Britain, the Gold Standard, and the Standard Formula
Take care of the pence, and the pounds will take care of themselves.
—William Lowndes
With the Great Recoinage of 1696, Britain withdrew from its earlier experiments with token subsidiary coins and reaffirmed the medieval idea of a full-bodied commodity money throughout the denomination structure. It thereby arrested its earlier substantial progress toward the standard formula. Nevertheless, in the eighteenth century, market pressures and a government policy neglecting subsidiary coinage propelled Britain again toward the standard formula. These developments contributed to the gradual process by which Britain approached the gold standard in the early nineteenth century, when it explicitly made its subsidiary coins into tokens that it promised freely to exchange for gold coins. 1
When the eighteenth century began, the unit of account for Britain was still in silver. In terms of the silver unit of account, the gold guinea issued by the government fluctuated in price. Because they affected tax collections, those fluctuations attracted the government’s attention in the mid-1690s. The government tried in vain to regulate the price of the guinea from the 1690s to the early eighteenth century. A leading monetary authority, Sir Isaac Newton, asserted the futility of those attempts. However, gold eventually emerged as unit of account.2
In addition, the price of silver coins expressed in terms of gold stabilized after 1717. Persistent depreciation of small denomination coins, the telltale sign of shortages of small change found in figure 2.1 on page 16, is absent in eighteenth-century Britain. Thus, besides transforming the unit of account, eighteenth-century Britain effectively solved the problem of small change. Counterfeiters and other suppliers of tokens somehow produced enough small change to have allowed the exchange rate between large (now gold) coins and smaller silver coins to stabilize after 1720.
Meanwhile, the government’s neglectful policy toward the coinage allowed the stocks of both gold and silver coins to deteriorate until many coins became substantially underweight. The prevalence of underweight coins widened the effective intervals between minting and melting points for both gold and silver coins. Wider intervals facilitated a fixed exchange rate between gold and silver coins because they allowed larger fluctuations in the relative price of uncoined gold and silver before they would trigger melting of underweight coins or minting of full-weight coins.
Thus, by the end of the eighteenth century, wear and tear of the coinage and a government policy of benign neglect had eradicated many intended effects of the Great Recoinage. By 1770, Britain again used underweight and token coins, many of them privately issued. After 1787, Boulton’s steam press could be applied to make high-quality tokens. The government was slow to use this technology, but private firms were not. They soon issued substantial numbers of high-quality convertible token coins. They created a system of convertible token coins that the government eventually nationalized. That is how the standard formula was implemented in Britain.
The accidental standard
The standard formula takes for granted that a large denomination coin is the unit of account, and that the values of subsidiary coins are stated in terms of the larger coin. Before the standard formula was implemented in Britain, a large gold coin, the guinea, had already become the unit of account. But the eighteenth century began in England with a smaller silver coin being the unit of account and most everyone taking for granted that the guinea was a commodity whose price would fluctuate.
How did the guinea convert from commodity to unit of account? We do not know for sure. In this section, we describe some evidence of the government’s attitude toward the guinea that hints at how fluctuations in the price of the guinea raised concerns about their redistributional consequences. By the end of the seventeenth century, the government was trying to influence or regulate the price of the guinea, without much success. However, these regulations showed a concern with units of account that eventually led private contracts to be denominated in guineas. This section describes some of the government’s policies that aimed to influence or regulate the price of the guinea. It also describes some of Isaac Newton’s views on the market for gold and guineas.
Laws and ceilings
After the Great Recoinage, the government set a legal ceiling on the value of the guinea, and lowered it several times, until 1717. After that the ceiling value remained 21s. per guinea. That value became the definition of the unit of account in Britain.3 Accepting that unit of account put Britain on a de facto gold standard. By making a large denomination coin the unit of account, guinea-denominated debts made people think of small denominations as fractional coins, and prepared the way for the standard formula.
The guinea and legal tender laws
The government’s main purpose in setting a legal ceiling on the guinea was to protect the revenues it received when it accepted gold in payment of taxes at a high rate. Until 1696, the government had never tried directly to regulate the price of the guinea; indeed, during the crisis of 1695 it found that it had little power to do so.4 When the guinea was first issued, 20s. was treated as a reference or “par”; until the early 1690s, the price of the guinea was quoted as a premium over 20s.5 Government officials accepted the guinea at a discretionary rate that tracked market variations. The dashed line in figure 16.3 shows the rate at which Treasury tellers were instructed to accept guineas in payment for taxes or loans. 6 During the height of the crisis, the Treasury was reluctant to accept the guinea at its rapidly rising market value, inspiring the government to try to deprive the gold coin of its legal tender status for payment of taxes. But on June 1, 1695 the cashiers of the Exchequer stated that their refusal to accept guineas for more than 25s. “would not at all keep down the price in the town” and that it also interfered with marketing government loans. Thus, they were forced to take guineas at 29s. In July 1695, the cashier of the Excise was instructed to “discountenance taking guineas, but if there be a necessity he must not refuse them, as this might cause the King’s subjects not to bring in the tax; he must make the collectors think that guineas will not be taken for the next installment of the tax.” This last instruction reveals an attempt to discourage demand for the guinea. Tax collectors in Bristol found in October that “by reason of the badness and scarcity of silver it was impossible to make good their collections unless they were permitted to receive and pay guineas at 30s.,” which the government refused to do. On October 16, the king was informed of the impact on tax collection but told his officials to “hold it as long as they can.” The question of the rate at which to accept the guinea was set aside when the recoinage was being planned, and no measures were taken until the recoinage was under way.7
In 1696, the House of Commons began to regulate the guinea, by passing resolutions that put a maximum on the coin’s market price, first 28s. (February 15), then 26s. (February 26). An Act (7 & 8 Will. III c. 10, s. 18) confirmed this last value, by stipulating that after March 25, 1696 “noe Person shall receive take or pay any of the pieces of Gold Coine of this Kingdom commonly called Guineas att any greater or higher rate than 26 shillings for each Guinea” (Horsefield 1960, 81; Horton 1887, 243). The penalty for violating this law was a fine of £20 plus twice the value of the guineas so exchanged, and the fine was shared with the informant. But it was made clear that “nothing in this Act contained shall extend or be construed to compel any person or persons to receive any Guinea or Guineas at the said Rate of 26 Shillings.” The guinea was therefore not made legal tender at 26s. (or any other rate). The ceiling was lowered to 22s. on April 10, and remained there until 1717. These successive values are represented by the solid line in figure 16.3, and also serve as the basis for “par” after 1696 in figure 16.4.8
We do not know the effect of these laws on the market price of the guinea. 9 By 1698, however, the price of gold had fallen to a level consistent with the guinea at 22s. or less (fig. 16.4).
In October 1697, the Treasury decided to accept guineas at 21½s. in payments for loans only, but maintained the rate of 22s. for taxes (Horsefield 1960, 82). The following year, a report to Parliament, signed by John Locke among others, recommended lowering the guinea’s rate by 6d., because the gold/silver ratio was lower in the rest of Europe (Li 1963, 126–28). No legislation followed, but in February 1699 the Treasury decided to refuse guineas at more than 21½s. Although the legally enforceable ceiling of 22s. was not modified, the market price of gold immediately adjusted to the Treasury’s new buying rate, so no change in the ratio of that market price to “par” is visible in figure 16.4.
Finally, in December 1717 a royal proclamation was issued to “ forbid all persons to utter or receive any of the pieces of Gold, called Guineas, at any greater or higher rate than 21 shillings for each guinea … upon pain of our highest displeasure and upon pain of the greatest punishment that by law may be inflicted upon them for their default, negligence and contempt in this behalf” (Li 1963, 156).10
These are the laws upon which Britain’s monetary system was based until Lord Liverpool’s Coinage Act of 1816 made gold the basis of the British coinage, with gold coins unlimited tender at a fixed value.
Newton’s forecasts
Sir Isaac Newton’s writings tell us something about the purposes behind the government’s monetary actions in 1717, and whether it foresaw the consequences of its actions. Unlike Locke, Newton was a professional monetary authority. He was master of the mint from December 1699 until his death in 1727. Between 1701 and 1717 he advised the Treasury on monetary matters, including public policy about the price of the guinea (texts in Horton 1887 and Shaw [1896] 1967; see also Craig 1953, 212–29; Feavearyear 1963, 153–58; Li 1963, 143–60).
In 1702, Newton estimated that the guinea was higher in England than it would be in France, Holland, Germany, or Italy, and recommended lowering its rating by 6d., 9d., or 12d.; but he advised against an “alteration of the standard,” by which he meant the silver coinage (Shaw [1896] 1967, 137). In 1717, Newton estimated that the gold/silver ratio was 15.57 in Britain with the guinea at 21½s, compared to 14.8 to 15 in the rest of Europe, 12 in East India, and 9 or 10 in China and Japan (Shaw [1896] 1967, 168). He expected arbitrage to reduce these disparities and that the ratio in Britain would fall into line with the rest of the world. John Conduitt, who succeeded Newton as master of the mint, identified further reasons to expect the ratio to fall throughout Europe, including increased supplies of gold from Brazil, and increased demand for silver from the European affluent (Shaw [1896] 1967, 187). Those expectations were broadly confirmed, as figure 17.1 shows.11
Figure 17.1 Market ratio of gold to silver prices, London (dots); legal ratio (ratio of par prices); market ratio in Hamburg (annual averages, lines); 1694–1797. Sources: figures 16.4, 17.3 for London; Soetbeer 1879, 128–30 for Hamburg.
Newton suggested again that England’s ratio be brought closer to that in the rest of Europe, by reducing the value of the guinea by 6d. But he did not think that government action was necessary: “if things be let alone till silver money be a little scarcer, the Gold will fall of itself. For people are already backward to give Silver for Gold, and will in a little time refuse to make payments in Silver without a premium.… And so the question is whether Gold shall be lowered by the government, or let alone till it falls of itself by the want of silver money.”
Newton still regarded silver as the standard of value. According to Cantillon’s testimony, Newton rejected the suggestion to debase silver, saying: “according to the fundamental Laws of the Kingdom silver was the true and only monetary standard and as such it could not be altered” (cited in Li 1963, 159). Newton thought that even if gold coins were not valued properly by the government, they would be correctly priced by the market. Similarly, Locke in 1698 had written that it was “impossible, that more than one Metal should be the true Measure of Commerce,” that “Gold as well as other Metals is to be looked upon as a Commodity, which varying in its Price as other Commodities do, its Value will always be changeable” (Li 1963, 127).12
Gold becomes the unit of account
Belying Newton’s dictum that “silver was the true and only monetary standard,” the unit of account in Britain turned from silver to gold during the eighteenth century.
In 1717, Parliament had lowered the guinea by 6d., which brought the “legal ratio” to 15.2 (see fig. 17.1). The gold/silver ratio fell further. But no silver was minted, silver coinage went at a discount, and the guinea stayed at 21s. By the late eighteenth century, England was on a de facto gold standard: the unit of account (in Locke’s words, what “men in their bargains contract for”) had become an amount of gold, namely, 
of a guinea.
Figure 17.2 Annual minting of gold (left scale, solid line) and market price of gold bars as premium over par (right scale, dots), London, 1718–1800. Par is based on a guinea at 21½s. before December 1717, 21s. after (£3 19s. 8¾d. per troy oz of standard [22k] gold before December 1717, £3 17s. 10½d. after). Source: figure 17.3, Challis 1992a.
Underweight coins
Figure 17.2 presents evidence that the amount of gold in a guinea was not actually stablilized until the last quarter of the eighteenth century because wear and tear left a distribution of underweight coins.
Assuming that melting occasionally took place, the maximum value of the percentage premium of gold bars estimates the size of the interval between the minting and melting points. 13 Figure 17.2 shows that the width of the interval increased significantly after 1760 until 1774, when the government recoined underweight gold coins. The widening interval before 1774 reflects a gradual deterioration of the weight of the outstanding stock of gold coins in Britain. That the market price of bullion could deviate as much as 4% above par indicates that an average guinea, if valued at 21s., could contain as much as 4% less than its official content. Progressive wear and tear widened the distribution of weights among circulating coins, thereby widening the interval between the melting point and the minting point. A wider interval allowed the price of bullion to fluctuate more.
By 1774 it was estimated that the average guinea was 9% underweight (Li 1963, 163). In that year, Parliament ordered a recoinage of gold at government expense. That had the effect of reducing to a point the distribution of guineas by weight, as the recoinage of 1696 had for silver. It therefore equated the guinea to its official content, and the market price of gold remained slightly below par.14 Figure 17.2 confirms that the margin of movement for gold narrowed after 1774.
Figure 17.2 shows both the premium on gold bullion over par and minting volumes between 1718 and 1773. Although the coincidence is imperfect, before 1760 it appears that when the market price of bullion was markedly above the mint price, minting slowed. From 1760 to 1771 there was substantial minting of gold despite what was often a significant premium on gold (the price level was higher than the minting point for gold).
Figure 17.3 shows that the interval beteween the minting and melting points for silver coins widened even more for silver than for gold. By weighing bags of coins at the mint, Conduitt estimated that the wear on silver coinage ranged from 1 to 3.4% in 1728. The government’s policy toward silver, however, differed from that toward gold. The recoinage of 1774 only involved gold. Silver was left to depreciate further, so that by 1798, the wear on silver ranged from 3 to 38% (Li 1963, 163, quoting Liverpool). On the contrary, the government progressively deprived silver of its legal tender quality: between 1774 and 1783, and again between 1798 and 1816, it was limited to £25 by tale, and any amounts larger could be paid in silver by weight only. In the late 1790s, when silver bullion fell to a discount for a sustained period of time, the government acted temporarily to suspend the free coinage of silver.
Figure 17.3 Market price of silver bars in London 1713–1800, shown as deviation from par. Par is 5s. 2d. per troy oz of sterling (92.5%) silver. Sources: Commons (1811), Castaing (1713–1800), Shaw ([1896] 1967, 195).
Neglect the pence
We have already described how the British authorities tolerated private supplies of small change in the eighteenth century even in the form of counterfeit copper coins. Benign neglect extended to the silver coinage, which gradually became token, while the pound sterling, taking care of itself, became gold.
We remain ignorant about many details of the process by which Britain adopted the gold standard. However, an important piece of the story was the persistence of a stable exchange rate between the guinea and the subsidiary coins in the eighteenth century. This somehow emerged when the government got out of the business of supplying small change, the policy of benign neglect alluded to earlier. Thus, the path of the exchange rate of large for small coins in eighteenth-century Britain did not display the age-old persistent depreciations depicted in figure 2.1 on page 16. We interpret the absence of depreciations as evidence that somehow the entrepreneurs who produced small change managed to supply enough to avoid the shortages that our model registers as a binding penny-in-advance constraint.
To understand how the exchange rate between coins of different metals could remain constant, we use a modified version of our model. We allow underweight coins of both metals to coexist with full-weight coins, as described in chapters 16 and 22. The modified model allows a constant exchange rate between two coins made of different metals so long as two conditions are met: (1) that the intervals be broad enough to accommodate movements in the relative prices of gold and silver while still overlapping at a constant value of e; (2) that there occur no shortages of small denominations that would trigger a binding penny-in-advance constraint and an increase in e.15 The wear and tear of the coin stock can account for the first condition. We conjecture that the “benign neglect” policy resulted in sufficient amounts of small change to maintain the second condition. 16
Implementation of the standard formula
We described on page 271 how Boulton’s application of James Watt’s steam engine allowed private firms to issue high-quality convertible tokens. After 1787, many firms issued such tokens, and they formed much of the subsidiary coinage of Britain. Thus, by 1800, the market had chosen a gold coin as the unit of account and had also found a technology and set of private institutions to provide token subsidiary coins convertible into gold. Britain had mostly solved the problem of small change in practice. The complete triumph of the standard formula in Britain waited for laws to catch up with practice and for the government to nationalize the business of supplying small change.
With the Coinage Act of 1816, the British government began to implement the standard formula not just for small denomination copper or bronze coins, at the lowest end of the denomination scale, but for all of its silver coinage. In doing so, Britain implemented the first full-fledged gold standard. But it did so in a piecemeal fashion, rather than carrying out a preconceived plan.
Redish (1990) discusses in detail the adoption of the gold standard shortly after the installation of Watt’s steam-driven minting presses at the Royal Mint in 1805. The gold standard was officially adopted by the Coinage Act of 1816, also known as Liverpool’s Act (56 Geo. III c. 68), based on Lord Liverpool’s 1805 proposal. The act proclaimed gold coin to be the sole standard of value, and made silver coins representative; however, it still allowed for free minting of silver, at the existing mint price of 62s. per pound. Seigniorage, which had been abolished in 1666, was levied again on silver by reducing the weight of the coins from 62s. to 66s. per pound troy, and their legal tender was limited to £2. Finally, the Act of Suppression of 1817 (57 Geo. III. c. 46) made private coinage illegal. The Mint resumed copper coinage in 1820.17
Formally, the 1816 act did not completely implement the standard formula. A key element, that the small denominations be minted on government account, seemed to be contradicted by a clause of the act specifying that free minting of silver would begin at a date to be announced. But because the market price of silver remained below the mint’s legally set price of 62s., the mint forever refrained from announcing a date, lest it be flooded with silver. Free minting of silver was not actually removed from the statutes until 1870. In addition, convertibility of the token coinage was not explicitly mentioned. Instead, the Bank of England voluntarily adopted a policy of accepting silver coins at face value in exchange for gold or its notes. Only in 1836 did the Bank come to an arrangement with the Treasury to sell its inventory of silver coins at par value to the Mint, establishing convertibility in fact if not in law (Redish 2000, 152).
According to Redish (2000, 149), Mint officials believed that restricting the quantity of silver coins would be sufficient to maintain a constant exchange rate with gold coins, and it took the Bank some time to convince them otherwise. Four ingredients of the standard formula were thus progressively and quietly put in place from 1816 to 1836: (1) a token coinage; (2) a government monopoly of token coin production; (3) minting on government account; and (4) convertibility.
The gradual process by which Britain slid into practicing the standard formula rendered its monetary arrangements and practices ambiguous to some observers. Thus, in 1829, when the United States Congress debated changes to its monetary system, several influential voices rejected the British model, at least as they understood it. Jefferson’s former Secretary of the Treasury, Albert Gallatin, thought that British token money “is in fact, an issue of adulterated money which does not regulate itself” because it is “irredeemable” (cited in Redish 2000, 220). John Stuart Mill’s description of the British monetary system (1857, book III, ch. 10, sec. 2) lists the first three ingredients from the previous paragraph, but not convertibility.18 As late as 1918, Cannan (1935, 39) did not consider convertibility to be part of the English system, but rather a “slight improvement” made by other countries to Britain’s system, a system that he nevertheless deemed “perfectly successful.” Despite having no formal commitment to convertibility, Britain practiced the standard formula.
1 The importance for the gold standard of using token subsidiary coins is one of the important themes of Angela Redish (2000).
2 The temporary French reform of 1577 had made a large gold coin the unit of account as part of a set of policy measures to cope with depreciating silver coins (see chapter 11).
3 It replaced the 5s. 2d. per ounce of silver that had preceded it.
4 Proclamations on coinage did assign rates to other gold coins, which were declared to be “current” at certain values, but no penalties were edicted.
5 Examples in Houghton (1692–1702) before 1694, in Calendar of Treasury Books 1904–58, 9:30.
6 Calendar of Treasury Books 1904–58, 9:30, 44, 225, 266, 679, 750, 1038, 1195, 1310, 1554, 1650, 1805, 1899; 10:174, 521, 1383, 1395.
7 Calendar of Treasury Books 1904–58, 10:1383, 1395, 1409; Calendar of Treasury Papers 1868–89, 1:464.
8 The “par” value of a troy pound of standard gold was taken to be 44.5 times the value assigned to a guinea.
9 The source for figure 16.3 is a weekly newspaper called Collection for the Improvement of Husbandry and Trade. The last price for the guinea was for Feb. 4 until the guinea reappeared on April 14 at 22s., and remained unchanged until it ceased to be reported in 1699. But Horsefield (1960, 82) cited contemporary evidence that the legal ceiling was evaded. A buyer, knowing that a seller had a guinea, would bet 3s. with the seller that the seller couldn’t find him a guinea at 22s. The seller would promptly produce the guinea and the buyer, losing the bet, would pay 22 + 3s.
10 What penalties this clause actually implied is not clear. The proclamation was issued under the royal prerogative, albeit at the request of Parliament, but was not backed by statutory penalties.
11 We lack observations on the gold/silver ratio in London from 1698 to 1717, which is why we use the ratio in Hamburg as a proxy.
12 Locke regarded attempts at rigidly setting the value of gold to be “prejudicial to the Country which does it.” He explained why in his 1691 tract (Locke 1691, 162–64): setting the guinea above its value forces creditors to accept it, but setting it under its value doesn’t. In the latter case, either gold is hoarded or else the law is ignored, but in the former case gold is imported by debtors. See Garber (1986) for more about the option value of currency under bimetallism.
13 This is δ−1 in the model of chapter 22.
14 The effect of wear and tear on coinage continued in Britain and concerned Jevons (1875). Another recoinage at government expense took place in 1896.
15 Above, we have interpreted the rise in e observed in 1695 as reflecting a binding penny-in-advance constraint.
16 Feavearyear (1963) and others have also argued that Britain, as it grew more prosperous, needed less small coinage.
17 The coins were issued at 24d. per pound; the price of copper in 1820 was around 10d. per pound. Britain switched to bronze in 1860.
18 That convertibility would be sufficient to maintain the value of token small change had nevertheless been made clear by Jean-Baptiste Say in 1803: “If the government were to pay silver on demand for the copper coins presented, it could, almost without inconvenience, give them a very low intrinsic content.… There would only be counterfeiters to fear” (Say 1841, 262).
