Rounding Rules and Cash Inflation When We No Longer Make Cents

Rounding Rules and Cash Inflation When We No Longer Make Cents

Cash was the third-most-used payment instrument in 2024, accounting for 14 percent of all consumer payments in the United States and 21.3 percent of all payments that were made in-person (see here ). About 70 percent of cash payments were for $10 or less. In August 2025, the US Treasury ceased production of new 1-cent coins (pennies). The penny is not phased out yet and will continue to serve as a legal tender until the market runs out of pennies or until users no longer find it valuable for transactions. The first penny was minted in 1787, and the long history of the penny coin is described here .

The elimination of the penny coin has three long-term implications:

1. It eliminates the high cost of minting the penny coin (which was 3.68 cents in 2024; see here ).
2. It will reduce the burden of paying cash by reducing the number of coins that need to be exchanged between the payer and payee.
3. However, it will require rounding of cash payment amounts to their nearest zero cents, 5 cents (nickel), or 10 cents (dime).

The first item needs no explanation. The second item-no more heavy pockets or wallets full of coins-was already analyzed in this Macroblog post. Therefore, this post will focus on the third item, which is the effect of rounding payments made in cash.

Rounding rules for cash payments
The slow phase-out of the penny coin implies that sooner or later transactions will have to be rounded up or down to either zero, 5, or 10 cents. Yet, there are no official rounding rules in the United States, although H.R.3761, a bill that would allow rounding, was introduced in 1989. This post focuses on "symmetric" rounding (defined below). Other possible rounding methods (upward and downward) are defined and analyzed in the full report .

The "symmetric" rounding rule implies that: (i) cash payments ending with 1 and 2 pennies are rounded down to 0 pennies; (ii) cash payments ending with 3 and 4 pennies are rounded up to 5 pennies; (iii) cash payments ending with 6 and 7 pennies are rounded down to 5 pennies; and (iv) cash payments ending with 8 and 9 pennies are rounded up to 10 pennies.

The symmetric rounding rule was adopted in Canada in 2012 after the Royal Canadian Mint stopped supplying new penny coins (see here ). The symmetric rounding rule was also adopted by the European Central Bank (see here ), even though the 1 euro-cent is still in circulation. This rule was adopted because several euro zone countries already allow rounding of the 1 euro-cent and 2 euro-cents to their nearest zero, five, or ten euro-cents. The list of countries includes Belgium, Estonia, Finland, Ireland, Italy, Lithuania, the Netherlands, and Slovakia.

Data and preliminary findings
The data are taken from the 2024 Survey and Diary of Consumer Payment Choice (SDCPC). The SDCPC is a sample of US consumers age 18 and older and conducted yearly each October. The data are publicly available and can be downloaded from here.

The 2024 SDCPC data contain 3,896 in-person cash payments conducted by 1,994 respondents who made cash payments. Therefore, on average, each cash-using respondent has made 1.95 (or almost two) cash payments during three consecutive days in the month of October 2024.

Perhaps the most important finding from the data with respect to rounding is that survey respondents reported that 72 percent of their cash payments already end with 0 (zero) pennies. Clearly, these payments will not be affected by the elimination of the penny because even now pennies are not used for most cash payments. Figure 1 shows the percentage of cash payments that end with zero and five cents.

In figure 1, long bars (light color) and short bars (dark color) measure percentage of cash payments ending with zero and five cents, respectively. Figure 1 implies that:

1. The percentage of cash payments ending with zero pennies increases with the value of the cash payment.
2. The percentage of payments ending with five cents declines with the cash payment value.

Observation (a) reveals that payers and payees care less about pennies when they transact with large amounts. Therefore, both parties to a transaction intentionally disregard the penny component of large cash payments. Observation (b) follows from observation (a) because large payment amounts with five-cent endings are also rounded to end with zero pennies.

The observation from the data-that 72 percent of all cash payments already end with 0 (zero) cents-is rather surprising and therefore requires an explanation. Three reasons possibly explain why our survey respondents recorded these payment amounts:

1. Some consumers might be billed for payment amounts that end exactly with zero pennies. For example, payments for partial bills such as installment payments (instead of the full amount) are generally rounded and do not involve any pennies.
2. Payees (merchants or sellers) want to reduce their transaction burden of handling pennies and therefore ask payers (buyers) to disregard the penny component. Alternatively, some consumers who do not have pennies round up the payment amount to the nearest nickel or dime.
3. Survey respondents disregard the survey instructions and round the penny component to zero.

There is no way of knowing which of the above three reasons causes a cash payment to end with zero pennies. All that we observe from the data are the payment amounts that end with zero pennies. Still, these observations might be telling us that the transacting parties do not care much about a few pennies more or a few pennies less for completing a cash transaction, especially for large dollar transactions.

Empirical analysis of rounding cash payments
Suppose that the penny coin is removed from circulation. It is important to emphasize that rounding applies to the total payment amount and not to individual prices! That is, payments are generally made for multiple items (not just one item) as figure 2 illustrates.

Therefore, the use of the terms "inflationary" or "deflationary" impacts of rounding apply to the final total payment amount (not to individual prices). This usage is different from the common use of the term "inflation," which applies to individual prices (or to a weighted average of individual prices). Also, note that rounding applies only to cash payments; that is, electronic payments could continue to be denominated in pennies (or even fractions of pennies, as commonly observed in gas stations). Table 1 displays the consequences of rounding for cash payments.

The computations in table 1 are based on two sets of data:

• The middle column corresponds to all 3,896 cash payment observations. Because most of these payments end with either zero or 5 cents, this dataset will be used to compute low bounds on the level of rounding and the cash inflationary impact of rounding.
• The column on the right corresponds to the 881 payment observations that do not end with either 0 or 5 cents. All these payments will be rounded once the penny is phased out. Therefore, the computations based on the restricted data provide high bounds on the level of rounding and the cash inflationary impact.

Rows 1 and 2 in table 1 display the average and median cash payment amounts in the two datasets. Row 3 shows that the average number of pennies that are rounded lies between 0.043 of a penny to 0.192 of a penny. Row 4 shows that the median rounding adds between 0 to 1 penny to the cash payment amount.

Finally, the cash inflationary impact (row 5 in table 1) is computed by averaging all the ratios of the number of pennies rounded in each cash transaction to the payment value of this cash transaction (denominated in pennies). Thus, as expected, the cash inflationary impact is negligible as it lies between 0.001 of a percent to 0.01 of a percent. In fact, the cash inflationary impact is not statistically different from zero under symmetric rounding.

What comes next? Some thoughts about nickel elimination
Currently, there are no plans to eliminate the nickel (the 5-penny coin) although at least one bill has already been introduced in Congress: H.R.1270 (2025). But, in the next few years, inflation will erode the real value of the nickel, which could necessitate the elimination of the nickel coin.

There are two reasons why the elimination of the nickel coin (after the penny is no longer in circulation) will be significantly more complicated than the elimination of the penny only:

1. Payments ending with exactly 5 cents will have to be rounded either upward to 10 cents (a dime), or downward to 0 (zero) cents. Therefore, symmetric rounding will not be feasible for payments that end exactly with 5 cents.
2. For some payment amounts, the elimination of the nickel will increase the burden of paying cash.

To demonstrate the first complication, how should a cash payment of $87.25 be rounded? Up to $87.30 or down to $87.20? Rounding up will put the burden on the payer (consumers, buyers) whereas rounding down will shift the burden to the payee (merchants, sellers).

To understand the second complication, first note that the rounding of payments to 0 cents or 10 cents will reduce or eliminate the need for the 25-cent (quarter) coin, which is now heavily used both in the US and Canada. Second, as table 2 shows, the United States and Canada are missing a 20-cent coin, which is used extensively in the euro area.

Without a 20-cent coin, elimination of the nickel would increase the burden of paying with coins for some transaction values. For example, to currently make a 30-cent payment, the payer can use one 25-cent (quarter) coin and one 5-penny (nickel) coin. Thus, the burden is to exchange two coins. However, rounding to the nearest zero or dime after the nickel is eliminated would require three coins (three dimes) for making a 30-cents payment. Thus, the burden of paying cash would increase from two coins (1 quarter and 1 nickel) to three coins (three dimes). In the euro area, a payment of 30 cents would require only two coins (one 20-cent coin and one 10-cent coin).

Final remarks
The most difficult part of this type of research is to obtain accurate data on the exact amount of cash payments. Scanner data (data recorded on cash registers) would not provide accurate information because these data show what consumers should have paid and not what they actually paid if the final payment amount was rounded to the nearest nickel or dime. In the other extreme, consumer survey data, such as the data presented in this blog post, might not properly distinguish between the amount printed on the receipt and the amount actually paid. Furthermore, respondents who take payment surveys might choose to round the small penny amount, despite being asked not to do so.

Therefore, the only way to obtain completely accurate data on actual cash payment amounts is to physically observe the consumer and the seller at the point of sale and to record the actual banknotes and coins that are exchanged in both directions (initial payment and the amount of change received). Unfortunately, this type of data collection is very costly.

Author's note: I would like to thank Tom Heintjes for most valuable comments and suggestions on earlier drafts and Whitney Strifler for the interactive charts.