bar under the $10. This is very important because as it turns out—and you're going to see in future slides—what people have in their pocket is going to matter on their decision when to pay cash and when to pay with a card.

Greene: There's also another blue spike at $50. Maybe, Shaun, you'd like to comment on that?

O'Brien: Yes. So this is more anecdotal, but one thing we've seen from the Fed is an increased number of fifties that are being paid into circulation. With additional discussions with financial institutions, we've learned that in addition to twenties, they're also beginning to stock $50s in some of their ATMs as well. So it's just kind of interesting to see the increase in the number of ATM transactions that are centered around $50.

Shy: I just wanted to mention the median cash withdrawal was $40, and the average cash withdrawal was $100.

Greene: I think the ATM median was $100.

Shy: The ATM mean was $100—that's the average—and the median was $40.

Greene: So to summarize how we get cash–by number, mostly from friends and family; by dollar value, from our bank accounts, often on Fridays, and usually in multiples of 20.

Now we're onto our third W: where do we use cash (slide 25)? Cash is used for three-quarters of person-to-person payments, and as you can see on this slide, the use of cash for person-to-person payments is not changing very much.

The teal areas on this slide, these tiny slices, correspond to a card payment, a PayPal payment, or an account-to-account payment, all of which we consider to be the payment instruments underlying some of these 21st-century wrappers for making a payment to another person. That would include PayPal, Venmo, Zelle, Google Pay, Apple Cash, Square—all of which are funded by some underlying payment instrument. From the picture here, it looks like the teal bars are taking share away from the lighter green, which represents paper checks. In 2015, 18 percent of person-to-person payments by number were made with a paper check, compared to 10 percent in 2017.

Moving to slide 26: cash is the most-used payment instrument at the in-person retail point of sale.

O'Brien: One of the questions that we often get about cash use and the recent decline that we've seen in the share of cash use from the diary is, is this decline the result of people shopping online more often? And while we do recognize that as people begin to shop, or as people continue to increase the number of times that they shop online, the vast majority of payments take place in person. And so what this chart is showing is only those payments where the participant indicated that it was an in-person payment, and still between 2017 and 2018 we see a 4 percentage point decline. So while online shopping may reduce those opportunities, people who are making payments in person are using cash less often.

Greene: And just as a reminder, about 87 percent of purchases are in person and just 13 percent online, which is why this data that Shaun is citing is so important. Moving to slide 27: in this slide, we look at the percentage shares of payments to particular retailers or merchants, and how these different payees are paid. We find that cash is used for about one-third of payments for nonprofits, for food purchases, and for gas. So by category, those are the most popular. It is important to note that nonprofits—which consists of charities, religious groups, government—represent just 3 percent of all the payments in our data set. So while this is a large percentage share of the nonprofit payments, it's not necessarily a very big number of cash payments.

When we turn this around and we look at all the cash payments, we find that more than half of them are made in three categories: fast food, grocery, and gas. And in fact, 25 percent of all cash payments are made for fast food.

So summing up W number 3: Where do we use cash? To pay another person, when we're shopping at the in-person point of sale, and especially to buy food.

Moving to slide 28: When do we use cash? This slide points out that even in 2018, as we have previously mentioned, we do use cash for small-dollar-value payments. And you can see here the steady decline in the percentage shares of cash used as the payment amount gets larger, moving from about 50 percent of payments less than $10 to just 6 percent (by number) of payments more than $100.

The next slide, slide 29, dives more deeply into this idea of cash use for small-dollar payments. Oz used machine learning to model 14,000 in-person payments from the 2017 and 2018 Diary of Consumer Payment Choice.

Shy: So remember the slide with how much people get cash from an ATM, and we showed that most cash is with multiples of 20? Well, we fed all the data to a machine—and the data includes transactions, their values, and all demographics, age, and household income—and we asked the machine to construct a classification tree, how the machine would classify the users of cash. So the machine learning has breaking points, and as you can see the first breaking point is $10: transactions lower than $10 are in cash. Transactions above $10…well, it depends.

Then the machine's second breaking point is household income. So if you have very high household income, over 110, you just pay with a credit card, or you're most likely to pay with a credit card. If you're below 110…well, it depends again. What's the second breaking point in terms of amount? $20. That rings the bell. $20 is what people have in their pockets if they went to an ATM before they went shopping. And so if you have left an ATM, you just pull out the $20 and you pay cash. If the transaction's value is more than $20, then it depends on age and household income.

Greene: So I like this chart because it brings out a lot of the points we've raised in other discussion today: in addition to the availability of the denomination, the importance of dollar value for payment instrument choice, the importance of characteristics of each of us as consumers—for example, our age, our gender, our ethnicity—and also the critical importance of household income for payment instrument adoption and also use.

Summarizing our fourth W: When do we use cash, especially for amounts less than $10, and also in multiples of $20?

Our fifth and last W: Why might cash users switch (on slide 30)? This is an examination, also by Oz, of the potential impact of cashless retail. He looked at in-person transactions in six categories, which you can see described in the pie chart on the left. If you look at the right hand-half of that pie chart, you'll see that half of the payments he looked at are for food (again, referencing the idea that we do use cash a lot to buy food).

Then he looked at two different groups of consumers, shown on the right: first, people who do not own either a debit card or a credit card. People without either kind of card make on average nine out of 10 of their purchases in cash. If you look at the median person in that group and say, what percentage of purchases does he or she make in cash? It's 100 percent.

The second group Oz looked at were people who own both a credit card and a debit card. On average, those people make about one-third of their purchases in cash. The median person—again, the person in the middle—makes just about one-fifth of his or her purchases in cash. So this leads us to the important policy issue on slide 31.

Shy: So as more stores become cashless, that becomes a policy issue that several states and several municipalities are trying to regulate, whether to allow for stores to refuse cash. In Massachusetts, cashless stores are prohibited. I should just qualify that because there are a few exceptions: parking lots, bus drivers, and some similar services, they're allowed not to take cash. But the regulation does exist in Massachusetts.

Now, since more stores became cashless in 2019—including stadiums, sports stadiums—Philadelphia has a new rule from July that prohibits stores from not receiving cash, followed by New Jersey, San Francisco, and now there is a proposed bill in Congress.

Greene: Moving to slide 32: We can look at another important use for cash in addition to providing access, which is as something that's helpful in emergencies.

O'Brien: Yes. So anytime there is an emergency—and whether that is a forecasted event like a hurricane, or another type of event like fires or an earthquake—we see an increase in the public's demand for cash. So part of the Fed's mission is to make sure that we meet that demand, whether that's in normal times or during a crisis. And so there's always a lot of coordination between the Federal Reserve Banks, the Board of Governors, and the Cash Product Office to ensure that we are meeting our mission. In addition, we are always keeping at least one month's supply of cash payable to correspond with any types of fluctuations that the public sees with respect to contingency situations.

Greene: Right. So there is a cushion for emergencies. If you were to take a look back at the currency in circulation chart that we showed at the beginning of this presentation, you would note a tiny little spike right around the time of Y2K, when people stocked up on cash in anticipation of what might potentially go wrong with computers at the turn of the century.

O'Brien: And to Oz's point before, that was mostly in twenties, if I remember correctly.

Greene: So, for use as payments. Moving onto slide 33: In addition to full-blown emergencies, cash is also useful in the case of various glitches of modern life. For example, the Staples in Harvard Square, which had its payment system stop working just a month ago. So cash is a public good, and we all do share in some of its benefits.

On slide 34: We have been talking about cash for us as people making payments, but there also are considerations for the businesses, people, or other entities who are receiving the payments. And this picture shows a candy store in downtown Boston that prefers to receive cash, and also requires a minimum amount of payment for a credit card.

Shy: So this presentation, and the data we have, reflects the consumer side. But let's take a minute to talk about the merchant side. As it turns out, card transactions are costly, and they are more costly in the United States for merchants than in Europe. So for that reason, you see some merchants refuse to accept cash, especially for low-value transactions, especially on items that have very low profit margins. And that only has to do with the fact that merchants face card-processing fees that include also interchange fees.

Greene: Now moving to slide 35: We've been talking about data from the Survey and Diary of Consumer Payment Choice research projects of the Federal Reserve Banks of Atlanta, Boston, and San Francisco. All this data and the accompanying research reports are available free online.

Moving to slide 36: You can see references to the research we talked about today. This Talk About Payments webinar is a project of the Retail Payments Risk Forum at the Federal Reserve Bank of Atlanta. I and my colleagues there blog every Monday about payments—all kinds of payment instruments, paper, cards, digital, imagined in the future—and looking at ways to mitigate and understand risks.

Now I'm going to turn it over to Jean to manage the questions.

Roark: All right, Claire. Thank you so much. I just want to remind our participants that they can still submit their questions. We did get quite a few already, but you can use the Ask Question button if you've joined us in the webinar. Just type your question in there and we will get those questions queued up for our presenters today. So with that, Claire, I'll just turn it back over to you and you can start at the top with questions.

Claire, you may not have heard that transition or you may not have unmuted. So, we're ready for you to start with those questions if you are ready.

Greene: Okay. So I'm going to start with a question. When we think about "family and friends" movement, I assume for family it is allowance and one partner sharing with their spouse. Friends. Did you learn about why, for what purpose, friends exchange cash?

O'Brien: No. When we ask individuals in the diary who they're paying, if they indicate that they are paying a person, we ask them to break down the payment a little bit more and just say, "You paid a person, but was it for a good or service, or was it a friend or family member?" And so we don't get the particular reason why an individual got money from a friend or family member. But the reasons that were stated, at least in my opinion, are probably accurate: there's one individual in the household who tends to get cash and distributes to everybody else. But from a friend, that I don't have a good answer for.

Greene: Here's another question: How much cash is withdrawn from the point of sale as cash back, and does it mirror ATM demand patterns?

Shy: So I cannot recall offhand. I just want to tell the listener that we do have this information in the summary reports that are posted online with the Atlanta Fed. Do you remember offhand, what is the cashback fraction? I remember we do have some tables or some charts on that.

O'Brien: We do, yes. I think—if memory serves me—the number's been consistent over the last couple of years, and it hovers around $30, and that's basically just a reflection of individuals getting either $20 or $40 for cash back at the point of sale.

Greene: It was interesting for me to see, as a person who might frequently get cash back at the point of sale, that most of the activity is in ATM. There's a lot more ATM withdrawal than point of sale cash back.

O'Brien: Yes, people are not completely substituting point of sale withdrawals for ATMs.

Greene: Okay. Here is another question: Isn't it fair to assume that credit usage and/or preference will recede when the economic cycle reverses?

Shy: We don't know the answer to this question, because we are trying not to assume. We just observe the real data. But we did observe, I think, in the reports—we wrote down a few years back that there was a drop in the use of credit cards after the recession, immediately. So there's a working paper somewhere that we reported that. So you don't have to assume, I guess. This is a fact that we recorded, right after the recession

Greene: And you would find what happened in the last recession, of course, where some people had their cards canceled or some people had their card limits reduced. So there were two things going on: they had less access to credit, and also maybe they were more nervous about using credit and turned to cash as that sort of old-fashioned, envelope-budgeting technique.

Next question: At the large college campus where I work, I rarely see a student use cash. Even at bake sales, clubs are usually taking Venmo rather than cash. Has Gen Z cash usage been included in the surveys? Thanks.

O'Brien: I will plead some ignorance. Is there a definition of Generation Z?

Greene: Yes. So Gen Z are 18- to 24-year-olds—and I would just point out that not all of them are in college, and that we do see in payment instrument use that education does make a difference. Shaun, do you want to comment on that?

O'Brien: Yes. I guess the other part that we're relatively restricted, from a definitional perspective, is the diary only looks at individuals who are 18 and older. So if the generation does kind of fade into the lower age groups, we're not going to pick up that effect.

The other part, especially on the university front, is it's my understanding that a lot of times there's like a closed, prepay card for their meal plan. So they might get a certain number of meals a week, plus X amount of cash that they can spend anywhere. So it's not really an "apples to apples" with respect to the college campus, but I do know that there has been, especially for—Claire, you had pointed out before—the nonprofits, for like Boy Scouts, Girl Scouts sales, their big move beyond just taking cash.

And I don't know if that's represented over time—a change of decreased cash—but I'd be surprised if we were able to go back and see the data over time that that wouldn't take place.

Shy: Indeed, the largest group of Venmo users are college students. But we tend to record the data or to display the data by age, so that includes students and non-students; and we do use weights to make sure that the right age groups are properly represented. But we generally don't distinguish between students and non-students.

Greene: That's correct. It's also the case that there are multiple papers that look at all the demographic variables in context with house-holding—so race, education, age, gender, ethnicity, first language, whether or not the person is new to the country—and combine all those factors in modeling, and there are papers at the website that look at that.

Here's a question, I think, probably for Shaun: the latest cash orders published by the BEP [Bureau of Engraving and Printing] shows a 10-year low on cash production orders. Is this fitness, or a function of declining need?

O'Brien: I think it's less fitness and more a function of cash is remaining out in circulation longer—and I guess you could say in part that's fitness. But it also has to do with, there are a lot of supply-side effects of merchants using recyclers and smart safes (which help reduce inefficient ordering and depositing), and so you get a more efficient recycling of cash. But as the chart that we showed early on—I think it was slide 10, or maybe 9—that currency in circulation continues to increase. So the overall demand for cash continues to increase, even if the transactional use has been going down in our data.

Greene: Do you think that the rising cost of access fees, both from ATM owners and from issuing banks, causes people to use less cash? I would just point out that I am not an expert on changes in the cost of access fees. We do ask our survey respondents to rate each payment instrument on various criteria—for example, security, ease of use, convenience, and also cost. I haven't looked at ratings of the cost of using cash recently, but it certainly would be possible to do that with our data. Do you want to comment on this?

Shy: I just wonder whether the rising cost is not absorbed by the banks, because many banks allow you to withdraw cash for free, although there are rising costs of ATMs. So we don't have the data of how the cost is being shared by the financial institution that provides the cash, and the consumer.

Greene: Here is a question: When we think of the circulation of notes, is there any correlation between the migration to electronic and the number of bills held by any party? Or, said another way: Has online shopping led to a reduction in cash use?

I think that Shaun addressed that question during the presentation. Do you want to just reiterate, Shaun?

O'Brien: Yes. So if I'm understanding the question correctly, online shopping will reduce opportunities to use cash. I wouldn't say there's a one-to-one substitution between increased online shopping and the decline in cash use, because I don't know of any papers that have done any type of counterfactual estimates for whether those payments would have been conducted using cards. But we have seen cash use for in-person payments decline, so obviously it reduces the opportunity. To what magnitude that reduces the number, I don't have a good answer for you.

Greene: This is a technical question about the way the slides work. There was a question on slide 15: What does "other" mean? And there are other slides where we use the term "other" and that generally means "all the rest of the payment instruments we haven't explicitly specified in labels on the slide." So that would be, of the eight payment instruments we study, the three paper types, the three card types, and the two ways of paying electronically from your bank account.

Here's a question: Have you looked at preferences for cash by consumers, and is it influenced by income level or family of origin? For example, lower socioeconomic background despite current level of income. Also, have you looked at the preference for cash usage broken down by race, national origin or gender, to see if there are different preferences?

So the second part of that question: there are indeed different preferences, and there are papers reporting on some of those differences—which can vary depending upon whether or not you're looking at preferences for purchases or for bills, just for example. As far as I know, we do not have data about the economic status of the survey respondents' family of origin.

O'Brien: I don't believe so.

Shy: We do have a lot of demographics, so you may be able to deduce something. We have household income, we have their ethnic group, gender, we have education—but we don't record specifically the country of origin, especially the parents.

Greene: One interesting thing that people tell us is they tell us how much responsibility they have for various tasks in their household, financial tasks—for example, paying bills, making decisions about investments. That's an interesting data point.

What do you believe the impact of contactless cards will have on cash usage?

O'Brien: That depends on who you ask. I—again, reiterating Claire's statement earlier, these are our opinions—I don't know what the effect will be in the U.S. I know that in other countries, it has been quite large.

Shy: Yes—Canada.

O'Brien: I believe the United Kingdom as well. I don't know. I think about contactless from a fundamental usage perspective—how different is it to tap your card versus insert your card or swipe your card? I don't know how much the time difference will make an effect, I don't know if the action of what you need to do will make an effect. But there's evidence in other countries that it might. Oz?

Shy: Yes; I'd just like to mention that, whether they call these a swipe card, or chip card, or a contactless card is a technical issue. Those who cannot get credit cards just don't get it, and they'll have to use cash. You can also ask the question: What about Apple Pay, what about Google Pay? These are all funded by credit cards or debit cards, so you first have to get the cards and then you have these wonderful services. And if you cannot get the cards or you're unbanked, then cash is the only option—with the exception of prepaid cards that you can always buy, but prepaid cards are somewhat costly to reload.

Greene: And this I think will be our last question: Is the Fed considering polymer-based notes like other countries?

O'Brien: The reason a lot of countries switched to polymer-based notes is due to quality issues, and those quality issues usually correspond to a shorter note life. In the U.S., we do not have a problem with note life. Our cotton substrate we feel holds up quite well, and the note life calculations that we use show that the average note life for the current set of notes continues to increase.

So, not anytime soon do I think we'll be switching to a polymer-based note.

Greene: Thanks, Shaun. Thanks, Oz. Thanks, everyone, for joining us at this webinar. I'm going to turn it back to Jean.

Roark: All right. Thank you so much, Claire, and I echo your thanks to Shaun and Oz as well—and I'd like to thank our participants for joining us today as well. A survey is now available to those joining via webinar, and shortly everyone will also receive a survey link via email. Now, you only need to fill it out once, but please do take just a couple of minutes to let us know how we did.

I appreciate you joining us. This concludes today's Talk About Payments webinar. Enjoy the rest of your day.