Ideas, Policies, and Outcomes
Evolutions of ideas and institutions
James Laurence Laughlin (1931, 87) concluded his textbook exposition of the standard formula as follows:
It might seem at first blush that, as subsidiary moneys play only a secondary part, the principles governing them are not of first importance. On the contrary, the questions raised in early monetary experience down to the present day of necessarily combining different metals of varying relative values in one monetary system have led to the evolution of general principles of far-reaching influence.
Laughlin cited the role of these principles in managing the silver money stocks in formerly bimetallic countries, and also in maintaining “the value of paper and credit representatives of the established metallic standard” and in supporting “devices for avoiding the actual transfer of the valuable standard metal” such as paper money, banknotes, checks, and bills of exchange.
Contemporaries of Laughlin also sought timely lessons from episodes in the history of small change. The German hyperinflation of 1922–23 prompted German scholars to study how medieval jurists had dealt with units of account during periods of fluctuations in exchange rates. We owe to Germany’s 1923 hyperinflation a wealth of scholarship devoted to those concerns (Nussbaum 1950, 171–80, 217; see also Täuber 1933, 20–29).
This chapter summarizes what we have learned from studying the history of the big problem of small change. We recall themes that connect many episodes.
Our story began after Charlemagne first issued the silver penny, which for hundreds of years was the only coin and unit of account in Europe. In the 1200s, states began to issue at first silver and then gold coins in larger denominations. The unit of account remained the penny. The larger denominations contained more silver than the penny, and were meant to be that much more valuable; coins of different denominations were supposed to exchange for one another like lumps of silver of different sizes. Governments set up a mechanism to supply the coins needed for trade. They chartered mints and instructed them to sell (but not buy) particular coins for specified amounts of silver. The public decided how many coins to mint and how many to export or to melt. This mechanism was designed automatically to supply the amounts of coins that traders wanted.
But almost from the beginning, that supply mechanism produced shortages of the smaller denomination coins. Those shortages were often accompanied by appreciations of the large denomination coins in terms of the penny. The administrators of the mints learned how to cure shortages, at least temporarily, by debasing small denomination coins. Debasements provided incentives for people to bring silver to the mint to purchase new small coins. That cure for shortages produced centuries of appreciating rates of exchange of the large for the small denomination coins. Those debasements also caused price levels to drift upward, because pennies were the unit of account in which prices were stated. A principle cause of many debasements was not governments’ thirst for seigniorage revenues, but their wish to provide adequate supplies of small denomination coins.
Recurring appreciations of large denomination coins relative to small denomination coins preoccupied lawyers who confronted conflicts about the units of account used in contracts. Beginning in the twelfth century, secular lawyers created a body of opinion about money, mainly by selecting from among conflicting passages that they found within the body of Roman law that had been systematized under Justinian. Roman law and ancient commentaries on it had expressed an array of views on money. They ranged from a commodity view that money acquired value from the metal within it, to the view that money was worth more than its constituent material because it facilitated exchange by alleviating the absence of double coincidence of wants, a view that had been expressed by the Roman jurist Paulus. The medieval lawyers chose to ignore Paulus and to adopt the view that money is a commodity whose value is determined like that of any other commodity. They thought that people bargained and contracted for the amounts of silver within coins, not the names on those coins.
Fluctuations in exchange rates of different denomination coins and features of the technology for producing coins bedeviled the medieval lawyers’ theory. Small denomination coins cost proportionately more resources to produce than large ones. Allowing producers of coins to recover their costs and governments to tax the coin producers put a wedge between the value of the metal inside and the exchange value of newly minted coins, for otherwise no one would have the incentive to take silver to the mint to purchase coins.1 Because the costs of production were greater, wedges were typically larger for small coins than for large. Those wedges meant that coins were actually valued partly by tale (i.e., in proportion to their number rather than their weight) and made room for exchange rate fluctuations between coins of different denominations, even in the absence of any change in their metallic contents.
For centuries medieval and Renaissance lawyers and monetary theorists tried to apply an idealized theory that held that money is a commodity to facts indicating that the market often valued coins partly by tale. The theory had to adapt to fit the facts. Theorists emerged who recognized and even made virtues of the facts. Eventually, they used their new theories to make proposals to amend the medieval money supply mechanism in ways that would avoid the recurrent shortages of small change.
Those theorists did not work in a vacuum. They sought to explain outcomes of accidental and purposeful monetary experiments. Lawyers adjudicating disputes occasioned by exchange rate fluctuations sometimes argued that what had implicitly been contracted for were not particular quantities of silver but values of a more comprehensive bundle of goods. That pushed value in exchange to the foreground, not weight. Towns under siege issued tokens of various forms. Although some of them appear to have been backed by a leader’s promise that they would be redeemed for metal coins after the siege was lifted, others seem to have been unbacked. During a siege, token coins were valued, even when they were not backed by an explicit promise of convertibility, supporting Paulus’s ancient observations.
Partly inspired by such observations, theorists discussed circumstances under which it would be wise or just to issue token coins. From experiences with siege monies, they knew that counterfeiters threatened to undermine a system of token coins. As late as the seventeenth century, coins continued to be created by the same ancient technology of hammering them by hand from dies. That technology required little capital, so that the costs of entry into the counterfeiting business were low. The technology produced low quality coins that were easy to copy. The medieval way of deterring counterfeiting was simply to abstain from issuing token coins and to let the high cost in terms of metal content of making full-bodied coins deter counterfeiters.
Experiments
Disturbances to supplies of small change in the sixteenth century brought new ideas and experiments to France and Spain. In France, units of account shortages of small change played an important part in generating the persistent rise in the price level of the sixteenth and seventeenth centuries called the Price Revolution. The inflation was partly caused by inflows of Spanish treasure into Europe from the New World, which drove down the relative prices of both gold and silver in terms of goods in general. But more inflation occurred than could be accounted for by cheaper gold and silver alone. The penny remained the unit of account for most of the sixteenth century in France. Exchange rates for larger silver and gold coins appreciated persistently, to the accompaniment of recurring shortages and debasements of small denomination coins. Those debasements and depreciations of the small coins accounted for perhaps half of the inflation in France during the sixteenth century. The exchange rate fluctuations between large and small denomination coins provoked disputes about the unit of account.
In response, in 1577 the government imposed a reform that made a large coin, the gold écu, the unit of account. The French government also experimented with issuing token small denomination coins in small amounts. These were remarkably progressive reforms, because making the large coin the unit of account and using small coins as tokens were to be crucial elements in what two centuries later would become a standard formula for solving the problem of small change. But for some reason, after 25 years, France abandoned the reforms of 1577 and returned to the medieval mechanism. The unit of account reverted to the penny and token small coins were abandoned.
In 1596, Castile started a longer and more comprehensive experiment that tested elements of the standard formula, including token small change. Fiscal exigencies motivated Castile’s monetary experiments. For years before 1596, it had been known that a government could raise resources by replacing small denomination silver coins with token coins. But the prevailing hammering technology had made token coins easy to counterfeit, which meant that a government would lose much of the potential revenue to counterfeiters. In the late sixteenth century an expensive new technology for producing high-quality coins arrived, the cylinder press. Knowing how token coins could be substituted for silver coins without causing inflation, the monetary advisors of King Philip II of Spain proposed that the cylinder press be used to make inconvertible token copper coins. Philip II and his successors performed that experiment.
The experiment succeeded for twenty-five years. Substantial numbers of token coins were issued without deteriorating in value relative to silver coins, raising substantial revenues for the government in the process. However, eventually so many of these small denomination token coins were issued that inflation began. The Spanish monetary authorities then struggled for half a century, sometimes to arrest inflation, and sometimes to extract more revenues from the token coinage. In the process, they performed marvelous experiments and tried to manipulate exchange rates. After the experiment was over in 1660, Spain renounced token coins, and returned to the medieval monetary system.
The Castilians had tried the “token” part of the standard formula, but had omitted convertibility. In 1661, the British monetary official Sir Henry Slingsby understood how convertibility would automatically have regulated the quantity of token coins issued and thereby prevented the Castilian inflation. The advisors to the king of Spain hadn’t recognized the role of convertibility, or had recognized it too late. But the Castilian experiment generated valuable time series data that led empiricists to draw an association between the quantity of such token coins issued and the price level. The wreckage of the Castilian experiment gave token coins a bad name, but created an important database that helped analysts like Sir William Petty induce the quantity theory of money.
Shortages of small change in seventeenth- and eighteenth-century Britain led to further and more successful experiments with elements of the standard formula. During the seventeenth century, under the opportunistic Stuart kings, Britain let private firms and cities supply tokens. Those tokens formed much of Britain’s small denomination coinage. Sometimes the firms or cities offered to convert those tokens into silver. Towards the end of the period that those tokens flourished, in the 1660s, there emerged a new technology for making milled coins that were difficult to counterfeit. By deterring counterfeiting, that technology improved the prospects for implementing a convertible token coinage. The government under the Stuart kings used that technology to produce some silver coins, but not to produce tokens. Perhaps that was because to implement a system of convertible token coins, the public had to believe the government’s commitment to redeem the tokens into full-bodied money on demand. As North and Weingast (1989) have reminded us, the Stuart kings could not be trusted.
Presumably the new regime formed by Parliament and King William III after the Glorious Revolution of 1688 was more trustworthy. Ironically that new regime never contemplated issuing token coins, precisely because it cared about credibility in monetary and fiscal affairs. At the time, John Locke made a famous argument that by restoring the weight of the small denomination coins, the government would be upholding the sanctity of contracts. Locke reiterated the medieval view that when people signed contracts for deferred payment of coins, they understood a coin to be a quantity of metal. For asserting that, the statesman Lord Liverpool and the Whig historian Lord Macaulay continued to praise Locke in the mid-nineteenth century.
In 1696, Parliament in Britain went beyond Locke’s recommendations, and recoined at the old weights at government expense. It thereby incurred expenses that accounted for a considerable fraction of its budget in the midst of a European war. The new regime thus expended considerable resources to restore a full-bodied coinage and to reinstate the medieval monetary mechanism. It thereby turned its back on the progress that the market had made in moving toward the standard formula and reinstated full-bodied and revalued silver small denomination coins.
Despite the setback of the Great Recoinage, the progress that Britain had made toward the standard formula under the Stuart kings resumed again during the second quarter of the eighteenth century. Wear and tear on the coins and a neglectful government policy toward subsidiary coinage allowed counterfeits and private issues of subsidiary coinage to proliferate. After 1720, the exchange rates of large for small denomination coins stabilized. By mid-century, the guinea, a gold coin, had replaced silver coins as the unit of account. That transformation in the unit of account came not from government legislation but from the decisions of private contractors. In 1787, the invention of Boulton’s steam press made it feasible to issue token coins that were difficult to counterfeit. Private firms soon exploited the new technology to issue convertible token coins. After a generation in which those coins served as the small change in Britain, the government nationalized the business of supplying small change, thereby practically implementing the standard formula. Meanwhile, the standard formula was more deliberately adopted by the German monetary union, which remained on a silver standard. Later in the 19th century, France and the United States embraced the standard formula when they abandoned bimetallism for the gold standard.
Major themes
We conclude part IV by recalling recurring themes.
Beliefs and interests
Much of modern economics and political economy assumes that while people might have diverse interests, they nevertheless share a common and correct model of the economy. That all decision makers know the correct model underlies both the Nash and the rational expectations equilibrium concepts. Different people might prefer different policies, but all know their consequences.
Though the formal model in part V assumes rational expectations, our narrative assumes that policy makers frequently used incorrect or incomplete models. Only after a long process of learning did society and policy makers discover what we think is a correct model. Only then were they able to set up a good mechanism for managing supplies of coins. Much of our story is about how intelligent and well-informed people used and debated different models, contributing to a social process of model adaptation and discovery.
That process of model adaptation and discovery would be difficult to formalize, and we have not tried to do. Most formal theories of learning in economics study easier problems in which people know the model. Models that use Bayes’ law assume that people know a correct specification or set of specifications that they never alter. The accumulation of evidence allows them to sharpen estimates and resolve uncertainty, but does not prompt them to uncover new specifications. As Marimon (1997) noted, a Bayesian learner knows the truth from the beginning: he may lack data but believes he has an adequate model specification.
Disputes about models were central to many episodes and propelled a slow social process of learning that eventually led societies to abandon commodity money and to embrace the idea of token money. Progress occurred when someone proposed a new model, or added a new feature to an old one. We have not modeled those leaps of imagination and paradigm shifts, but have observed them in our history.2
Thus, we assign an important role to model disagreement and discovery in explaining outcomes,3 while modern political economy assumes away any such disagreements and explains outcomes as reflecting the polity’s mechanism for resolving conflicts of interest. Conflicts of interest play a role in our story too, but they are not the whole story. Numerous episodes of inflation in units of account set debtors against creditors. The debate about the Great Recoinage between Locke, Lowndes, and Newton brought out both their different models and their different preferences about the distributional consequences of proposals to recoin at an historic or a depreciated rate.
Units of account and nominal contracts
How people choose to denominate contracts is an important question in modern macroeconomics. “Price stickiness” and “nominal rigidities,” if they exist, emerge from contracting decisions. An enduring question in macroeconomics, often posed as a challenge to particular models of sticky prices, is why people don’t index contracts against inflation.
That question runs through our narrative. By focusing on the content of the coins initially lent and the intent of the parties, medieval law seemed equipped to index contracts and protect parties from fluctuations in exchange rates. Yet when exchange rate stability prevailed, contracts ceased to take note of the form in which money was lent, and parties seemed happy enough to denominate all transactions in the same unit of account. Only when rates fluctuated again was it urgent to reconsider conventional units of account and index the unit to one coin or another.
At times, the monetary authorities tried in vain to defend a hitherto conventional unit of account by legislating the exchange rate. Their efforts invariably failed to affect the exchange rates that traders actually used, and succeeded only in producing a string of ghost monies, units of account that let traders evade government-mandated units of account.
By the sixteenth century, jurists were assigning to the public authorities the power to regulate the unit of account. Those public authorities saw it as their duty to provide a stable unit of account in which all transactions, small or large, could be denominated. That was the intention of the French reform of 1577, with the innovative shift from the small coin to the large coin as unit of account. At the same time, theorists described money, not as a lump of metal, but as “the measure of all things,” and struggled with ways to keep it as constant as the yard or the ounce, a concern not shared by their medieval predecessors.4 But the inflationary experiences of the seventeenth century also made those authorities reluctant to relinquish the tight constraints on the price level that full-bodied coinage provided. Their desire to maintain a constant vector of exchange rates and full-bodied coinage for both gold and silver ultimately gave birth to bimetallism.
John Locke said that although people made contracts in units of account, by a unit of account they meant a fixed quantity of metal. In his view, it was the government’s duty to fix that quantity, and thereby to uphold the intentions of all parties in a way that private law could not. Disregarding Locke, who had said that people bargain for weights of metal and not “for sounds” (see page 284), on June 5, 1933 the U.S. Congress declared that contracting for a fixed quantity of metal was “against public policy” (48 Stat. 112) and at one fell swoop substituted sounds for intrinsic value.5
Small change and monetary theory
The evolution of monetary doctrines about small change was an important part of the process by which a managed fiat currency system came to be understood and implemented. Apart from the inessential detail of the substance on which a “promise” is printed, a token coin is like a paper bank note. A token coin is an IOU written on a piece of metal whose intrinsic value is less than the amount owed. At first, token coins were issued by local governments and private firms, as well as by some national governments. Eventually, national governments monopolized the issuing of banknotes and token coins. Thinking about the feasibility of a system of token small coins uncovered and refined the quantity theory of money as well as its main rival, Adam Smith’s real bills doctrine (page 101).
Currency boards, dollarization, and the standard formula
Many of the theoretical justifications for the standard formula apply when modern policy makers consider the drastic step of abandoning their national currency or the somewhat less drastic step of linking their national currency to a foreign currency via a currency board. A currency board behaves much like the managers of small change under the standard formula: it issues and redeems tokens convertible into foreign currency. Delegating monetary policy to a currency board is supposed somehow to alter a government’s incentives in a way that increases the markets’ trust in the future value of its currency. (Recall that an essential prerequisite for the standard formula was having a government that could be trusted to redeem tokens for full-bodied money, and that it took a long time before the idea was accepted that governments could be so trusted.) Like token subsidiary coinage, the notes of a currency board tie up fewer resources in a nation’s currency stock than does a fully “dollarized” system. In a dollarized system, a government exits the business of creating token currencies and uses the currency issued by a foreign government.6 Thus, the choice of a currency board versus dollarization involves the same basic theoretical principles that his advisors taught King Philip II.7
Learning by markets and by governments
Both governments and private sectors played important roles in solving problems and pushing forward the learning process. A slogan in economics is that if something is a big problem, the market will adapt to solve it. We have encountered examples that validate that bromide. One set of examples were the repeated “invasions” of foreign coins that were the market’s way of implementing a spontaneous debasement to cure shortages of small coins in the absence of timely government action. A second set of examples occurred when the market created ghost monies, units of account that private traders used to circumvent a government-mandated unit of account. Third, firms issued tokens, and eventually convertible tokens, thereby giving the government working examples of the standard formula. Since convertible tokens are interest-free IOUs, firms had incentives to issue them.
Although the standard formula calls for the government to take the lead in administering a system of token coinage, it was actually the market that took substantial steps toward the standard formula in seventeenth– and eighteenth–century Britain. Ultimately the British government implemented the standard formula by nationalizing a smoothly operating system of privately issued tokens.
1 This made the wedge necessary. What made the wedge possible was that coins fulfilled a particular purpose that could not be served by uncoined metal.
2 See Kreps (1998) for a discussion of the issues. Kreps sketched a learning process that features model respecification and experimentation.
3 See Sargent (1999) for a similar perspective on the U.S. inflation and deflation of the 1970s and 1980s.
4 Medieval philosophers thought of money as a measure (Kaye 1998), as did Oresme. This led them to condemn debasements as an alteration of this measure, not necessarily to advocate policies to maintain its constancy (see the footnote on page 186 for an exception).
5 The Duchess paraphrased Lowndes’s dictum (see page 291): “take care of the sense, and the sounds will take care of themselves.” Alice in Wonderland, ch. 9.
6 By the act of dollarizing, the government in effect performs the reverse of the “ small country” experiment of Adam Smith that we described on page 101.
7 See page 231.