The Hidden Job of Cash in a Digital Economy
This series examines the resilience of retail payment methods when the infrastructure behind digital transactions fails. This opening chapter focuses on the role cash can play when electricity, communications, or payment networks are disrupted, so the reader can distinguish normal-time convenience from the ability to keep transacting under stress. The central question is not whether cash should replace cards and digital wallets. It is why a payment method used less often in ordinary life may still matter to the resilience of the whole system.
Cash is not one thing
The simplest way to judge cash is to count how often people use it. That measure matters, but it misses the difference between a payment method's everyday volume and its value when alternatives stop working.
The Federal Reserve's 2026 Diary of Consumer Payment Choice offers a useful baseline. In 2025, U.S. consumers made an average of 47 payments per month. Six were made with cash, or 14 percent of the total, leaving cash behind credit and debit cards in transaction volume. It would be easy to read that number as evidence of cash's steady disappearance. The rest of the Diary points to a more complicated conclusion. Four out of five consumers used cash at least once in the previous 30 days, 90 percent said they planned to keep using it, and 76 percent carried cash on their person. Forty-five percent stored cash somewhere else, with average store-of-value holdings of $364.
These numbers describe several different relationships to the same instrument. Someone may use cash for a small purchase, prefer a debit card for most in-person transactions, carry notes in case a card fails, or keep money at home for emergencies. Those are not interchangeable behaviors. A falling share of routine payments can coexist with a persistent need for a backup and a reserve.
Preference data makes that distinction visible. In 2025, 16 percent of consumers said cash was their preferred method for in-person payments, compared with 38 percent for credit cards and 40 percent for debit cards. Yet cash transactions were not limited to people who preferred cash. About 64 percent of in-person cash payments came from people who preferred debit or credit cards. In other words, many people use cash without regarding it as their first choice. They may turn to it because the transaction is small, the merchant accepts it, budgeting is easier, or their preferred electronic method is unavailable.
That fallback role is easy to overlook because it appears only when a choice is constrained. A person who normally pays by card may have no strong reason to mention cash until a terminal is offline or an authorization request cannot be completed. In ordinary statistics, the backup looks like a minority preference. In a disruption, it can become the difference between completing a purchase and postponing it.
Cash also remains unevenly distributed. Older consumers made an average of 10 cash payments per month in 2025, compared with two among consumers aged 18 to 24. Rural residents made nine, compared with six in urban and suburban areas. Five of every six cash payments were for transactions under $25. These are descriptive differences, not proof that age, income, or geography alone causes cash use. They do show why an average payment share cannot stand in for every household's exposure. A system that works for a digitally well-served urban consumer may impose different costs on a rural household, a lower-income household, or someone making many small in-person purchases.
The Diary therefore establishes a starting point rather than a crisis result. Cash remains part of the payment portfolio for ordinary reasons, and it is held by people who may not describe it as their preferred instrument. The more demanding question is what happens when the digital route is no longer available.
Resilience begins where convenience ends
Digital payments are powerful because they make authorization, record-keeping, remote commerce, and rapid transfer convenient. They are also dependent on a chain of conditions. Electricity must power terminals and communications equipment. Telecommunications must connect the merchant, customer, and payment network. Data and authorization systems must be reachable. The store itself must be open, and the customer must be able to reach it.
Cash has a different point-of-sale dependency profile. A note can settle a transaction without a live authorization request or the same connection to a card network. That difference does not make cash disruption-proof. People need access to cash, merchants need to accept it, stores need to operate, and physical supplies still need to reach the area. But the failure of a communications or authorization network does not automatically invalidate a note already available to the buyer and acceptable to the seller.
This is the difference between adoption and resilience. Adoption asks which method is convenient and widely used when the system is functioning normally. Resilience asks which transactions remain possible when one or more parts of that system fail. A payment system can perform well on the first question and poorly on the second. The two measures are related, but neither substitutes for the other.
A disaster makes the distinction concrete. A hurricane can close stores, damage roads, interrupt electricity, reduce mobility, and disrupt payment networks at the same time. If spending falls afterward, the decline cannot automatically be assigned to failed cards. People may be unable to reach a store, the store may be closed, or needed goods may be unavailable. Yet payment failure can add a separate constraint: even where a store is open and a needed item is on the shelf, a customer may be unable to complete the purchase through their usual electronic method.
The resilience study On the Resilience of Payment Methods treats this separation as an empirical question. Its evidence combines storm records, county-level electricity outages, store transactions that record payment instrument, broader card-spending data, household purchase records, and surveys. That combination matters because no single data source can show the whole chain. A payment share shows what method was used, but not always why. An outage record shows infrastructure failure, but not whether a household had cash. A spending measure shows the outcome, but not which constraint caused it.
The practical implication is not that every household should return to a cash-based economy. Digital methods provide real benefits in normal times, and cash itself depends on a functioning distribution and retail system. The implication is that payment resilience is a portfolio property. A portfolio is stronger when one failure does not eliminate every usable path to an essential transaction.
The diagram captures a point that is easy to lose in arguments about whether cash is “obsolete.” Cash and digital payments are not simply competing brands on a single scale from old to new. They expose the user to different dependencies. When those dependencies are intact, digital convenience may dominate. When a dependency fails, the value of another route can rise sharply even if that route was used only occasionally before.
Two kinds of evidence, one question
The two supplied sources illuminate different sides of this problem. The Diary follows payment behavior over time in a nationally representative U.S. survey and diary, making it useful for understanding persistence, preferences, holdings, and demographic differences during ordinary conditions. The resilience study follows the timing of storms, outages, spending, and payment composition, then adds survey evidence from the United States, Spain, and Sweden. One source tells us why cash remains in the portfolio; the other asks what that portfolio can do when infrastructure fails.
That distinction also sets a standard for the claims that follow. The Diary does not by itself show that cash prevents a crisis-related spending decline. Its figures are descriptive and its future-use question measures intention. Similarly, a hurricane event study cannot treat every post-storm dollar lost as a payment failure. Physical damage, store closures, mobility, and supply interruptions remain part of the event.
What the combination can establish is more precise. Cash has a durable normal-time role, including a backup role for people who prefer cards. Electronic payments have infrastructure dependencies that can fail during severe weather. The resilience study's central task is to determine whether changes in spending and payment composition become sharper where the digital infrastructure actually fails, rather than merely where bad weather is observed.
That is the hidden job of cash: not to win every ordinary transaction, but to remain usable when the set of feasible transactions contracts. Seeing that job requires more than counting cash payments. It requires following the payment system through the moment when convenience ends and asking which options remain open.
The next unresolved issue is therefore concrete. Before deciding how much value a cash fallback has, we need to trace what fails first, how households prepare before the disruption, and why spending can move from cards to cash even as total spending falls. That is where the difference between a weather shock and a payment shock becomes visible.