Chapter 3. Until Further Notice: Suspensions of Payment and Convertibility

3.1What a suspension promises

A suspension is an act with a particular shape. It stops one obligation while leaving the thing itself standing. The Bank of England's notes still circulate; the currency still exists; the debt is still recorded against a name. Convertibility is the promise that gives a note or a currency its stated worth on demand: that the holder may present it and receive a fixed quantity of gold, or a fixed quantity of another currency. A peg is a promise about an exchange rate, a commitment to buy and sell currency at a stated parity. When a suspension stops a payment, it stops the promise without destroying what was promised on, and that is why the first thing a later reader needs is not the date of the act but what the act's own record says about its return: a term, a date, a condition, or only the word temporary.

Each of this chapter's four cases is a suspension of this kind, and each record holds a different amount of that return. The Bank Restriction of 1797 answers that the Bank's obligation was stopped and gives a period:

The same record's boundary answer refuses to let the suspension be read as the extinction of the currency:

The record of Britain leaving gold in 1931 holds a single dated act set against a background, and no return:

Its boundary answer warns that a date of transition settles nothing about its effects:

The bank holiday of March 1933 is the one record here that holds three dated acts rather than one:

Its boundary answer separates what was reopened from what was recovered:

The closing of the gold window in August 1971 is an announcement with domestic measures attached, and the record keeps the deliberations that preceded it apart from the act itself:

What was announced beside the suspension, the record says in one line:

What each record therefore holds about a return differs sharply. The Bank Restriction's chronology gives its period as 1797–1821 and its record speaks of a later resumption; the proclamation of 6 March 1933 fixes Monday 6 to Thursday 9 March, and the record of the bank holiday gives reopening permits from 13 March; the record of 1931 gives only the date of 21 September and a background and names no end, no outcome and no bearer; the address of 1971 gives the word temporary, an exception clause and no date, no outcome and no bearer. Only one of the four records names a bearer of the cost, and it is the bank holiday's.

3.2The Bank Restriction, 1797–1821

The Restriction is the one suspension in this chapter whose length its own record gives, and it is the least fully recorded. Skeptron's record takes its period from the Bank of England's own chronology, as quoted above, and the period it gives runs from 1797 to 1821. The ledger's reading of this case treats the word temporary and the period 1797–1821 as a stated term and an observable condition of ending; this study reads them as the official chronology's retrospective presentation, and finds in the record no term first intended, no reason for the period's length, and no statement of what would end it. The record itself states the obligation stopped, the period the official history presents, and, in passing, the issuance of low-value notes and counterfeiting; it names no group of people holding the notes during those years.

The minute of the Privy Council of 26 February 1797 names no date and no other condition for the end of the suspension. The account of the suspension that this study can therefore give comes from four directions: the record itself, the Privy Council minute, Henry Thornton's Enquiry of 1802, and the House of Commons committee's Report of 1810.

Thornton, a banker and a member of Parliament, treats the Bank's conduct during and after the restriction in his Enquiry of 1802. In the Chapter IV excerpt of that text Thornton writes that if there was any fault in the conduct of the Bank of England it lay rather in restricting its notes in the late seasons of alarm than in enlarging them, and that the Bank may have in some measure aggravated rather than lessened the demand upon itself for guineas through the suppression of too many notes at the time preceding the suspension of its cash payments. He answered the objection that Parliament should not have interposed: the question, he wrote, was not whether one holder's claim should be interfered with but whether all holders of notes and all persons entitled to demand cash payments should be allowed to press their claims at once, when almost every person entitled to demand payment was also a debtor. He described the suspension as giving effect to what must have been the general wish of the nation in new and extraordinary circumstances, and described what followed: the Bank extended the quantity of its notes nearly to the amount usually in circulation, credit revived, and guineas became remarkably abundant, with gold flowing into the country and the course of exchange turning in Britain's favour.

In the Chapter V excerpt of the Enquiry Thornton argues that the repeal of the law suspending cash payments in the last period of the war would have been inexpedient, because repeal would have come when gold had recently been drawn out of the country by an unfavourable exchange and when the country was again subject to alarms of invasion. His conclusion is worth holding beside the record, because it speaks to the question of term: the renewal, he wrote, stood on the particular circumstances of the times, and not on any principle which necessarily implied the permanence or even the long continuance of the suspension.

The Bullion Report of 1810 gives the terms under which Parliament had left the question open. In its opening the committee found that the market price of standard gold had risen well above the Mint price and that exchanges with the Continent had run far below par; most of the witnesses it called attributed the high price of gold to an alleged scarcity arising from an unusual demand on the Continent and to the practice of hoarding, while the committee's own judgment pointed to something in the state of the domestic currency. Its opinion was that the circulating medium ought to be brought back, with speed compatible with caution, to cash payments at the option of the holder of Bank paper, and that the manner of the return should be left largely to the Bank's discretion while Parliament fixed the time. On that time the committee's words were "the Bank would be compelled to pay in cash at the end of six months after the ratification of a definitive Treaty of Peace" (Report from the Select Committee on the High Price of Gold Bullion, 1810), and its proposal was "Your Committee are therefore of opinion, that even if Peace should intervene, two years should be given to the Bank for resuming its payments ; but that even if the War should be prolonged, Cash payments should be resumed by the end of that period" (the same report).

Two features of that proposal bear on this chapter's question. The condition in the law as it stood was an event, not a date: peace would be ratified at some future moment, and six months would run from then. The committee replaced that condition with a fixed period, two years from 1810, which is itself a condition of a different kind, and left the mechanics to the Bank's own discretion, with a further provision that the chartered banks of Ireland and Scotland and the country banks should not be compelled to pay in specie until some time after the Bank of England had resumed. A reader can see a term being argued for, and can see who was to be waited for in the meantime; what no text in this chapter's evidence gives is the term intended when the restriction began in 1797.

Two histories of British monetary thought put the Restriction inside a longer argument. Frank Whitson Fetter's Development of British Monetary Orthodoxy, 1797–1875 traces the evolution of British thinking about money and banking from the Restriction itself, through the bullion controversy and the debates between the currency and banking schools that produced the Bank Charter Act of 1844, to the later suspensions of that Act in 1847, 1857 and 1866. The Restriction, on Fetter's account, is the beginning of a period in which an orthodoxy about convertibility and about the management of the note issue had to be formed and reformed under pressure. John Clapham's two-volume institutional history of the Bank of England, written with access to the Bank's own records, describes the restriction of cash payments from 1797 as one of the crises the Bank faced as it grew from a lender to the state into an institution holding the banking system's reserves. This study reads the notes and the texts of both works as giving no term for the Restriction beyond the period itself.

What the record leaves unnamed is the people. It tells the reader that low-value notes were issued and that counterfeiting occurred; it does not say who held notes during those years, how many held them, or what happened to anyone's claim while the obligation stood suspended. Among the eighty-four cases of this study's ledger the record states who bore the cost in one case and names nobody in thirty-four; this record names none. The texts of 1802 and 1810 say something about holders: in the Chapter IV excerpt of the Enquiry Thornton writes that it was important to continue for a time to pay the labouring people in money and to circulate the new one and two pound notes first by the medium of the higher classes, and in the conclusion of the Report of 1810 the committee speaks of the rights of commercial property and of cash payments at the option of the holder of Bank paper. Those are the authors' words in 1802 and 1810, not the record's, and the record's silence is the thing this chapter's question is about.

3.3The bank holiday of 1933

March 1933 is the case in which a suspension and its return are both written down by the actors themselves. It supplies a fact the other three cases cannot: the return of a suspended payment can itself be an act, issued bank by bank, by an authority that had reserved the power to permit what it had suspended. The proclamation fixes the holiday's term inside the instrument; the Act of 9 March confirms what had already been done and widens the power; the address of 12 March describes the return as a staged reopening rather than as a date. A suspension with a stated term is not the same thing as a suspension whose end is an observable event, and this case holds both in one week.

This is a synthetic scene, assembled only from the instruments and the record quoted below. It adds no fact, date, person or outcome to them.

On the evening of 6 March 1933 the President's signature stands at the foot of a proclamation that names four days, and the prohibition inside it reaches every banking institution and every branch in the country, in the territories and insular possessions included. Behind the instrument sits the Treasury, to whom the holiday hands three powers: to permit any or all banking institutions to perform any or all of the usual banking functions, to direct the issuance of clearing house certificates against the assets of banking institutions, and to authorise special trust accounts for new deposits, held separately in cash or on deposit in Federal Reserve Banks or invested in obligations of the United States. The banks are shut; the country is waiting to learn which of them the Treasury will permit to open on Monday; and the only instrument that speaks to that question does not name them.

The proclamation's words on the holiday and its term are these: "do hereby proclaim, order, direct and declare that from Monday, the Sixth day of March, to Thursday, the Ninth day of March, Nineteen Hundred and Thirty-three, both dates inclusive, there shall be maintained and observed by all banking institutions and all branches thereof located in the United States of America, including the territories and insular possessions, a bank holiday, and that during said period all banking transactions shall be suspended" (Franklin D. Roosevelt, Proclamation 2039, 6 March 1933). Its recitals state what it took itself to be doing: heavy and unwarranted withdrawals of gold and currency for the purpose of hoarding, speculative activity abroad in foreign exchange producing drains on the nation's stocks of gold, and a national emergency, in which a period of respite was provided in the interest of bank depositors, drawing on the authority given to the President by the Trading with the Enemy Act of 1917. The prohibition allowed no banking institution or branch to pay out, export or earmark gold or silver coin, bullion or currency, to permit its withdrawal or transfer by any device, to facilitate its hoarding, to pay out deposits, to make loans or discounts, to deal in foreign exchange, to transfer credits from the United States abroad, or to transact any other banking business. The instrument's term is a pair of days, Monday 6 to Thursday 9 March, both inclusive.

The Emergency Banking Relief Act of 9 March 1933, passed three days after the proclamation, confirmed what had been done: its first title approved and confirmed the actions and proclamations of the President and the Secretary of the Treasury since 4 March 1933, and widened the President's power over transactions in foreign exchange and the hoarding and export of gold and currency during a national emergency declared by him; its second title provided for conservators of national banks. The Act's architecture therefore holds a suspension whose period is set by a proclamation, and powers whose operation begins by confirming an act already taken, their own period left to a proclamation made three days earlier.

What a listener was told about the return came six days later, in the radio address of 12 March 1933. The account was mechanical: a bank does not hold deposits in a vault but invests them in credit, so a sudden demand by all depositors cannot be met without selling assets at panic prices; the holiday supplied the currency necessary to meet the situation, and the new law allowed the twelve Federal Reserve Banks to issue additional currency on good assets, so that the banks which reopened would be able to meet every legitimate call. Roosevelt told listeners that no sound bank was a dollar worse off than it had been when it closed its doors. The reopening was staged rather than general: banks in the twelve Federal Reserve Bank cities would open on the Monday after the address, those in cities with recognised clearing houses on the Tuesday, and banks in smaller places on the Wednesday and succeeding days, subject to the Government's ability to complete its survey. The address warned that some banks would be unable to reopen without reorganisation, and made no promise that individual losses would not be suffered: Roosevelt said that there would be no losses that possibly could be avoided, and that there would have been more and greater losses had the drift continued. The one sentence later quoted everywhere from that address tells the listener where to keep money: "I can assure you that it is safer to keep your money in a reopened bank than under the mattress" (Franklin D. Roosevelt, Fireside Chat on Banking, 12 March 1933).

The record keeps these three dates apart:

and it adds a boundary that reopened institutions and individual recovery are different things:

That second answer restates a boundary the 1933 texts had already drawn: six days earlier the fireside chat had said, in the actors' own words, that there would be banks unable to reopen without reorganisation. The record, written in 2026 by an archive that reads public sources, gives three dates and one boundary.

Arguments about the causes and effects of the week rest on evidence the record does not hold; this chapter reads them as testimony from outside the instrument, beside a record whose own words stop at three dates and one boundary.

Here the condition is an act rather than a date: a reopening permit, issued bank by bank from 13 March. Whether the permits were issued as fast as the country waited, whether the depositors who had fled returned, and how many of the permanently closed banks reopened later are questions this record does not answer; the study's question is only whether the record gives a reader what would settle them. What this case adds is that a suspension can carry its term in its own instrument, that its end can be written as an act of a named authority, and that even then the record names who bore the cost while it stood — the one case in the chapter that does.

3.4Britain in 1931 and the dollar in 1971

Two of this chapter's four suspensions come with the actor's own text attached, and in both the text itself carries terms that the record does not restate.

The Act of the United Kingdom Parliament passed on 21 September 1931 suspended the operation of the subsection of the Gold Standard Act of 1925 that obliged the Bank of England to sell gold at a fixed price. Its opening words were "Unless and until His Majesty by Proclamation otherwise directs, subsection (2) of section one of the Gold Standard Act, 1925, shall cease to have effect, notwithstanding that subsection (1) of the said section remains in force" (Gold Standard (Amendment) Act 1931, section 1(1)).

The Act's second subsection discharged the Bank from liability for anything it had done against the suspended subsection since 18 September 1931, and its third subsection made it lawful for the Treasury to make orders authorising such measures on the exchanges and otherwise as might be expedient for meeting difficulties arising in connection with the suspension; that third subsection alone was to continue in force for a period of six months from the passing of the Act. The asymmetry is the point for this chapter. The suspension itself has no term: it runs until a proclamation directs otherwise. Only the Treasury's power over the exchanges was given a term, and a short one. The condition of return in the text is a decision by a named authority, taken at an unnamed time.

The record of this case, read from Bank of England research, gives one date and a background, and its boundary answer, quoted in 3.1, refuses to attribute effects to the date. It names no end, no outcome and no bearer of the cost while the constraint was in force. The record's own fifth question compares conditions of a different kind and asks what must be observed for a comparison to be kept or withdrawn; it is a boundary, not a statement of the act's return.

The address of 15 August 1971 is the announcement, and it is explicit about the word the reader needs: "I have directed Secretary Connally to suspend temporarily the convertibility of the dollar into gold or other reserve assets, except in amounts and conditions determined to be in the interest of monetary stability and in the best interests of the United States" (Richard Nixon, Address to the Nation Outlining a New Economic Policy, 15 August 1971).

The address announced a package rather than a single act: a freeze of wages and prices for ninety days, which the address itself described as temporary; a surcharge on imports, which the address also described as temporary, saying that when the unfair treatment ended the import tax would end as well; and the suspension of convertibility. The import tax is the one measure in the address to which a condition of ending was attached in general words. The convertibility suspension got the word temporary, an exception clause, and no date. The address also promised, to the international banking community, that the United States would press with the International Monetary Fund and its trading partners for the reforms needed to set up an urgently needed new international monetary system.

Joanne Gowa's study of the closing of the gold window explains the American decision of August 1971 as a product of domestic politics: the administration's priorities, including lower unemployment and the electoral calendar, together with pressure on the balance of payments, led to suspension rather than to external events alone; the book examines the choice among the options available to the administration and the role of domestic actors. Harold James's history of international monetary cooperation since Bretton Woods follows the par value system's rise and breakdown, the era of floating rates and the changing role of the International Monetary Fund, and argues that cooperation was shaped by repeated crises and by the interests of the major powers. This study reads the notes and the texts of both works as naming no condition for the dollar's return to convertibility.

What followed on the records' own dates is what the records say. For 1931 there is one date and one background, and no later date inside the record. For 1971 there are two sets of dates: the policy consultations of 13 to 15 August, which the record keeps separate from the announcement, and the announcement itself, with which the wage and price controls were announced at the same time. The record's own chronological question answers how many days stood between the meeting and the announcement, and answers two, on the civil calendar; the record then warns that the sequence of two events does not establish causation between them. Both records, in short, hold the act and what was said beside it; neither holds a return.

3.5The debate over the rule and its escape clause

A suspension of convertibility is a departure from a rule, and the literature asks whether such a departure can be understood as part of the rule rather than as its abandonment. Two positions frame the question, and what each needs from a record — a return, an expectation, a threshold — can be tested against the four records read in this chapter.

The escape-clause reading, held by Bordo and Kydland, conditions a suspension's return not on a date. It is an expectation held by the public.

Kydland and Prescott's warning sits against that reading: a rule binds only so long as departing from it looks worse than keeping it.

The two positions differ in what each asks of a later reader. The warning about discretion concerns the temptation to depart from a plan once agents have acted on it. The escape-clause reading depends on the expectation that a suspension will be reversed, and that expectation is itself a judgment about future discretion. A record that holds only the word temporary, without a date or a threshold, gives a later reader nothing on which to test whether that expectation was ever warranted; the address of 1971 is exactly such a text, since it attached a condition to the import tax and none to convertibility.

Two of the four records hold a return, two leave it unnamed. The Bank Restriction of 1797 is the first returning record, and its return is the thinnest evidence in the chapter. The record's boundary answer separates the scope of the Restriction Period from its later resumption and quotes an official chronology presenting the period as 1797–1821. The minute of the Privy Council of 26 February 1797 named only the sense of Parliament, and the committee of 1810 proposed two years, peace or war, leaving the manner of resumption to the Bank's discretion. A chronology giving the period as a whole is how the episode came to be presented; it is not a term anybody stated at the beginning.

The bank holiday of March 1933 is the second returning record, and here the return is written by the actors themselves. The order of 6 March fixed a term in days, the Act of 9 March confirmed and widened the powers, and the record's answer gives reopening permits from 13 March; the address of 12 March set out the order of reopening — the twelve Federal Reserve Bank cities on the Monday, cities with recognised clearing houses on the Tuesday, smaller places on the Wednesday and succeeding days, subject to the Government's ability to complete its survey — while stating that some banks would be unable to reopen without reorganisation. Where the Restriction carries its return only in a later chronology, the holiday carries its return in the order, the statute and the address.

The other two records hold a departure and name no return, and in both the actor's own text differs from the record's summary. The Act of 21 September 1931 suspends the operation of the subsection of the Gold Standard Act of 1925 that obliged the Bank of England to sell gold at a fixed price, and it does so in these words: "Unless and until His Majesty by Proclamation otherwise directs, subsection (2) of section one of the Gold Standard Act, 1925, shall cease to have effect, notwithstanding that subsection (1) of the said section remains in force" (Gold Standard (Amendment) Act 1931, section 1(1)). The condition in the text is a decision by a named authority at an unnamed time, and it is a condition the record does not restate. Only the Treasury's power over the exchanges was given a term — six months from the passing of the Act — and the suspension itself was given none. The address of 15 August 1971 is explicit in the other direction: "I have directed Secretary Connally to suspend temporarily the convertibility of the dollar into gold or other reserve assets, except in amounts and conditions determined to be in the interest of monetary stability and in the best interests of the United States" (Richard Nixon, Address to the Nation Outlining a New Economic Policy, 15 August 1971), while "As a temporary measure, I am today imposing an additional tax of 10 percent on goods imported into the United States" carried the condition that when the unfair treatment ended the import tax would end as well. Convertibility got the word temporary and an exception clause; the tax got a condition of ending. Neither text names a date on which the suspension of convertibility would be lifted.

Read in the light of that pair, the Bank Restriction and the bank holiday are suspensions within a rule, since each carries the shape of a return — a period argued for in the first, a staged reopening written by the actor in the second — though neither record holds the expectation that Bordo and Kydland make the condition of their clause. The 1931 and 1971 records are departures whose return they leave to an unnamed act at an unnamed time, which is what Kydland and Prescott's warning asks a reader to distrust: the value of the rule rests on the expectation of a return, and these two records hold no expectation, no threshold and no date on which that expectation could have been tested.

The four records give no reasons for the length of any suspension, no account of what the public expected, and no evidence of what the exchange rate, the output or the price level did after the date. The record of 1931 says only that the background included economic contraction and pressure on the pound; the record of 1797 gives a period and says nothing of why it ended when it did. Across the cases in this study, most texts with an act name no observable condition, date or event on which the act would end or be revised, and of the forty-two cases with an act one carries a holiday presented without a stated term together with an observable condition of ending, who bore the cost, and two dated moments — this study reads that one as the record of March 1933. A debate about expectations cannot be settled on records that hold no expectation, and a debate about recovery cannot be settled on records that hold no outcome.

3.6The debate over the emergency that stays

The debate this section serves asks what becomes of emergency power once the emergency's reason is gone. The four suspensions in this chapter show what an authority wrote at the moment of acting and almost nothing about how long it expected the emergency to last; they can test the shape of a legal accommodation, but not the ratchet, which is a claim about what happens afterwards.

The four records are the moments at which the step is taken; none holds the calm afterwards. The bank-holiday record names three dates, 6, 9 and 13 March 1933, and its boundary says that banks permanently closed remained. The Restriction record carries no moment of resumption, only the period the Bank's own chronology retrospectively gives. The 1931 record gives 21 September 1931 as a transition date and, in its boundary answer, attributes no effects to it. The 1971 record keeps the consultations of 13 to 15 August apart from the announcement of 15 August, names wage and price controls beside the suspension, and says what became of neither. This study reads the ratchet as a claim these records can neither bear nor refute: they hold the act and its stated reason, and leave persistence out.

Higgs's proposition has a sibling in the ledger: that a restriction outlives the memory of its reason. The reading across the eighty-four cases went against it: in most cases the record gives a reason for the act. The four records here do so in the actors' own words. The Privy Council's minute of 26 February 1797 records alarms and an apprehended want of cash for the public service; the proclamation of 6 March 1933 recites hoarding, speculative activity abroad and a national emergency; the address of 15 August 1971 names speculators attacking the dollar; the 1931 record names economic contraction and pressure on the pound as background. What the records almost never give with the reason is its expected length, and a reason recorded without a duration cannot be set beside the act's own duration.

Gross and Ní Aoláin's Law in Times of Crisis sets out the models a legal system meets in an emergency: that ordinary law applies fully; that emergencies lie outside the law; and that emergency powers are accommodated within the law, with rights derogated under defined conditions. Its subject is terrorism and threats to national security; here its relevance is the shape of the accommodation. The 1933 record shows that shape most clearly. The proclamation of 6 March drew its authority from Section 5(b) of the Act of 6 October 1917 and set a holiday of fixed dates; the Act of 9 March approved and confirmed the President's and the Secretary of the Treasury's actions since 4 March, amended the Trading with the Enemy Act to reach "any other period of national emergency declared by the President", and gave the President's power over gold a fresh basis in subsection (n) of section 11 of the Federal Reserve Act. The emergency is inside the statute from the first day, and the restriction ends in the record not by a date in the instrument but by a permit issued bank by bank; the fireside chat of 12 March puts it that the reopening of smaller banks would come "subject, of course, to the Government's physical ability to complete its survey". The 1931 Act is another version: Parliament suspended a subsection, discharged the Bank from liability, and gave the Treasury's power over the exchanges a term of six months while giving the suspension none. The 1797 minute and the 1971 address show the accommodation less clearly, since neither holds an instrument of return.

This study reads the four records as documentation of deferral in sequence. The 1933 record's three dates put an executive proclamation on 6 March, a statute on 9 March and permits from 13 March: the order came first and the confirming law after. The 1971 record separates deliberation from announcement, and the announcement from the measures attached to it. The 1931 record holds a single statute, the 1797 minute a council's unanimous opinion addressed to the Bank's directors; one holds a date and one a period, and both sit less well under this lens.

The final point belongs to the records. Each of the four cases is recorded at a single moment — a year for the Restriction and a day for the other three — and each leaves a different remainder. Chapter 4 turns to the second kind of act these records show, lending and rescue in a panic.