The most potent censors do not hold elected office. They own the switch that decides whether your money moves. COVID-era payment lockouts and the Canadian trucker freezes were not outliers. They were stress tests that revealed how economic life now runs through a handful of private chokepoints. If power is the ability to say no, then the modern payments stack says it quietly, at scale, and often without appeal.
Banks and payment processors insist they are neutral pipes. They are not. They run risk-based compliance systems aimed at anti-money laundering, sanctions enforcement, and fraud. Those systems trigger alerts that are overwhelmingly false positives. In many institutions, more than 90 percent of AML alerts do not indicate actual crime. Yet each alert creates liability, reputational risk, and a paper trail. The rational response is to overblock. Terms of service stretch to cover amorphous categories like misinformation or reputational harm. What starts as a risk filter turns into a speech filter by proxy. The result feels like regulation, but it is enforced by contracts, not courts. The decision arrives as an email: account closed, funds on hold, appeal denied.
The case for going cashless is familiar: speed, safety, shrink reduction, cleaner books. It is true and it is incomplete. In engineering, redundancy is not waste. It is insurance against correlated failure. Cash is the offline mode of commerce. When systems stall, power fails, or compliance tools glitch, cash keeps the lights on. Replace a mixed system with a single rail, and you trade day-to-day efficiency for tail risk. Digital outages are not rare. Payment networks go down, real-time rails hiccup, and fraud controls misfire at scale. Some jurisdictions are moving to protect access to cash because they see the systemic risk. A society that cannot transact without permission is brittle. A store that cannot make a sale when a terminal times out is learning that lesson the hard way.
There is precedent for private policy performing public aims. Operation Choke Point, a Justice Department initiative a decade ago, encouraged banks to treat certain industries as high risk. Many were not illegal. They were politically disfavored. Even after the program ended, the lesson stuck: reputational risk is a supervisory risk. Similar dynamics surface when sanctions orders or emergency powers meet payment platforms. The Canadian trucker episode showed how fast financial plumbing can convert policy into frozen livelihoods. Scholars have noted how banks, operating under a regime of privilege and regulation, can become de facto regulators themselves. The line between public authority and private enforcement blurs. You may not get due process, but you will get deplatformed.
It is tempting to frame banks versus government as a power contest. It is more accurate to see a partnership. Banks manage debt contracts. The state enforces them. In crises, the tether tightens. The 2008 meltdown exposed the doom loop between sovereigns and banks. Lenders stuffed with government bonds magnified sovereign stress, while governments backstopped lenders to prevent collapse. Deposit insurance, lender-of-last-resort facilities, and bond purchases keep the system running. That mutual dependence did not end with the recovery. It is the operating model. The practical implication is stark: the system has two keys. Either the regulator’s order or the processor’s policy can lock you out. Asking which one is more powerful misses the point. Together, they shape the boundary of permissible economic life.
Payments are network businesses. That breeds coordination games. If one major processor blacklists an entity, competitors face a prisoner’s dilemma. Do they take the business and inherit the compliance and PR risk, or do they follow the blocklist and preserve standing with supervisors and media? The rational move is herding. Over time, that herding becomes a Schelling point: the safe default is refusal. Meanwhile, Goodhart’s law bites. Once a compliance metric becomes a target, actors game it. False positives rise. Flags spread between banks via shared utilities and background checks. A single minor infraction can cascade into a durable financial scarlet letter. Small error rates become system-level fragility when decisions are correlated. What looks like prudence in isolation becomes a kill switch in aggregate.
Power is measured by the cost of exit. If using cash is treated as suspicious, if every sale requires a KYC’d, surveilled, and centrally routed transaction, then exit is priced out. Central bank digital currency proposals raise this to code. Programmable money is efficient, auditable, and policy responsive. It is also a lever. Even without CBDCs, we already live with programmable payments via card networks and app stores that can deny service with a terms update. Investors and operators should model this as operational risk: concentration in a few processors, dependencies on a handful of banks, and policies that can shift overnight. In a downturn, or during an emergency, those dependencies tighten, not loosen. That is when optionality is worth more than throughput.
Nature avoids monocultures for a reason. Financial systems should too. An antifragile economy tolerates multiple forms of exchange and custody. Cash offers offline resilience and anonymity. Bank deposits offer scale and deposit insurance. Open banking improves portability. Community banks and credit unions diversify counterparty risk. Alternative rails like wires, ACH, real-time payments, and even well-structured stablecoins and crypto in self-custody add redundancy. None are perfect. That is the point. A portfolio of imperfect options is safer than a single perfect system. At the policy level, the test is simple: can a lawful business or person transact, appeal, and switch providers without ruinous friction? If not, the system is steering toward brittleness.
Do not ask whether banks are more powerful than governments. Ask how much it costs to opt out of the stack they jointly control. Map your dependencies to a handful of processors, a short list of megabanks, and a web of compliance signals that can follow you across providers. Model the tail: an erroneous alert, a mass outage, a political order. Then model the exit: secondary rails, secondary banks, cash acceptance, and legal recourse. The lesson of recent years is not that digital payments are bad. It is that convenience hides correlated risk. Freedom in markets is not the absence of rules. It is the presence of alternatives. Memory keeps those alternatives alive. Forget, and you will find out who really has the switch when you need it most.
