Money as Software: Digital Payments and the People Banks Missed

A bank branch is an expensive object. It needs a building, a vault, staff, security and enough customers within travelling distance to justify all of it. That arithmetic explains the shape of global finance more than any policy ever did: banking went where density and wealth already were, and the places that failed the arithmetic went without. Not because anyone decided to exclude them — because a branch there would lose money.

What changed is that the arithmetic changed. Once a payment is a message rather than a physical transfer, the marginal cost of serving one more customer approaches zero, and the question "is this village worth a branch?" is replaced by "does this village have phone coverage?" That is a much easier question to answer yes to.

What actually moved

It is worth separating three things that get bundled together as "fintech", because they have different effects on ordinary lives.

• Payments — moving money between people and merchants instantly and cheaply. This is the part with the most transformative reach and the least glamour.
• Storage — a place to keep money that is not a mattress, a jar or a relative. Less discussed, arguably more important, because the primary financial problem of low-income households is not access to credit but the safety and liquidity of small savings.
• Credit and insurance — the layer built on top of transaction history, which is where both the greatest promise and the greatest hazard sit.

The first two are close to unambiguously good. The third is where the story gets complicated.

The person who benefits most from instant payments is not the one who wanted convenience. It is the one who was previously carrying cash home in a bag.

The everyday consequences

Ask people in newly-digital payment markets what changed and the answers are strikingly practical, and rarely about technology.

Safety. Cash has to be carried, hidden and defended. Market traders who used to close early to avoid travelling home with a day's takings now stay open. This is a direct income effect from a security improvement, and it is invisible in any technology metric.

Distance. Sending money to a parent in another region used to mean a bus ride, a courier or a friend travelling that way. It now takes seconds and costs a fraction. Remittance flows — which in many countries exceed foreign aid — got materially cheaper, and the difference stayed in the recipient household.

Granularity. Digital payments make small amounts economic. Pay-as-you-go solar, water and cooking gas exist because collecting fifty cents a day is now feasible. That converts an unaffordable capital purchase into an affordable operating expense, which is how a great many households got electric light for the first time.

Legibility. A transaction history is a record. For a small trader with no accounts, no payslip and no collateral, it is often the first documentary evidence of income they have ever had — and the basis on which credit becomes possible at all.

What it replaced

Cash, in part — though far less completely than the coverage suggests, and cash retains two properties nothing digital has matched: it works when the network does not, and it does not create a record. Both matter more than they seem.

More significantly, it replaced informal financial infrastructure: the moneylender, the bus courier, the savings group, the shopkeeper who held your money. These systems were not primitive. They were sophisticated, socially embedded and expensive, and they persisted because nothing better existed. Replacing them with something cheaper is a genuine gain — but the informal systems also provided flexibility, forbearance and human judgement that an automated ledger does not.

What it cost

Frictionless credit. The same transaction data that makes lending possible makes it possible to lend badly, instantly and at scale. Instant micro-loans delivered by phone, with fees expressed as flat percentages that conceal annualised rates in the triple digits, have produced real over-indebtedness in several markets. The technology removed the friction that used to function as a safeguard — the walk to the branch, the conversation, the delay in which you might reconsider.

Fraud that scales. Instant, irreversible transfers are perfect for social engineering. The victim authorises the payment themselves, which places them outside most fraud-protection schemes. Older users are targeted disproportionately, and the losses are often unrecoverable within seconds.

Exclusion by dependency. As merchants stop accepting cash, everyone without a phone, an ID, a reliable network or the confidence to use the interface is pushed out of ordinary commerce. That population is disproportionately old, poor, disabled or undocumented. A cashless society is a closed door for whoever cannot open the app.

Concentration. Payment rails have strong network effects, which tends toward a small number of dominant operators. When a single private platform becomes the way a country pays for things, its outage is a national event and its fee schedule is effectively a tax. The markets that handled this best treated instant payments as public infrastructure with mandated interoperability, rather than a product.