The puzzle of electrum coins

From the Israel Museum in Jerusalem’s 2013 exhibition White Gold

 [Originally published at Bullionstar.]

For several years Brits have been hearing rumours that their 1p and 2p coins were on the cusp of being discontinued. Not so. Last month the UK Treasury announced its commitment to both coins. The 1 and 2p coins will continue to be produced for ‘years to come.’

Few bits of monetary technology have enjoyed as long an existence as the coin. The earliest coins were produced around 640 BC, some 2600 years ago, by the Lydians, who had built an empire in the western half of what is now Turkey.

To most of us, the usefulness of coins is self-evident. Sure, small coins like the 1p are a bit of a nuisance. They tend to accumulate in our pockets or piggy banks, never used. But compared to barter, or exchanging bits of unrefined metal, coins are a much better alternative.

One would assume that’s why the Lydians created coins in the first place: convenience. But the true story is much more puzzling than that. To this day we don’t entirely know why the Lydians began to turn precious metals into circular discs.

The traditional origin story for coins

The classic story for the adoption of coinage involves the efficiency gains that society enjoys when trade can be conducted by tale rather than by weight. Tale is a sum or a tally. All modern payments are done by tale. A payor counts up the right amount of coins (or notes), then passes the stack to the payee who – if they wish – can glance at the inscription on each coin’s face to ensure that it is legitimate. Circulation by tale is a convenient way of doing business.

But we take it for granted. Before coins appeared on the scene 2000 years ago, numismatists believe that people typically transacted with silver ingots and bars, otherwise known as hacksilber. These pieces could be cut up into smaller amounts in order to cover a range of different transaction sizes.

Because the bits of hacksilber were irregularly shaped, or non-fungible, they couldn’t by counted. Rather, they had to be weighed first, and only then could the transaction proceed. Weighing different bits of silver is a laborious process. A scale must be produced along with a set of weights that both the buyer and seller can trust.

Counting is much easier than weighing. If the stamp on the coin is reliable, buyers and sellers can trust to issuer to have already pre-weighed and standardized the metal for them. And so coinage would have dramatically reduced lineups and waiting time in busy markets all across the ancient world. What a fantastic invention.

Perfectly standardized

At first glance, Lydian coins have all the hallmarks of this classical origin story.

To begin with, they are quite beautiful. Each coin was typically stamped on the obverse side with a design in the form of an animal, human, or myth. On the reverse, or back-side of the coin, a square or rectangular design appears (see image at top). Did these designs constitute some sort of official guarantee of the coin’s weight and fineness? Or did they symbolize something else?

The coins generally lacked any sort of writing on them. Numismatists are thus unsure who actually issued the coins. Was it the city, the king, a merchant or some other rich individual?

One fact that all numismatists agree on is that the Lydians were assiduous to a fault about ensuring standardized weights for their coins. The biggest denomination, the stater, weighed 14.1 – 14.3 grams. Half staters contained half as much metal, followed by third staters (or trites), 1/6, 1/12, 1/24, 1/48, and 1/96th staters, the last of which contain just 0.15 grams of metal.

Smoothed distribution of Lydian coin weights around each denomination. Source: On the Origin of Specie, (2012)

Francois Velde, an economist at the Federal Reserve who dabbles in numismatics, has catalogued thousands of Lydian coins owned by private collectors and museums around the world. Using this data, one can see the remarkable precision of Lydian coinage (see chart above). The weight of the largest coins – staters and trites – tend to be tightly clumped near the standard weight.

Interestingly, the smallest denominations – the 1/96th staters – are much more loosely distributed around the standard weight (see the dark blue line). Velde (2012) attributes some of the lower accuracy of smaller denominations to the fact that they would have circulated more, and thus deteriorated faster.

The inconvenience of electrum

By carefully calibrating the weights of each denomination and stamping them with a seal, surely Lydia qualifies as the first society to make the technological leap to circulation by tale. But it’s here that the story begins to fall apart.

One of the curious facts of early Lydian coins is that they were made from a material called electrum. Electrum is a naturally occurring alloy of silver and gold, often found in streams and rivers. The problem with natural electrum is that the mix between gold and silver is variable. The silver content can be anywhere from 10% to 30%, according to numismatist Robert Wallace (1987).

Given this variability, Lydians must have had difficulties valuing electrum. A given electrum coin wasn’t fungible, or interchangeable, with its cousins. A coin with more gold in it would have a slightly different colour than one with less gold, as the chart below implies. And since gold was probably worth around 10 times more than silver in ancient times, electrum coins with more gold in them would have had a much higher intrinsic value than those with less. But how much more? According to Wallace, this lack of certainty would have caused “endless doubts and disputes over particular coins."

Approximate colours of Ag–Au–Cu alloys [Wikipidia]

What a contradiction Lydian coins are! The Lydians had evidently gone to extreme lengths to perfectly calibrate coin weights, and thus potentially exchange coins by tale, only to undo all the benefits of standardization by making coins with an arbitrary gold-silver mixture. Now buyers and sellers would have to settle on some laborious means of determining a given coin’s mixture, say like using a touchstone, before they could consummate a trade.

The Lydians could have avoided this problem at the outset by issuing coins using silver rather than electrum. Silver, after all, was already traded in ingot form. With silver coins, at least there would be no confusion about intrinsic value. But the Lydians chose not to go this route.

Which leads us to what may be the most popular theory for electrum coins, what I will call the “token" theory.

Electrum coins as tokens

It is Robert Wallace who can be credited with creating what is probably the most widely-accepted theory for electrum coins. Wallace (1987) began by imagining himself in the shoes of an owner of an electrum hoard around 640 BC. This individual had the following problem. His stash of metal was not uniform, and so fellow Lydians didn’t really trust its quality. How could our electrum owner get his suspicious counterparts to accept his metal for its full value?

The easiest solution available to our Lydian would be to refine his electrum into its silver and gold constituents, then sell each separately. But Wallace tells us that the technology for “parting" gold and silver – cementation – would not be available for almost a hundred years, circa 550 BC. So our electrum owner was stuck with his mongrel metal.

According to Wallace, our Lydian stumbled on the solution: turn his raw electrum into stamped coins. Why would a potential buyer trust electrum in coin form but not bar form? The answer is that the owner of electrum didn’t create just any regular coin. Rather than issuing discs that were valued for their (uncertain) metal content, our Lydian electrum owner designed them as tokens.

Electrum coin from Ephesus, 625-600 BC with a stag grazing [source]

A stamped piece of metal can be valuable either because of the material of which it is made or the symbol that is stamped on its face. A token is of the latter sort. By contrast, a piece of hacksilber is the former. It gets its value from the silver itself.

How did a mere stamp create value? Wallace hypothesizes that the issuer’s stamp indicated a promise to “accept back or redeem his coins" at a fixed rate. A skeptical buyer would therefore have no problem receiving an electrum token in trade. After all, the stamp guaranteed that the issuer would buy it back at that very same rate.

Fungibility regained

By setting his redemption price for tokens high enough, the issuer ensured that the market value of his coins would always exceed their intrinsic electrum value. This would have had the beneficial effect of making all his electrum coins fungible. After all, since both a silver-rich electrum token and a gold-rich one could both be redeemed at the issuer for the same fixed price, neither coin was any better than the other.

Electrum could now circulate freely rather than being handicapped by non-uniformity. The decision to turn electrum into coinage had converted “stocks of what was otherwise a doubtful and uncertain substance into negotiable currency large and fixed value," says Wallace. In the process, our electrum owner had become a much wealthier man than might otherwise have been the case.

So what about circulation by tale?

Despite the fact that the weight of electrum coins was so precisely calibrated, numismatists believe that Lydians exchanged the coins by weight rather than by tale, much as they had with hacksilber. The main bit of evidence for this is that electrum coins were never clipped.

Clipping is when someone scrapes or snips a bit of metal off of a coin before passing it on. The clipper keeps the shavings for themselves. Coins that circulate by tale are easily attacked by clippers. Since sellers will accept coins with little more than a glance to the stamp on the coin’s face, a buyer who scrapes off a bit of metal before handing the coin can easily get away with it.

A lack of clipping is consistent with the practice of weighing coins and only accepting those that are up to snuff. If a coin was even a bit too light, then the seller would not take it. And so no one would bother clipping them in the first place.

But if electrum coins circulated by weight and not tale, this hardly seem like a technological improvement over hacksilber. Lydian trade was still as slow and awkward as before.

However, the necessity of weighing electrum coins may have served a purpose. It may have been a security feature designed to protect the issuer’s wealth. Imagine that our issuer of electrum tokens has spent some staters into circulation. Prior to being returned to him for redemption, these staters had all been clipped. Since he has less electrum than what he started out with,  our issuer’s wealth has deteriorated.

To protect himself from this sort of theft, Wallace (1989) suggests that the issuer wouldn’t redeem just any of his tokens. As a security measure, he would only take back those that were still of their original weight. Since no merchant would want to be stuck holding a coin that could not be redeemed, they would always weigh each coin that was offered to them in order to avoid accepting light ones.

They only circulated domestically

Wallace’s theory explains another odd feature of electrum coins. Given the distribution of electrum coin hoards, numismatists believe that they didn’t circulate outside of their area of production. This is unusual for ancient coinage. Roman coins have been found as far afield as Sumatra, while Sassanian coins (minted in modern day Iran) have been unearthed in England.

But if circulation of electrum coins was premised on the guarantee that their issuer would redeem them, then that explains why they wouldn’t have circulated very far. People in a distant city would not recognize or trust the redemption promise of an unknown issuer, and so they wouldn’t accept them in trade.

Electrum diluted with silver

Another oddity of electrum coins is that they often contain far more silver than the natural electrum out of which they were manufactured. Electrum found in naturally-occurring deposits usually contains no less than 70% gold, but the coins themselves often contain just 45-55% gold. For some reason, Lydians coin issuers chose to introduce a bit of pure silver into the electrum mix before coining it.

In the chart below, for instance, the vertical column that represents the 1/6 stater denomination contains around 14 different coins. The majority of these coins contain less than 65% gold.  Only two contain more than 80% gold.

Most electrum coins contained less than 70% gold. Source: Velde

Why would the Lydians have chosen to dilute electrum with silver? The intrinsic value of natural electrum was quite high. Given that gold was worth around ten times the value of silver, numismatists estimate that the most commonly available coin, the trite, was worth several sheep, or ten day’s wages (de Callatay, 2013). Converting into modern terms, the trite would be worth the equivalent of a $500 bill. This hardly seems a very convenient denomination. The smallest coin, the 1/96th stater, was worth about a day’s wages, and thus not useful as small change (Velde, 2012).

Wallace (1987) suggest that by mixing some silver into the natural electrum, the intrinsic value of the coin would have been reduced. The price at which the issuer promised to redeem the coin could now be lowered. This reduction would have permitted electrum coins to participate in a wider range of transactions, thus increasing their usefulness.

Still more questions

New data and theories about electrum coins have improved our knowledge. Unfortunately, it seems that we remain “confused but on a higher level!" remarks historian Francois de Callatay (2013). While Wallace’s theory is elegant, it leads to only to more questions.

Velde asks some of the more glaring ones. If electrum coins were redeemable, what did the issuer promise to redeem their coins with? Gold? Silver? Perhaps they be used to discharge taxes? If gold and silver were to be used to redeem electrum coins, why not use these materials as the basis of coinage instead?

And what did the issuer keep in reserve to “back" his guarantee, wonders Velde. If each coin had to be 100% backed by gold, then our issuer would have had to incur the costs of storing and vaulting the yellow metal. This would have meant that issuing coins wasn’t very profitable. One wonders why our electrum owner would have bothered producing them in the first place.

Electrum coins, what happened to them?

Whereas the Brits still seem to be quite fond of their 1p and 2p coins, the Lydians quickly discontinued their electrum coinage. About a hundred years after electrum coins were first issued, they disappear from the numismatic record.

Around 550 BC, King Croesus decided to issue individual silver and gold coins. This switch from electrum to pure gold and silver coincides with the discovery of the process of cementation, the ability to separate gold from silver. Presumably decomposing electrum into its constituent parts in order to create a uniform currency was deemed superior to issuing electrum discs.

Except for a few smaller city-states that continued to issue electrum coins for another century or two, never again would a mixed silver-gold coin be issued. All that remains is a mystery for modern numismatists to puzzle over.

---

Sources:

de Callatay, Francois. White Gold: An Enigmatic Start to Greek Coinage. 2013. [link]
Velde, Francois. On the Origin of Specie. 2012. [link]
Velde, Francois. A Quantitative Approach to the Beginnings of Coinage. [link]
Wallace, Robert. The Origin of Electrum Coinage. 1987. [link]
Wallace, Robert. On the Production and Exchange of Early Anatolian Electrum Coinage. 1989. [link]

Hacksilver

1. Over the last month or two I've been following an interesting archaeological debate over the discovery of coinage. I thought I'd share it with you.

2. It's generally accepted by archaeologists and numismatists that the first coins were invented in Lydia, modern day western Turkey, in the 7th Century B.C.E. (i.e. 610 B.C.E. or so). The idea quickly spread to Greece. The Lydians used electrum, a strange silver/gold mix, to make their discs. (I wrote about electrum coins here). I've included an example below.

Electrum coin from Ephesus, 625–600 BC [Source]

We don't know exactly why the Lydians used electrum, or even if they treated their discs in the same way that future generations would use coins. But when the Greek city states copied Lydian coinage in the 6th Century, they didn't use electrum. Their coins were pure silver.

3. Lydia's electrum coins aren't the topic of this post. The debate that I'm going to describe revolves around the belief among some archaeologists that a form of proto-coinage had been invented prior to the Lydians and their electrum coins. This proto-coinage came in the form of sealed and regulated bags of hacksilver (more on hacksilver later).

Others archaeologists disagree. They are adamant that Lydia remains ground zero for coinage.

For lack of better terminology, I'll call the first group of archaeologists, those who think there was a predecessor sort of coin, the proto-coiners.

So relax and follow along.

4. By the way, mine is an outsider's account on the archaeology of money and coinage. I am not an archaeologist, so I will certainly get a few things wrong. Nevertheless, I am hoping that my regular monetary economics readership will enjoy learning how archaeologists attack the problem of money.

5. How popular was silver in ancient society?

"Silver served as the main measure of value, the means of payment and credit, and as an indirect form of exchange in Near Eastern economies from the mid-3rd millennium onward," write archaeologists Tzilla Eshell, Ayelet Gilboa, Naama Yahalom-Mack, and Ofir Tirosh (Eshel et al) in a 2018 article entitled Four Iron Age Silver Hoards from Southern Phoenicia. The Near East is a catch-all term for modern-day Israel, Iraq, Iran, Jordan, and Syria. I will return later to Eshel and coauthors' paper.

Morris Silver, an economist who researches ancient economies, describes Mesopotamian texts of the middle of the second half of the third millennium that show silver being used by street vendors, to pay rent, purchase dates, oil, barley, animals, slaves, and real estate.

According to archaeologists Seymour Gitin & Amir Golani (2004), Assyrian economic texts from the 7th C B.C. show that the majority of all types of payments were already being made in silver, including those for tribute, craftsmen obligations, and for conscription and labor commutations.

Cuneiform tablet, loan of silver [Source: The Met]

The Old Assyrian cuneiform tablet above from around 1900 B.C.E. says that 6 minas (c. 3 kg) of silver are owed by two men to the merchant Ashur-idi. One third of the loan must be paid by the next harvest and the rest at a later date. If it is not repaid by that time it will accrue interest charged at a monthly rate.

6. If silver had already become a sort of medium of exchange sometime between 3000 B.C.E. and 2000 B.C.E., it wasn't in coin form, but as hacksilver. By hacksilver, what is usually meant by archaeologists is silver ingots, hacked pieces of ingot, silver scrap, and cut up bits of silver jewellery.

Below are some examples of hacksilver:

Hacksilver (i.e. cut silver ingots) from four Phoenician hoards, 11th-9th century BCE.

Hacking may have been used to check the quality of silver. Pure silver was soft. It would have been easier to chisel than silver alloyed with copper.

See: https://t.co/IIlwXKSTaz pic.twitter.com/vWsV0PCUOg

— John Paul Koning (@jp_koning) January 14, 2021

7. One reason for the hacking or cutting-up of silver may have been to make small change. If 2 grams of silver was required to make a payment, but a payee only had a single 10 gram ingot, then a small part of it had to be cut off.

8. Another less obvious reason for hacking, suggested by Eshel & coauthors, is that it may have been a way for merchants to check for quality. Pure silver is soft. Mixing copper into silver makes for a harder ingot. A solid smash to the ingot may have been the accepted way of verifying whether an ingot was good silver or not.

9. It wasn't till 610 BC or so that the Lydians made the first coins. So for over a thousand years, silver circulated as a medium of exchange, in hacked form.

10. The big innovation with coins is the stamp. Because we trust the issuer's brand, we needn't weigh out or assay (i.e. smash/hack) silver prior to engaging in trade. So trade was much more fluid.

If you think about it, branding metal is a pretty big step for a society to take. It means that laws, norms, and institutions have become established enough for people to be confident in something as abstract as an issuer's emblem. Too much fraud, warfare, and lawlessness, and branding breaks down—you've got to go back to weighing and hacking silver yourself.  

11. The proto-coiners don't agree that Lydia was the first to "brand" silver. They suggest that bagged and sealed hacksilver was already circulating in a way similar to coins. Some authority, perhaps a government administrator or a merchant, pre-weighed a certain amount of good hacksilver, bagged it, and affixed their seal to it. And so anyone who was offered the bag in trade could treat it just as they would a coin. As a verified amount of silver, it needn't be weighed or hacked. The bag would have been accepted according to whatever information was inscribed on the seal.

12. If the proto-coiners are right, that means our ancestors were better monetary innovators than we originally thought. It pushes the effective date of coinage technology back by 500 or so years.  

13. It's a fascinating debate, especially because it invokes a set of mysterious old hoards that archaeologists have discovered over the years. These hoards are typically hidden in clay jars underneath the floors of houses by their owners, probably for safekeeping. And then they were forgotten or some disaster befell their owner, only to be rediscovered thousands of years later. Who were these people? Why did their hoard get forgotten?

14. One of the key hoards around which the debate revolves is the Tel Dor hoard, which was found north of Haifa in Israel. It was excavated in the 1990s by Ephraim Stern, an Israeli archaeologist at the Hebrew University of Jerusalem. One element of the Dor hoard is an old jug filled with silver, below.

The Tel Dor hoard (a) as displayed in The Israel Museum and (b) in situ, looking east [Source: Eshel et al]

15. Stern's description of this jug (published in this 2001 paper) quickly filtered into the archaeological community. Christine Thompson, archaeologist and co-founder of the Hacksilber Project, used Stern's findings to build a proto-coinage argument. It's worth getting into the details of her argument (and subsequent rebuttals) to see how archaeologists think. You can find it in her 2003 paper, Sealed Silver in Iron Age Cisjordan and the ‘Invention’ of Coinage.

Thompson (channeling Stern) tells us that the Tel Dor hoard dates to somewhere between 1000 B.C.E and 900 B.C.E. The hoard consists of a jug containing 17 bundles of hacksilver wrapped up by linen cloth (see photo of one of the bundles below).

16. Together the silver weighs 8.5 kilograms, which at today's silver price is worth around $8,000. But that's not a great way to think about how much this hoard was worth. According to a very readable article by Tzilla Eshel, a half-gram of silver was equivalent to a day-worker's wage. So the entire hoard was the equivalent of forty-six years of labour. In modern day terms, that would value the purchasing power of the hoard at well above $1 million.

17. Stern has speculated that the hoard may have belonged to a Phoenician merchant who used the silver to build and equip ships, or to buy merchandise for eventual exchange with countries in the western Mediterranean.

18. Most of the linen wrapping in the Tel Dor hoard has long since disintegrated. Thompson notes that the bundles were closed with bullae, or clay seals. (See below). But these seals do not contain a name, just a pattern.

One of the silver bundles found at Tel Dor, and an illustration of a clay bulla, or seal. Source: Ephraim Stern in The Silver Hoard from Tel Dor [pdf]

19. If the bundles found in the Dor hoard were treated by their owner as coins, then one would expect them to weigh a standard amount, just like all modern nickels and quarters weigh the same. And that is the gist of Thompson's argument. According to her, the 17 bundles all appear to be the same weight. .
 
20. One of these bundles had been removed by Stern to be weighed. It registered at 490.5 grams. Thompson suggests that this 490.5 grams might align with usage of the Babylonian shekel unit of account. Under this archaic weight standard, each shekel weighed 8.3g, and 60 shekels was worth 1 mina. Thus a mina would have weighed 500 grams.

So the 490.5 gram bundle found at Dor comes close to an even mina. It's as if the bundle was a very large denomination mina coin. Thompson attributes the missing 10 or so grams to loss of a few small pieces due to disintegration of the cloth.

21. The purity of the Dor hoard is quite high, notes Thompson, suggesting that the bagged silver, like a coin, had been checked and regulated.

22. Thus Thompson has created a plausible theory about bags of silver being treated as coins. Like a coin, the sealed bags of hacksilver found at Dor had a certain purity and weight. Presumably people who received them in payment didn't have to weigh the silver. Nor did they have to assay the silver by hacking or smashing it.

23. It's a convincing theory. Now for the counter-theory.

24. Raz Kletter, an archaeologist at University of Helsinki, is not convinced by the idea of a proto-coinage. In a 2004 paper, he points to the nearby Tell Keisan hoard, dated to around 1000 B.C., which also contained wrapped hacksilver bundles. The hoard includes 6 or 7 bags of cloth, says Kletter. Two of them weighed in at 24.5 and 25 grams, suggesting that they may have conformed to the same denomination. But two of the other bags measured at a 32 and 100 grams respectfully, which muddies the waters. Moreover, Kletter says that the weights of the bags does not correspond clearly to any known standard of weights and measures.

25. Tel Dor hadn't finished telling its story, either. Recall that at the time Thompson was writing, 2003, only one of Dor's 17 bundles had been weighed. It came in at 490.5 grams, and Thompson ascribed to this bundle the possible value of a clean mina of 500 grams, less a few grams due to thousands of years of wear and tear.

But in 2018, Eshel & co-authors opened a second Dor bundle. They found that it measured just 420.6 grams, which doesn't conform as closely to a mina. So that undermined some of the arguments in favor of proto-coinage.

26. That's not all. Eshel & co-authors did a chemical analysis of four hoards including both Tel Dor and Tell Keisan. Recall that Thompson suggested that the purity of Tel Dor silver indicated a degree of regulation, much like a mint controls the silver content of a coin. But Eshel & co-authors found that the silver from Tell Keisan, though "piously" packed in sealed bundles, was not so pure. It contained large amounts of copper, suggesting that it was a forgery. (See photo below). If the whole point of bagging and sealing was to create a trustworthy medium of payment, the deliberately-alloyed Tell Keisan silver seems to contradict this.

A bundle of hacksilver from Tel Keisan. Its green colour betrays its copper content. When silver corrodes it tarnishes black, but copper produces a green rust. Source: The Torah

27. This collection of counter-observations somewhat weakens the argument for an early proto-coinage in the Near East. But there are probably plenty of yet-to-be discovered hoards. Who knows, perhaps the next one will contain bundles of provably standardized hacksilver. It certainly is a provocative idea.

28. If the role of bundling and sealing of hacksilver wasn't to create a proto-form of coinage, than what was its function? Eshel & co-authors suggest that bagging was little more than a convenient manner of storing one’s wealth. Taking out a single cloth bundle and weighing it would have been much less awkward than removing individual pieces one-by-one and weighing them. 

If you're interested in learning more about the ancient hacksilver economy, I'd suggest reading How Silver Was Used for Payment, recently published in The Torah. Tzilla Eshel, the archaeologist who co-authored one of the papers I cite in my blog post, is the author and has written it with the lay-person in mind.

From ancient electrum to modern currency baskets (with a quick detour through symmetallism)

Electrum coins [source]

First proposed by economist Alfred Marshall in the late 19th century as an alternative metallic standard to the gold, silver and bimetallic standards, symmetallism was widely debated at the time but never adopted. Marshall's idea amounted to fusing together fixed quantities of silver and gold in the same coin rather than striking separate gold and/or silver coins. Symmetallism is actually one of the world's oldest monetary standards. In the seventh century B.C., the kingdom of Lydia struck the first coins out of electrum, a naturally occurring mix of gold and silver. Electrum coins are captured in the above photo.

While symmetallism is an archaic concept, it has at least some relevance to today's world. Modern currencies that are pegged to the dollar (like the Hong Kong dollar) act very much like currencies on a gold standard, the dollar filling in for the role of gold. A shift from a dollar peg to one involving a basket of other currencies amounts to the adoption of a modern version of Marshall's symmetallic standard, the euro/yen/etc playing the role of electrum.

The most recent of these shifts has occurred with China, which late last year said it would be measuring the renminbi against a trade-weighted basket of 13 currencies rather than just the U.S. dollar. Thus many of the same issues that were at stake back at the turn of the 19th century when Marshall dreamt up the idea of symmetallism are relevant today.

So what exactly is symmetallism? In the late 1800s, the dominant monetary debate concerned the relative merits of the gold standard and its alternatives, the best known of which was a bimetallic standard. The western world, which was mostly on a gold standard back then, had experienced a steady deflation in prices since 1875. This "cross of gold" was damaging to debtors; they had to settle with a higher real quantity of currency. The reintroduction of silver as legal tender would mean that debts could be paid off with a lower real amount of resources. No wonder the debtor class was a strong proponent of bimetallism.

There was more to the debate than mere class interests. As long as prices and wages were rigid, insufficient supplies of gold in the face of strong gold demand might aggravate business cycle downturns. For this reason, leading economists of the day like Alfred Marshall, Leon Walras, and Irving Fisher mostly agreed that a bimetallic standard was superior to either a silver standard or a gold standard. (And a hundred or so years later, Milton Friedman would come to the same conclusion.)

The advantage of a bimetallic standard is that the price level is held hostage to not just one precious metal but two; silver and gold. This means that bimetallism is likely to be less fickle than a monometallic standard. As Irving Fisher said: "Bimetallism spreads the effect of any single fluctuation over the combined gold and silver markets."  Thus if the late 1800s standard had been moved from a gold basis to a bimetallic one, the stock of monetary material would have grown to include silver, thus 'venting' deflationary pressures.

Despite these benefits, everyone admitted that classical bimetallism had a major weakness; eventually it ran into Gresham's law. Under bimetallism, the mint advertised how many coins that it would fabricate out of pound of silver or gold, in effect setting a rate between the two metals. If the mint's rate differed too much from the market rate, no one would bring the undervalued metal (say silver) to the mint, preferring to hoard it or export it overseas where it was properly valued. The result would be small denomination silver coin shortages, which complicated trade. What had started out as a bimetallic standard thus degenerated into an unofficial gold standard (or a silver one) so that once again the nation's price level was held hostage to just one metal.

The genius of Alfred Marshall's symmetallic standard was that it salvaged the benefits of a bimetallic standard from Gresham's law. Instead of defining the pound as either a fixed quantity of gold or silver, the pound was to be defined as a fixed quantity of gold twinned with a fixed quantity of silver, or as electrum. Thus a £1 note or token coin would be exchangeable at the Bank of England not for, say, 113 grains of gold, but for 56 grains of gold together with twenty or so times as many grains of silver. The number of silver and gold grains in each pound would be fixed indefinitely when the standard was introduced.

Because symmetallism fuses gold and silver into super-commodity, the monetary authority no longer sets the price ratio between the two metals. Gresham's law, which afflicts any bimetallic system when one of the two metals is artificially undervalued, was no longer free to operate. At the same time, the quantity of metal recruited into monetary purposes was much larger and more diverse than under a monometallic standard, thus reducing the effect of fluctuations in the precious metals market on aggregate demand.

While symmetallism was an elegant solution, Alfred Marshall was lukewarm to his own idea, noting that "it is with great diffidence that I suggest an alternative bimetallic scheme." To achieve a stable price level, Marshall preferred a complete separation of the unit of account, the pound, from the media of exchange, notes and coins. This was called a tabular standard, a system earlier proposed by William Stanley Jevons. The idea went nowhere, however; the only nation I know that has implemented such a standard is Chile. As for Fisher, he proposed his own compensated dollar standard plan, which I described here.

The urgency to adopt a new standard diminished as gold discoveries in South Africa and the Yukon spurred production higher, thus reducing deflationary pressures. None of these exotic plans—Marshall's symmetallism, Jevons tabular standard, or Fisher's compensated dollar—would ever be adopted. Rather, the world kept on limping forward under various forms of the gold standard. This standard would be progressively modified through the years in order to conserve on the necessity for gold, first by removing gold coin from circulation and substituting convertibility into gold bars (a gold bullion standard) and then having one (or two) nations take on the task of maintaining gold convertibility while the remaining nations pegged to that nation's currency (a gold exchange standard).

---

Let's bring this back to the present. In the same way that conditions in the gold market caused deflation among gold standard countries in the late 1800s, the huge rise in the U.S. dollar over the last few years has tightened monetary conditions in all those nations that peg their currency to the dollar. To cope, many of these countries have devalued their currencies, a development that Lars Christensen has called an 'unraveling of the dollar bloc.'

A more lasting alternative to re-rating a U.S. dollar peg might be to create a fiat version of electrum; mix the U.S. dollar with other currencies like the euro and yen to create a currency basket and peg to this basket. China, which has been the most important member of the dollar bloc, has turned to the modern version of symmetallism by placing less emphasis on pegging to the U.S. dollar and more emphasis on measuring the yuan against a trade-weighted basket of currencies. This means that where before China had a strictly made-in-the U.S. monetary policy, its price level is now determined by more diverse forces. Better to put your eggs in two or three baskets than just one.

Bahrain, Oman, Qatar, Saudi Arabia and United Arab Emirates are also members of the dollar bloc. Kuwait, however, links its dinar to a basket of currencies, a policy it adopted in 2007 to cope with the inflationary fallout from the weakening U.S. dollar. In an FT article from April entitled Kuwaiti currency basket yield benefits, the point is made that Kuwait has enjoyed a more flexible monetary policy than its neighbours over the recent period of U.S. dollar strength. Look for the other GCC countries to mull over Kuwaiti-style electrum if the U.S. dollar, currently in holding pattern, starts to rise again.

Modern day electrum can get downright exotic. Jeffrey Frankel, for instance, has suggested including commodities among the basket of fiat currencies, specifically oil in the case of the GCC nations. Such a basket would allow oil producing countries to better weather commodity shocks than if they remained on their dollar pegs. If you want to pursue these ideas further, wander over to Lars Christensen's blog where Frankel's peg the export price plan is a regular subject of conversation.