The AI Journal

In 1728, the newly chartered Royal Bank of Scotland opened for business by attacking its only large rival. It bought up notes of the Bank of Scotland from merchants and travellers across Edinburgh at a discount, held them quietly for weeks, and then presented the entire hoard at the older bank’s counter on a single morning, demanding coin. Bank of Scotland’s specie reserve could not meet the sudden claim, and within days it suspended payment. That episode, a “note duel” deliberately engineered to bankrupt a competitor by liquidity assault rather than by offering customers a better product, opened the very period now cited as the model case for currency issued by competing private banks with no chartered monopoly and no lender of last resort. What followed really was, by the standard measures, remarkably stable: roughly a century and a quarter of multi-bank Scottish note issue produced far fewer failures and far smaller losses to note holders than contemporary England’s more restricted system. What is usually skipped is why a system that opened with an attempted liquidity assassination settled into decades of calm, and the answer is not simply that no central bank was there to get in the way. It is a set of specific, deliberately built restraints that the modern habit of citing Scotland as proof markets alone suffice does not ask itself to reproduce.

Bank of Scotland’s response to the 1728 raid, adopted in 1730, was the “option clause”: a line printed on its notes reserving the right to delay redemption in coin for up to six months, in return for five per cent interest on the deferred sum. This did not remove the threat of a raid; it blunted it, because a rival buying up notes to force sudden mass redemption could no longer manufacture an immediate liquidity crisis, only an interest bill. Other Scottish banks adopted the clause through the 1730s, and duels of the 1728 kind did not recur, though James Gherity’s reappraisal of the episode disputes how much of that credit belongs to the clause itself rather than to a shared, clause-independent reluctance among the surviving banks to repeat 1728’s near-catastrophe. Even granting the clause some causal role, retrospective accounts often write as though it explains the calm of the whole free-banking era, but it did not survive to see most of it: the Bank Notes (Scotland) Act of 1765 outlawed it, three decades before the mature, many-bank system that observers actually have in mind when they invoke “Scottish free banking” had fully taken shape. Whatever kept the peace between 1765 and 1844 needed a different mechanism, because the one credited with ending the duels of the 1720s and 1730s had by then been withdrawn by statute.

That later mechanism was the note exchange. Scottish banks arranged, informally at first and then through clearing meetings regularised from around the 1760s, to net one another’s circulating notes on a fixed schedule rather than let them return for redemption unpredictably. This did two things a chartered monopoly, or a modern regulator, would otherwise have to do. It gave every bank continuous, granular knowledge of how much currency its rivals had in circulation, since a bank issuing beyond what its business warranted would accumulate a persistent net debit at the exchange. And it gave the group leverage to discipline that bank without waiting for a public run, by demanding coin settlement on unfavourable terms or threatening exclusion from the exchange itself. A raid like Royal Bank’s no longer needed a hostile competitor buying notes in the street; imprudent issue was caught and punished from inside the very cooperative structure the banks had built to protect themselves from being raided again. That is not the absence of restraint that citations of Scotland usually imply. It is restraint administered by a private cartel of the regulated rather than by a chartered regulator, which is a different claim, resting on a different set of institutional facts, and one the advocacy literature rarely states outright.

Tyler Cowen and Randall Kroszner’s reassessment of the episode adds a complication that citations of Scotland tend to omit rather than answer. Statutory restrictions bounded nearly every dimension of Scottish note issue that an unregulated system would have left to competition: minimum note denominations were fixed by law, interest-bearing and other alternative instruments were periodically restricted, and new entry, while easier than in England, was never simply open to anyone with capital. More consequential still, Cowen and Kroszner argue that the Bank of England functioned as an unacknowledged backstop for the Scottish system, since Scottish banks held correspondent balances with London houses and could draw on Bank of England liquidity through those channels during periods of general stringency, most visibly around the crises of 1825 and 1837. A system with a silent lender of last resort one channel removed is not the same system as one with none, even though no Scottish bank drew on Threadneedle Street’s discount window by name. Citing Scotland as proof that stability requires no central bank quietly assumes away the very institution it is offered as evidence against.

The strongest objection to this argument concedes the institutional facts and denies they matter. Defenders of the free-banking reading can reply that Bank of England assistance to Scottish houses was occasional and marginal rather than routine, that Scottish banks bore genuine failure risk in the long stretches between crises, and that the option clause and the note exchange were themselves market-generated arrangements, adopted voluntarily under general law by competing firms protecting their own interests, not impositions handed down by a directing authority. On this view, the “not really laissez-faire” charge conflates background statutory constraints on note terms with the actual disciplining mechanism, which was privately organised and could just as easily be cited as evidence for, not against, what unforced competition produces. This is a serious point, and it forces a narrower claim than the one with which this essay began. The private origin of the note exchange does not restore the stronger thesis that no coordinating institution was needed; it only establishes that the institution which did the coordinating was built by the banks rather than by a state. That is precisely why the arrangement counts as engineered restraint rather than its absence: it had to be deliberately constructed, maintained against free-rider pressure, and backed by a credible threat of exclusion, the same functional requirements Charles Goodhart identifies as the reason clearinghouse arrangements tend to evolve toward central-bank-like coordination wherever multiple note issuers operate side by side. Conceding that Edinburgh’s bankers built their own restraint rather than inheriting it from London does not restore the claim that no such restraint was required.

The freedom Scottish banks are remembered for was freedom from a single chartered monopoly, not freedom from coordinated restraint. What replaced the missing monopoly was not the price mechanism working alone on isolated competitors but a clearinghouse the competitors built for themselves, doing quietly and by private agreement much of what a chartered central bank would later do by statute, while a foreign central bank stood behind the whole arrangement as a resource nobody wished to name. Free-banking advocates are entitled to the finding that rivals can build that machinery without a state mandating it. They are not entitled to the further claim that the machinery was unnecessary, because the one stretch of the Scottish record in which banks briefly approximated that stronger condition, the years immediately after 1728, before the option clause and before the mature exchange, is exactly the stretch the celebrated stability figures do not cover.