6.1Three ways of writing an end
The documents read here write the end of a suspension in three forms. A procedural end names a step in government: the sense of Parliament in 1797, a royal proclamation in 1931. An event end names something that will happen in the world: six months after a definitive peace, under the law as the Bullion Committee described it in 1810. A dated end names a time: two years, peace or war, as the committee proposed; six months, for the Treasury's power under the Act of 1931.
The forms differ in what a holder of paper could learn from them. A procedural end tells the holder who will decide, and nothing about what they will decide or when. An event end tells the holder what to watch for, but if the event is a peace treaty in a long war, the holder cannot date it; the committee's complaint was precisely that such an end was indefinite. A dated end tells the holder when, and says nothing about whether the date will be kept, which depends on the authority's word.
The escape-clause reading, as the study's note on Bordo and Kydland states it, requires a suspension in an emergency and an expected return to the old parity. Of the three forms, the event end is the one closest to its logic: the suspension lasts as long as the emergency, war, and ends when the emergency ends. That is what the law as it stood in 1810 provided, by the committee's account. The committee objected to it on grounds of credibility and wanted a date instead. The Minute's procedural end and the Act's procedural end are the furthest from the logic: they say who will decide, and not on what ground.
In this article's reading, this yields a refinement of the escape-clause reading that the documents themselves suggest and that Bordo and Kydland's note does not address: an escape clause can be written as an event (the emergency's end), as a date, or as a procedure, and the contemporaries who argued about the Restriction thought the choice among these mattered for the credibility of the return.
6.2Where the cases bear on the debates
On the question whether a suspension is part of the rule or a departure from it, the documents of 1797 to 1810 support the premise of the escape-clause reading without supplying its mechanism in the order that began the suspension. Thornton ties the renewal to the circumstances of the time, with his hedged refusal of permanence; the committee assumes a return and argues only about when; neither treats the suspension as a rejection of convertibility. But the original order named no end except Parliament's sense, and the end the law wrote, by the committee's account, was one the committee thought too indefinite to sustain confidence. The committee expected a return and argued that an indefinite end endangered it; whether the public expected one is beyond what these texts can show.
The Act of 1931 supports neither reading by its words, but its form fits the golden-fetters side better. It keeps the gold standard's legal frame in place, switches off the obligation to sell gold, and leaves the restoration of that obligation to the government's discretion, without a condition or a date. If Eichengreen's account is right that leaving freed governments to expand (Eichengreen 1992), the Act is written so as not to bind that freedom. If Kydland and Prescott are right that discretion invites worse outcomes than rules (Kydland and Prescott 1977), the Act is written as discretion. Eichengreen and Sachs's comparison is by the date on which countries left gold (Eichengreen and Sachs 1985); the Act supplies a date, and the record keeps it, but the note on their article does not name Britain, and this article draws no conclusion about Britain's recovery.
On the question of who bore a suspension, the documents name more than the records do. The Minute names the public service as the thing to be protected and the alarmed demanders of specie as the cause. Thornton names the labouring people as protected by coin for a time, the higher classes as the first channel for the new notes, and argues that, since creditors are debtors too, the suspension served everyone. The committee names the country banks and the chartered banks of Scotland and Ireland as needing time of their own on resumption. The Act names the Bank as indemnified and bars proceedings by anyone harmed. The modern records name no one.
6.3What the modern records keep and lose
The records of both episodes keep their dates carefully and decline to put later movements down to the act alone. What they lose is what the actors wrote about the end. The record of 1797 keeps the word temporary and gives an end year, but the end year comes from a later chronology, and the record does not say that the original order named only a procedure, that the law as the committee describes it named a peace, or that a committee argued for a date. The record of 1931 keeps the date of the Act but none of its three other moments, and does not say that the suspension's end was left to proclamation or that a subsidiary power had a fixed term.
The ledger found a term and a checkable end in the first summary and neither in the second. Across the eighty-four cases it read, the study found that in most cases with an act the record names no observable condition of its end and does not name who bore the cost; the two records read here fall on either side of the first finding and squarely within the second (Appendix). The first record falls on the side of naming an end because of hindsight, not because the actors named one. A summary that takes an act's period from a later chronology will look, to anyone reading it for what the act promised, as if the act promised its own end.