Chapter 6. Who Pays While It Lasts

6.1The bearer of an emergency act

An emergency act in money and credit does not merely stop something; while it stands, someone holds a position that would otherwise have changed. A suspension of payment leaves note holders without the exchange they were promised. A bank holiday leaves depositors without access to their balances. A deferral moves a payment date and leaves the interest bill standing. A ceiling on emergency lending leaves the bank that would have drawn on it short. A rescue dilutes the old holders and hands control to new ones.

The possible bearers in this ledger are not abstract. A holder of notes who cannot convert them into gold is bearing the suspension. A depositor whose bank is closed, or whose balance is larger than the sum that deposit insurance covers, is bearing the closure. A debtor whose principal is deferred but whose interest is still paid on time is bearing the deferral in one form, and the creditor is bearing it in another. A taxpayer who stands behind a guarantee is bearing a risk that has been shifted onto the state. A trader who may not sell above a fixed maximum is bearing the price control. The state whose revenue is pledged to secure an interest payment is bearing the debt conversion. In each case the position is created by the act itself: the act produces the wait, and the wait has a cost that falls somewhere.

The ledger's field "who bore the cost" answers one thing and refuses the rest. It reads the words of the record, not the world: where a record leaves the bearer unnamed, the reading is that no bearer is named, even where a later reader believes the answer obvious. It does not measure the size of any loss, does not name private persons, does not decide who ought to have paid, and does not treat the studio's recurring sentences as evidence. Where the two independent readings of the record disagree about the point, it is left unsettled, and an unsettled point counts as unclear in every verdict.

The ledger's separate axis on who bears the wait was settled in only 62 of the 84 cases (74%); of the values read, 54 were unknown, 22 unsettled, seven those nearest the stoppage, and one third parties or the vulnerable. The axis itself is set out in Chapter 2.2. The series' hypothesis that the cost of waiting falls not on whoever decides but on those nearest the point where things stop is read with the other series hypotheses in Chapter 12.6; this chapter is about what the records name, not the verdict.

The four cases here were chosen on this axis and no other. Diocletian's edict of 301 is the only one whose text names a group at fault in its own words. The South Sea record of 1720 separates four things and names no bearer at all. Mexico's deferral of 1982 and Greece's ceiling of 2015 are acts with terms a reader can see, and both records stop at the moment where a payer would have to be named.

6.2Diocletian's preamble

Skeptron's record of the edict of 301 reads its preamble as blame laid on traders:

The same record keeps that claim apart from the evidence about transactions:

The source text that goes with this chapter is Roland G. Kent's translation of the preamble, published in the University of Pennsylvania Law Review in 1920. In his own paragraph Kent says that he is moved by the desirability of having the preamble accessible to students in English, that he is not unaware of the temerity of attempting the task, and that he is reassured by the kindness of Professor Rolfe, who consented to the use of his and Tarbell's translation so far as it goes, with such verbal changes as Kent cared to make. Kent then prints the document in paragraphs of one sentence each, eighteen numbered sentences. The copy that reached this study is a machine transcription of a scan of the journal article; the header of that transcription marks it uncorrected and notes errors of reading that remain in it, so this study describes the sentences in its own words instead of quoting them.

In Kent's translation the preamble opens by invoking the national honour and the dignity and majesty of Rome, and by placing the state's fortune under the debt owed to the gods after the wars that established a tranquil peace. It turns next to avarice, described as inflamed almost every hour and even every moment, unchecked, and to common fortunes which, under an unhappy star, are ripped to pieces from day to day. The emperors present themselves as watchful parents of the whole human race who step in as arbiters, because humanity cannot achieve the long-hoped-for result by itself. The argument then becomes concrete: who, the preamble asks, can be ignorant that in commodities bought and sold in markets men force up prices wherever the general welfare requires the armies to be directed, in villages and towns and along every highway, not fourfold or eightfold but to a degree that language cannot evaluate; that a soldier is robbed of both bounty and pay by the outlay on a single article; that the contributions of the whole world for maintaining the armies accrue to the gains of plunderers. The remedy follows: the emperors resolved not that prices should be fixed but that a maximum should be fixed, so that avarice would be confined by the bounds of the statute while places where goods abound keep their cheapness; the list is to be held in attention throughout the whole domain, and all men are to understand that freedom to exceed the limits is removed. The preamble closes with penalties — capital punishment for anyone who acts with boldness against the letter of the statute, the same peril for a buyer who comes to an agreement with a retailer who violates it, and a heavier penalty still for one who withdraws goods necessary for food or usage from the market — and with an appeal for obedience, noting that a few have raged exceedingly for the ruin of the whole world.

Simon Corcoran's study of imperial pronouncements under Diocletian and the Tetrarchy, The Empire of the Tetrarchs of 1996, includes this edict among the pronouncements of the period, notes that it sought to fix maximum prices and wages and was issued along with the monetary reform of the period, and discusses the problem of inflation. This study uses his work only at the level of that description.

The preamble's order matters for the axis of this chapter, because it separates two things at once. It names a group as the cause: merchants, those who retail, those who haggle over the sowing of the seed and the business of retail dealers, buyers who deal with violators, and those who withdraw goods from the market. It distributes penalties among those groups. It does not, in the record's words or in Kent's translation, name a group as the payer. The preamble speaks in general terms of common fortunes and of people brought by extreme poverty to perceive their wretched condition; those are the closest the text comes to a bearer, and they are general. The record's two readings differ on whether it names who bore the cost, which shows how thin the line is between a group described as suffering and a group named as paying; this study leaves the point unsettled. Elsewhere the record is plain: it does not present the act as temporary, names no condition for its end, gives no figure for it, and names no two dated moments.

The record's fifth question asks what must be observed to maintain or withdraw a comparison with price control. Its answer names an observation about effect — how quantity, price and the possibility of transaction change after control — and a condition for withdrawal: that if announced figures have been substituted for actual transaction prices, calculations of profit and cost of living are withdrawn. Neither sentence names who bore the cost while the edict stood. The edict's penalties name who risks punishment; they do not name who pays for a maximum price.

6.3The South Sea Company

Skeptron's record for 1720 keeps four things apart:

Its boundary answer repeats the separation as a rule:

The record's own date is the year 1720, with no month or day entered, and its named source is the Bank of England's chronology of its founding, the South Sea Company and the convertibility restrictions.

The source text that accompanies this chapter is Charles Mackay's Memoirs of Extraordinary Popular Delusions, first published in 1841. The copy used here is the illustrated edition of 1852 as served by Project Gutenberg, which this study treats as uncorrected, so what follows is Mackay's account paraphrased. Mackay writes that the company was originated in 1711 by Harley, Earl of Oxford, with the view of restoring public credit after the dismissal of the Whig ministry, and of providing for the discharge of army and navy debentures and other parts of a floating debt amounting to nearly ten millions sterling; that the company took the debt upon itself; that the government agreed to secure a certain period of six per cent interest; that duties were made permanent to provide for that interest of 600,000l. per annum; and that the monopoly of the trade to the South Seas was granted, the company being incorporated by act of parliament. He describes expectations of wealth from the gold and silver mines of Peru and Mexico, the assiento contract with its limits on tonnage and cargo and its share of profits to the King of Spain, the proposals to parliament in 1717 and the three acts passed that year, and the plan of 1720 by which the company sought to take upon itself the debts of the state, upon five per cent per annum secured until Midsummer 1727, after which the whole was to become redeemable at the pleasure of the legislature, with the interest reduced to four per cent. He narrates the bidding of the company and the Bank of England for the government's favour, Walpole's warning that the scheme would raise the value of the stock artificially by exciting and keeping up a general infatuation, the stock's rise from about a hundred and thirty to near four hundred, its settling at three hundred and thirty as the bill passed the Commons by 172 to 55 and the Lords by 83 to 17, and the royal assent the same day; and he describes the crowds in Exchange Alley and the other schemes that followed.

Mackay's work should be read at the level of its reading note: it set the popular picture of financial manias as outbreaks of collective folly, a picture later scholars have questioned; it is a narrative written for a general audience, is not based on careful research into the sources, and should be read as a period text. This account was written in 1841, more than a century after the year it describes. Richard Dale's The First Crash of 2004 follows the same episode as a history: the scheme to take over government debt, the rise and collapse of the shares and the wider speculation of the time, the political and financial aftermath, and comparisons drawn with later crises. Peter Garber's Famous First Bubbles of 2000 argues that market fundamentals can explain much of the price movements in episodes often called manias, so that they should not be treated as simple cases of irrationality. Larry Neal's The Rise of Financial Capitalism of 1990 places the bubbles of 1720 in France, the Netherlands and England inside a developing system of integrated markets, and argues that the system survived them.

On the axis of this chapter the record says nothing at all. The two readings agree that the record shows an act of some kind but not what kind; they read it as not presenting the act as temporary and as naming no condition for its end. The record names no one who bore the cost, and whether it names who bore the wait is left unsettled. The record's third answer is an instruction about verification — a trade privilege may not be read as operating revenue, and the chronology may not be turned into daily prices — and an instruction of that kind is not a statement about loss. Who held shares at the collapse, whose debt terms were converted, whose duties stood behind the six per cent, and what any of them received afterwards are questions these records cannot answer. What the year can be read as a market episode belongs to Chapter 10.3, and the figures a record gives, read across the cases, are treated in Chapter 7.

6.4Mexico in 1982 and Greece in 2015

Two records, thirty-three years apart, show how far an emergency act's own terms reach toward naming who pays, and where they stop. Mexico's deferral moved the dates of principal payments on public-sector bank loans for ninety days and left the interest bill standing on its own schedule; Greece's record shows a ceiling on emergency lending kept at the level decided two days earlier, with the central bank stating that it stood ready to reconsider the decision. What this section adds to the question of who pays while an act lasts is a reading of what each record does name — a debtor state, banks, a period, a schedule of interest, a ceiling, a readiness to reconsider — and of what it leaves unnamed, which is a payer. The works that stand beside these cases, Boughton's official histories of the International Monetary Fund, Reinhart and Rogoff's catalogue of defaults, and Gourinchas, Philippon and Vayanos on the Greek crisis, speak to the shape and effects of lending programmes; the two records hold the shape of two acts and nothing at all about effects.

Mexico. The record for 20 August 1982 reads:

The record's setting is a negotiation. On one side is a state whose public-sector bank loans have become due; on the other, the banks that hold them. Between them runs a period in which principal is not repaid, and an interest schedule that continues untouched. What is waited for is the ninety days running out; the record does not say who waited, and this study adds no waiting party the record withholds.

The record answers the question of what the deferral may not be recorded as with a refusal, and then separates the periods:

Read plainly, the act moved the dates of principal payments on public-sector bank loans and left the interest bill standing; a request for 120 days, as another IMF source records it, is kept apart from the ninety days that were agreed. The deferral suspends part of an obligation without cancelling it, and the record's own third answer says that it may not be recorded as an exemption from all interest or as a write-down of principal. The record's named source is an IMF history chapter on the debt crisis and the Mexican negotiations of 1982. This study says of the negotiations only what the record says.

What the record names is specific: the negotiation with the banks, the public-sector loans, the ninety days, the treatment of interest. It names no bearer — no depositor, no taxpayer, no organ of the state named as the party that paid while the deferral stood. Two independent readings of the record disagree on this point, and this study does not resolve it. On the other axes this case stands out among the chapter's four: the ninety days is a term, and the condition of its end is stated observably, so a later reader could check whether the deferral ended when it said. The record's fifth answer states what would keep or withdraw a comparison with debt reduction — the change in repayment burden and available resources after a reduction, and the refusal to compare acts whose debt types, maturities and rules of treatment differ — and that answer names no payer either.

This study reads Reinhart and Rogoff's This Time Is Different as a long-run catalogue rather than as a source for the terms of a single act: it does not give the terms of the ninety days, and this record does not give the catalogue's scale.

Boughton's two official histories give the Fund's own account of the decades they cover. Tearing Down Walls (2012) follows the 1990s, including the Mexican crisis of 1994 to 1995 and the Asian crisis of 1997 to 1998. Neither book reaches 2015, so neither can say anything about the announcement of 28 June of that year.

Greece. The record for 28 June 2015 reads:

Emergency liquidity assistance is lending by a central bank to a bank that cannot obtain funds in the ordinary market, and in this record it is capped at a level set two days earlier. No public-domain text of the European Central Bank's own statement goes with this case, so the case is read from the record, which reports what the announcement did. The actors the record names are a central bank and the Greek banks it lends to; the events it references are a referendum decision and the non-extension of an adjustment programme. What is waited for is the ceiling holding and the decision being re-judged; the record names no depositor and no organ of the state as the party that bore the cost while the ceiling stood.

The record's third answer separates what a reader might otherwise merge:

The record's fifth question — which is not the weekday check that every case carries — asks how many days elapsed between the determination of the ceiling basis on 26 June and the stoppage on 28 June; its answer computes the interval as two days, treats the determination and the stoppage as two stages, and says that the order of the periods alone does not prove a single cause. The record's decision window is a post-hoc review kept apart from the facts of the announcement, and its third answer states that actual bank-by-bank withdrawal amounts and the timings of cash shortages are not present in the announcement and are not fabricated.

On the axes of this study the Greece record is far from a blank on structure and close to a blank on cost. The two independent readings did not agree on what kind of act this was, on whether the text presented it as temporary, or on what would end it; they did agree that the record shows an act, that two dated moments are named, that no sum is given, and that no one is named as bearing the cost. The separate question of who bore the wait reads those nearest the stoppage for this case, on the implication of the setting rather than in the record's own words, which is why the record itself names no bearer and why this case, among the four in this chapter, supports the series' hypothesis about bearers only on that implication. The record names no one who bore the cost of the ceiling, and this study supplies no payer the record withholds.

Pierre-Olivier Gourinchas, Thomas Philippon and Dimitri Vayanos analyse the Greek crisis in a 2017 paper: the build-up of public and external debt, the loss of access to markets, the official lending programmes, the restructuring of private-sector debt in 2012 and the contraction that followed, and the trade-offs between fiscal adjustment, official financing and debt relief. Their analysis begins in 2010 and reaches the restructuring of 2012; the reading note given with this study does not describe it as treating the announcement of 28 June 2015. Where paper and record stand together, both concern official lending inside a currency union and the burden that follows from it. Where they cannot meet, the record holds no programme terms, no amounts drawn, no bank's position and nothing after its own date.

Set beside each other, the two records share one shape. Each names an arrangement, a period and a counterpart, the banks in both cases, and each stops short of naming a payer. The deferral's interest schedule and the ceiling's level are terms a later reader can check; who stood behind them while the ninety days ran and while the ceiling held is not in either record.

6.5The debate over who pays in a debt crisis

The question of who pays in a debt crisis has a literature, and the literature argues about effects rather than about the contents of records. Chapter 1.4 maps it briefly; this section sets out what each position claims and where four records of this chapter touch it. These four records are dated readings of public sources, written for questions of chronology and scope, and none of them holds the evidence the literature argues over — a measured loss, a sum borne by anyone, or what happened after the date the record fixes.

This study reads Reinhart and Rogoff and Tomz as concerned with what happens to the payer over years rather than with the contents of a single emergency act's record.

James M. Boughton's two official histories give the institution's own account.

The critics take different evidence and a different object: Radelet and Sachs read the Asian crisis as a panic by short-term creditors, Feldstein argues that heavy conditionality beyond the balance-of-payments problem intruded on national policy and raises the question of moral hazard, and Paul Blustein's The Chastening (2001), a journalist's account of the crisis and its aftermath in Russia and Brazil, draws on interviews and reports on the debates within the Fund and the Treasury; it describes the programmes in Thailand, Indonesia and Korea and argues that policy mistakes, notably the handling of the Indonesian programme, damaged the Fund's reputation. Feldstein's and Blustein's readings differ in kind from Boughton's: a policy argument and a reported narrative against institutional histories written from the Fund's archives.

One document shows what it looks like when an act names its bearers, and it does so in the vocabulary of a domestic banking resolution rather than of a sovereign debt crisis. In the joint statement of 12 March 2023 by the Secretary of the Treasury, the Chair of the Federal Reserve Board and the Chairman of the FDIC, issued after a recommendation from the boards of the FDIC and the Federal Reserve and consultation with the President, a resolution of Silicon Valley Bank is approved that fully protects all depositors, with access to all their money from Monday 13 March. The statement names the groups it does not protect and the group that carries the cost: shareholders and certain unsecured debtholders "will not be protected", and any losses to the Deposit Insurance Fund to support uninsured depositors "will be recovered by a special assessment on banks, as required by law". Chapter 4 reads that act's record in full; this section takes from the statement only its vocabulary.

Mexico's record of 20 August 1982 shows a deferral whose interest was left standing:

An arrangement in which the debtor's payment dates moved and the creditor's claim was preserved rather than written down. That is a fact about the shape of an emergency act, and one a later reader can hold against any claim that the act forgave debt; it is also the kind of term any account of who paid would need, and the record gives the term and names no payer.

Greece's record of 28 June 2015 shows a ceiling kept at its level rather than raised, with the central bank stating, as the record reports it, that it stood ready to reconsider:

This study reads that as an act that looks revisable without any observable condition of revision, and the literature supplies nothing to fill the gap: a ceiling is a decision about who may draw and who may not, and the record holds no bank's position, no amount drawn, and nothing after its own date. Both records are dated readings of public sources, written for questions of chronology and scope rather than of cost.

Where the records cannot bear the debate is more of it. No record here measures a loss. None names a sum borne by depositors, creditors, the state or taxpayers. None records what happened after its own date, so none can test whether a programme's conditions were met, whether a deferral led to a restructuring, or whether a ceiling was lifted. Tomz's reputation question turns on how lenders later treat a defaulting sovereign, and a record of a deferral's terms does not observe that treatment. Feldstein's moral-hazard argument and the Radelet and Sachs reading of a panic are claims about effects; these records hold no effects. Boughton's histories rest on the Fund's archives; the two records here rest on public sources checked for scope and date. One line from the ledger's numbers states the limit in advance of the next section: across its eighty-four cases, one record names a bearer in its own words, two name one on the implication of the setting, thirty-five name none, and eight were left unsettled, while in thirty-eight the question cannot apply at all. No count in this study measures a loss to any of them, because no record supplies one.

6.6What the records name and leave unnamed

The ledger's counts answer this chapter's question in their own terms: in most of the cases in which an act is shown, the record names no one who bore the cost while it lasted.

In the ledger's reading of who bore the cost, across the 84 cases: not applicable in 38 cases, not named in 34, unsettled in 8, named by implication in 2, named in the record's own words in 1, and not named by implication in 1. Read as what they count, that is: of the 46 cases in which the question could apply, three named a bearer — one in the record's own words and two on the implication of the setting — thirty-five named none, and eight the two independent readings left in disagreement. The study's own hypothesis on this point is judged on the cases in which an act is shown; the 46 in which the question could apply are this chapter's base for the question, and the hypothesis counts apart the cases where it does not apply at all.

Counting only what the record states in its own words, it is supported by 31 cases, against 1, with 29 unclear and 23 not applicable — 97% of the cases the hypothesis decides. Counting also what the setting implies, it is supported by 35, against 3, with 6 unclear and 40 not applicable — 92%. By group, counting only what the record states, crisis history stands at 9 supporting, 0 against and 7 unclear; credit history at 17, 1 and 10; markets at 1, 0 and 1; money at 4, 0 and 1. The verdicts themselves are read in Chapter 12.4; what this section adds is the shape of the base: a field that applies in more than half the ledger and that most often stays empty.

Of the four cases in this chapter, the reading is not-named for the South Sea record and not-named for the Greek record, and unsettled for Diocletian's edict and for Mexico's deferral. None names a payer in the record's own words. The series' hypothesis that the cost of waiting falls not on whoever decides but on those nearest the point where things stop is unclear when only what the record states counts — 84 cases unclear, none supporting and none against — and it is still unclear when what the setting implies is counted too: 8 supporting, none against, 76 unclear, because the unclear cases outnumber the decided ones, even though those 8 are all the cases that decided anything. A hypothesis supported only by implication is a weaker thing than one the records state, and Chapter 12.6 says so.

One count stands beside all of this. Everything this chapter has read, four records at a time, is the shape of records that leave an act's end, its cost and its bearers unnamed.