Book IV, Chapter VI, 2

The same is true of coin. No one imagines, I believe, that even the greater part of the annual coinage—which for ten years before the recent reform of the gold coin amounted to upwards of £800,000 a year in gold—was added each year to the money already circulating in the kingdom. Where the government pays for coinage, the value of a coin, even one containing its full standard weight of gold or silver, can never greatly exceed that of the same quantity of uncoined metal: anyone can obtain coin for uncoined gold or silver merely by taking it to the mint and waiting, perhaps, a few weeks. Yet in every country most circulating coin is almost always worn to some degree, or otherwise fallen below the standard. In Great Britain before the recent reform, this was decidedly so: gold was more than two per cent., and silver more than eight per cent., below its standard weight. Now if forty-four guineas and a half, at their full standard weight of a pound of gold, could buy scarcely more than a pound of uncoined gold, forty-four guineas and a half that had lost some weight could not buy a pound without an additional payment to make up the difference. The market price of gold bullion, therefore, instead of equaling the mint price of £46:14:6, was then about £47:14s., and sometimes about £48. Yet when most coin was in this diminished state, forty four guineas and a half fresh from the mint bought no more goods than the same number of ordinary guineas. Once they entered a merchant's coffers and mingled with the rest, distinguishing them afterward would cost more trouble than the difference was worth. Like other guineas, they were worth no more than £46:14:6. Melted down, however, they yielded, with no perceptible loss, a pound of standard gold, which could always be sold for between £47:14s. and £48, in gold or silver equally suitable for all the purposes of coin as the metal just melted. Melting new coin thus offered an obvious profit, and happened so promptly that no government precaution could prevent it. The mint's operations were rather like Penelope's weaving: the day's work was undone at night. Instead of adding daily to the coin, the mint was chiefly replacing its finest part, which was melted down daily.

If private people bringing gold and silver to the mint had to pay for coinage themselves, it would increase the value of those metals just as workmanship increases the value of plate. Coined gold and silver would be worth more than uncoined. Provided it was not excessive, the seignorage would add the whole amount of the charge to the bullion's value, for the government everywhere holds the exclusive right to coin, and no coin can reach the market more cheaply than it chooses to supply it. If the charge were excessive, however—that is, far above the actual labor and expense of coinage—the great difference between bullion and coin might induce counterfeiters at home and abroad to flood the market with enough false money to lower the value of government coin. Yet in France, where the seignorage is eight per cent., no appreciable inconvenience of this sort appears. The dangers facing a counterfeiter who lives in the country whose coin he imitates, or his agents and correspondents if he lives abroad, are far too great to risk for a profit of six or seven per cent.

In France, seignorage raises the coin's value beyond what the amount of pure gold in it would warrant. The edict of January 1726, for example, fixed the mint price of fine gold of twenty-four carats at seven hundred and forty livres nine sous and one denier one-eleventh for the mark of eight Paris ounces. [See Dictionnaire des Monnoies, tom. ii. article Seigneurage, p. 439, par 81. Abbot de Bazinghen, Conseiller-Commissaire en la Cour des Monnoies à Paris.] Allowing for the mint's permitted variation, French gold coin contains twenty-one carats and three-fourths of fine gold and two carats one-fourth of alloy. A mark of standard gold is therefore worth only about six hundred and seventy-one livres ten deniers. In France, however, this mark is coined into thirty louis d’ors of twenty-four livres apiece, or seven hundred and twenty livres. Coinage thus increases the value of a mark of standard gold bullion by the difference between six hundred and seventy-one livres ten deniers and seven hundred and twenty livres: forty-eight livres nineteen sous and two deniers.

Seignorage will often eliminate the profit from melting new coin altogether, and will always reduce it. That profit arises from the difference between the bullion ordinary circulating coin ought to contain and the bullion it actually contains. If the difference is smaller than the seignorage, melting brings a loss; if equal, neither profit nor loss; if larger, some profit remains, but less than without seignorage. If, for example, before the recent reform of gold coin, coinage had carried a seignorage of five per cent., melting gold coin would have brought a loss of three per cent. At two per cent. seignorage there would have been neither profit nor loss. At one per cent. there would have been a profit of only one per cent., rather than two per cent. Wherever money is accepted by count rather than weight, seignorage is therefore the most effective safeguard against melting coin, and, for the same reason, against exporting it. It is generally the best and heaviest pieces that are melted or exported, since they yield the greatest profit.

The law encouraging coinage by making it duty-free was first enacted for a limited period under Charles II., and then extended repeatedly until 1769, when it was made permanent. To replenish its coffers, the bank of England is often obliged to take bullion to the mint; its directors probably thought it better that the government should pay for coinage than that they should. Probably to oblige this great company, the government agreed to make the law permanent. Yet if the practice of weighing gold falls out of use, as its inconvenience makes quite likely, and English gold coin again comes to be accepted by count as it was before the recent recoinage, this great company may discover that here, as on other occasions, it has seriously mistaken its own interest.

Before the recent recoinage, English gold currency was two per cent. below standard weight; with no seignorage, its value was therefore two per cent. below that of the standard gold bullion it ought to have contained. When this great company bought bullion to have it coined, it consequently paid two per cent. more for it than the resulting coin was worth. With a seignorage of two per cent., however, the common gold currency, though two per cent. short of standard weight, would still have equaled in value the standard quantity it ought to have contained: the value added by coinage would have made up for the lost weight. The bank would indeed have paid the seignorage of two per cent., so its loss on the entire transaction would have been exactly two per cent.—the same as its actual loss, and no greater.

If seignorage had been five per cent. while gold currency was only two per cent. below standard weight, the bank would have gained three per cent. on the price of bullion. But having to pay five per cent. seignorage for coinage, it would again have lost exactly two per cent. on the transaction as a whole.

If seignorage had been only one per cent. while gold currency was two per cent. below standard weight, the bank would have lost only one per cent. on the bullion's price. But it would also have paid one per cent. seignorage, making its loss on the entire transaction exactly two per cent., as in every other case.

If a reasonable seignorage were charged while coin held its full standard weight, as it has very nearly done since the recent recoinage, whatever the bank lost on seignorage it would gain on the bullion's price, and whatever it gained on the bullion's price it would lose on seignorage. On the transaction as a whole it would neither gain nor lose; here, as in all the preceding cases, its position would be exactly what it would have been without seignorage.

When a tax on a commodity is moderate enough not to encourage smuggling, a merchant dealing in it advances the tax but does not truly pay it: he recovers it in the commodity's price. The final purchaser or consumer pays it. Money, however, is a commodity in which every person is a merchant. No one buys it except to sell it again, and ordinarily there is no final purchaser or consumer. Thus when a tax on coinage is moderate enough not to encourage counterfeiting, everybody advances the tax but nobody finally pays it, because everybody recovers it through the coin's increased value.

A moderate seignorage, then, would never increase the expense of the bank or of any other private person bringing bullion to the mint for coinage; nor does the absence of such a seignorage ever reduce that expense. Whether seignorage is charged or not, if the currency contains its full standard weight, coinage costs nobody anything. If it falls short of that weight, coinage must always cost the difference between the bullion it ought to contain and the bullion it does contain.

When the government pays for coinage, therefore, it not only incurs a small expense but forfeits the small revenue a proper charge might bring; neither the bank nor any other private person benefits in the least from this useless act of public generosity.

The bank's directors, however, would probably be reluctant to agree to seignorage on the strength of a theory promising them no gain, merely protection against loss. Given the present state of gold coin, and while it continues to be accepted by weight, they would certainly gain nothing from the change. But if weighing gold coin ever goes out of fashion, as seems very likely, and the coin again falls as far below standard as it did before the recent recoinage, seignorage would probably bring the bank a very considerable gain—or, more properly, saving. The bank of England is the only company that sends a substantial quantity of bullion to the mint, and the burden of annual coinage falls wholly, or almost wholly, upon it. If annual coinage merely had to replace the inevitable losses and ordinary wear of coin, it would seldom exceed fifty thousand, or at most a hundred thousand pounds. But when coin has fallen below standard weight, annual coinage must also fill the great gaps that export and the melting pot continually leave in circulating money. For this reason, during the ten or twelve years immediately before the recent reform of gold coin, annual coinage averaged more than £850,000. A seignorage of four or five per cent. on gold coin would probably have stopped both export and melting effectively even under the conditions then prevailing. Instead of losing about two and a half per cent. each year on the bullion coined into more than eight hundred and fifty thousand pounds—an annual loss of more than £21,250 pounds—the bank probably would not have suffered a tenth of that loss.

Parliament allocates only fourteen thousand pounds a year to cover the expense of coinage; and I am assured that the actual cost to the government, the fees of the mint officers, ordinarily does not exceed half that amount. Saving so small a sum, or even gaining another unlikely to be much larger, might seem too insignificant to deserve the government's serious attention. But saving eighteen or twenty thousand pounds a year if an event occurs that is not improbable, has often happened before, and is very likely to happen again, surely deserves the serious attention even of a company as great as the bank of England.

Some of the preceding arguments and observations might perhaps have belonged more properly in the chapters of the first book concerning the origin and use of money and the difference between commodities' real and nominal prices. But since the law encouraging coinage springs from the common prejudices introduced by the mercantile system, I thought it more appropriate to reserve them for this chapter. Nothing could better suit the spirit of that system than a kind of bounty for producing money—the very thing it supposes to constitute the wealth of every nation. It is one of the system's many admirable devices for enriching the country.

The same is true of coins. I doubt anyone believes that even most of the coins minted each year were added to the money already circulating in the kingdom. For ten years before the recent reform of the gold coin, the annual minting of gold amounted to more than £800,000. Where the government pays for minting, a coin containing its full standard weight of gold or silver cannot be worth much more than the same amount of metal in bullion. Someone with uncoined metal need only take it to the mint and wait perhaps a few weeks to receive the same amount in coins. But in every country, most coins in circulation are almost always somewhat worn or otherwise below standard. In Great Britain before the recent reform, this was especially true: gold coins were more than two per cent. under standard weight, and silver coins more than eight per cent. under. Forty-four and a half guineas at full weight contain a pound weight of gold and could buy only slightly more than a pound weight of uncoined gold. Forty-four and a half underweight guineas could not buy a pound weight without an extra payment to make up the shortfall. So the market price of gold bullion was not the mint price of £46:14:6. It was then about £47:14s., and sometimes about £48. Yet when most coins were underweight, forty-four and a half guineas newly made at the mint bought no more goods than ordinary guineas. Once they entered a merchant's cash box and mixed with other coins, identifying them again took more trouble than the difference was worth. Like other guineas, they were worth only £46:14:6. Melt them down, however, and they yielded, with no noticeable loss, a pound weight of standard gold. That gold could always be sold for between £47:14s. and £48, in either gold or silver, and was as suitable for making coins as it had been before melting. Melting newly minted coins therefore brought an obvious profit, and happened so quickly that no government precaution could stop it. The mint's work was rather like Penelope's weaving: what was done by day was undone at night. Instead of adding much to the stock of coins each day, the mint mainly replaced the best coins, which were melted down each day.

If private people taking gold and silver to the mint had to pay for minting themselves, minting would add to the metals' value, just as workmanship adds value to silverware. Coined gold and silver would be worth more than uncoined metal. If the seignorage, or minting charge, was not excessive, it would add its full amount to the value of bullion. Since governments everywhere have the exclusive right to mint coins, coins cannot reach the market at a lower price than governments choose to charge. But an excessive charge—one much higher than the actual labor and expense of minting—could encourage counterfeiters at home and abroad. The large gap between the value of bullion and that of coins might lead them to supply so many false coins that the government's coins lost value. In France, however, the seignorage is eight per cent., and no noticeable problem of this kind has arisen. A counterfeiter living in the country whose coins he imitates faces serious risks, as do his agents or contacts if he lives abroad. Those risks are far too great for a profit of six or seven per cent.

In France, seignorage raises a coin's value beyond what the pure gold in it alone would justify. Under the edict of January 1726, the mint price for a mark of fine, twenty-four-carat gold—a mark being eight Paris ounces—was set at seven hundred and forty livres nine sous and one denier one-eleventh. [See Dictionnaire des Monnoies, tom. ii. article Seigneurage, p. 439, par 81. Abbot de Bazinghen, Conseiller-Commissaire en la Cour des Monnoies à Paris.] Allowing for the mint's permitted margin, French gold coins contain twenty-one and three-fourths carats of fine gold and two and one-fourth carats of alloy. A mark of gold at that standard is therefore worth only about six hundred and seventy-one livres ten deniers as bullion. But France mints that mark into thirty louis d’ors worth twenty-four livres each, or seven hundred and twenty livres altogether. Minting thus raises the value of a mark of standard gold bullion by the difference between six hundred and seventy-one livres ten deniers and seven hundred and twenty livres: forty-eight livres nineteen sous and two deniers.

A seignorage will often eliminate the profit from melting new coins, and will always reduce it. That profit comes from the difference between the bullion ordinary coins should contain and the amount they actually contain. If the difference is smaller than the seignorage, melting causes a loss rather than a gain. If the two are equal, there is neither gain nor loss. If the difference is greater, some profit remains, but less than without seignorage. Suppose, for instance, there had been a five per cent. seignorage before the recent reform of the gold coin. Melting gold coins would have caused a three per cent. loss. At two per cent. seignorage there would have been neither profit nor loss. At one per cent., the profit would have been only one per cent. rather than two per cent. Wherever money is accepted by number rather than by weight, seignorage is therefore the most effective way to prevent coins from being melted down and, for the same reason, exported. The best and heaviest coins are usually melted or exported, since they offer the largest profits.

A law encouraging coinage by making it free of duty was first passed for a limited period under Charles II. Later extensions kept it in force until 1769, when it became permanent. The bank of England frequently has to take bullion to the mint to refill its cash reserves. Its directors probably thought it better for the government to bear the minting expense than for the bank to do so. The government probably made the law permanent to please this powerful company. But weighing gold coins is inconvenient, and the practice will likely fall out of use. If English gold coins are then accepted by number, as they were before the recent recoinage, the bank may discover that here, as on some other occasions, it badly misunderstood its own interest.

Before the recent recoinage, English gold coins in circulation were two per cent. below standard weight. With no seignorage, they were consequently worth two per cent. less than the amount of standard gold bullion they should have contained. Whenever the bank bought gold bullion to have it minted, it had to pay two per cent. more for that bullion than the resulting coins were worth. With a two per cent. seignorage, however, ordinary gold coins would have been worth as much as the standard quantity of gold they should have contained, despite being two per cent. underweight. In this case the value added by minting would have offset the loss of weight. The bank would indeed have had to pay two per cent. seignorage. Its total loss would still have been two per cent.—exactly what it actually lost, and no more.

If seignorage had been five per cent. while circulating gold coins were only two per cent. underweight, the bank would have gained three per cent. on the bullion price. But it would have paid five per cent. seignorage on minting, making its total loss exactly two per cent. again.

If seignorage had been only one per cent. and circulating gold coins two per cent. underweight, the bank would have lost only one per cent. on the bullion price. It would also have paid one per cent. seignorage, so its total loss would have been exactly two per cent., as in the other cases.

Suppose there were a reasonable seignorage and coins contained their full standard weight, as they have nearly done since the recent recoinage. Whatever the bank lost through seignorage, it would gain on the bullion price; whatever it gained on the bullion price, it would lose through seignorage. Overall it would neither gain nor lose. Here, as in the earlier cases, its position would be exactly the same as with no seignorage.

If a tax on a commodity is low enough not to encourage smuggling, the merchant selling it advances the tax but does not really pay it. The merchant gets it back in the commodity's price. The last purchaser or consumer ultimately pays. Money, however, is a commodity that everyone deals in. People acquire it only to pass it on again. In ordinary circumstances it has no last purchaser or consumer. If a minting tax is low enough not to encourage counterfeiting, everyone advances it, but nobody ultimately pays it. Each person recovers it through the coin's higher value.

A moderate seignorage, then, would never increase the expense borne by the bank or by other private people bringing bullion to the mint. Nor does having no moderate seignorage reduce their expense. Whether seignorage exists or not, minting costs nobody anything if coins have their full standard weight. If they fall short of it, minting always costs the difference between the bullion they ought to contain and the bullion they actually contain.

By paying for minting, the government bears a small expense and gives up the small revenue it could collect through a suitable charge. Neither the bank nor any other private party gains the slightest benefit from this pointless public generosity.

The bank's directors would probably hesitate to accept a seignorage based on an argument promising no gain, only protection from loss. With gold coins as they now are, and as long as they are accepted by weight, such a change would certainly bring them no gain. But weighing gold coins will likely fall out of use. If the coins then become as degraded as they were before the recent recoinage, a seignorage would probably save the bank a great deal. The bank of England is the only company that sends any substantial amount of bullion to the mint. It bears all, or nearly all, of the burden of annual coinage. If new coins were needed only to replace those unavoidably lost and those worn out in ordinary use, the annual amount minted would seldom exceed fifty thousand pounds, or at most a hundred thousand pounds. When coins fall below standard weight, however, the mint must also replace the large amounts continually removed from circulation by export and melting. For that reason, during the ten or twelve years immediately before the recent gold-coin reform, annual coinage averaged more than £850,000. A seignorage of four or five per cent. on gold coins would probably have stopped both exporting and melting effectively, even under the conditions then prevailing. Instead of losing about two and a half per cent. each year on bullion minted into more than eight hundred and fifty thousand pounds—an annual loss of more than £21,250 pounds—the bank would probably have lost less than a tenth as much.

Parliament assigns only fourteen thousand pounds a year to cover minting costs. I am told that the actual cost to the government, or the mint officers' fees, normally does not exceed half that amount. Saving such a small sum, or gaining another sum unlikely to be much larger, might seem too minor to warrant serious government attention. But saving eighteen or twenty thousand pounds a year if an event occurs that is quite possible, has often happened before, and is likely to happen again surely deserves serious attention, even from a company as large as the bank of England.

Some of these arguments and observations might have fit better in the chapters of the first book about the origin and use of money and the difference between real and nominal commodity prices. But the law promoting coinage comes from the common prejudices introduced by the mercantile system. I thought it better to keep the discussion for this chapter. Nothing suits that system better than a kind of bounty for producing money, the very thing it imagines makes every nation wealthy. This is one of its many remarkable devices for enriching a country.