The traditional arguments in favor of gold as the most rational exchange medium are invariably a combination of some or all of the following, all of which are undone by modern technology or empirical data:
1) It's easily divisible.
2) It's identifiable (and elemental - it can't be created ab initio).
3) It's durable.
4) It's rare.
5) It's useful.
Gold certainly qualifies for these categories, but today, so do most of the other transition metals. Let's dispense with 1, 2, and 3 immediately since they're no longer unique features of gold. Modern technology lets us work with and machine and divide into coins all metals - and identify them as well (we now have better techniques than color, oxidation and specific gravity). No, this technology wasn't available to past governments, but it's available to us, and we're talking about why would gold still be the standard now, not historically - or at least why it would be the fall-back.
In terms of durability, although gold is more resistant to oxidation than other metals, once mined and refined, metals are stored under contolled conditions (like vaults or pockets) that make the durability question moot. Pure gold nicks and bends easily too; many other metals do not.
So I ask again: why gold? In terms of usefulness, gold does have industrial and medical applications, but in that regard it's sure no uranium or silicon, and in any event it would be hard to quantify the innate industrial value of an element. That leaves us with only one physical attribute to be the source of gold's central role in human economics.
Rarity
I had always assumed that value correlated with rarity, but realized I'd never seen data on this. Therefore, to determine which metal would be the best value reserve, I looked at the value to rarity ratio by comparing the relative abundance of metallic elements in the Earth's crust and their modern values; most values came from the Los Alamos periodic table pages, or from current commodities markets for those metals that are commonly traded. For practical purposes, we can rule out as a potential value reserve for humans any elemental metals which are unstable or unhealthy to handle at room temperature. For example: radium is radioactive, tellurium stinks, cesium explodes, and francium exists only momentarily and as part of a nuclear decay path (which means it's both radioactive and becomes cesium, which explodes).
Ruling out the intolerables, we find that osmium and iridium are both rarer and more expensive than gold. Before you start celebrating the inherent rationality of the human valuation of metals, please find me someone advocating an osmium standard for a currency, or a single currency or economy that has ever been based on either osmium or iridium. By the reasons given for gold as the basic commodity metal, currencies should be based on osmium, at least now that we can collect and refine it. Then again, maybe two centuries from now the foolish age of fiat currencies will have passed and the world economy will be on an iridium standard and I will be hailed as a visionary. (Use the right side of my profile on coins, please.) Of course, I'm not making coin-portrait appointments yet, because on further investigation, any semblance of economic logic breaks down:
Platinum has the same value as gold (or historically, often higher) even though it's substantially less rare in the Earth's crust. It bears mentioning that for some reason platinum has never been the basis for any economies either. Palladium is about twice as common than gold. Rhenium is slightly rarer than gold but about a third the value. Rhodium is about 4x rarer than gold but only about 15% more valuable. Ruthenium is about three times rarer than gold but has similar value, even though it's more industrially useful, even if still obscure. Even so, you don't see it being traded in Chicago, and there have never been currencies based on it, or to my knowledge even the suggestion thereof.
One objection to my theory is that a commodity that's too rare is useless as a basis for wealth, and that there is a rarity "sweet spot" for precious metals. While self-evidently true, this really doesn't tell us anything since it's a circular argument. That is to say, by this argument gold has "just the right level of rareness", which we know because it's what people use, so it must be just right. No dice. I might concede that iridium, which is almost 9 times rarer than even gold, wouldn't be a great choice for a standard, but why not palladium or rhenium? They're both of similar abundance to gold, and they're both industrially useful too.
You might object that just because an element is in the Earth's crust doesn't mean it's jumping out of the ground and hammering itself into coins; and that production is a measure of said element's availability to human activities. So I ordered elements by global production:
There's a slight trend here, but it's still not the clear ordering you might expect to find. One thing I noticed was that (not surprisingly) gold was produced far out of proportion to its crustal abundance, which suggests that economic value isn't just a function of ease of extraction. So I also ordered the elements in terms of the ratio of production to abundance. Without showing you the scatterplot, the first ten most "overproduced" relative to their natural abundance were, in decreasing order, carbon, nitrogen, gold, lead, antimony, bismuth, tellurium, phosphorus, sulfur, and mercury. Gold is in the company of some rather economically indistinguished elements.
Oddly enough, copper is 20,000x more common in the Earth's crust than gold, and yet it has been used as currency, as has silver. What's going on? There are three things going on: oxidation potentials, animal behavior, and habituation.
A Behavioral Explanation
I added this paragraph after the fact, because of the timely eruption of a tailor-made scandal driving home my point: if you don't think irrational grouppthink psychology plays a major role in human economic behavior, how did the Japan Enten scandal happen? If you're still not convinced of the role of irrational human psychology, take the time to read MacKay's Extraordinary Popular Delusions and the Madness of Crowds.
First, gold, silver and copper are the only metals likely to be found at the Earth's surface that have negative oxidation potentials for the transition from pure to oxidized metal (meaning they can be found in nature in their unoxidized forms). Okay; so one day a few thousand years ago, someone was wading up a creek, and found gold nuggets glinting under the water. So what? Why was he interested in the physical properties of shiny gold nuggets any more than he would be in wood or rock or water?
It bears constant repeating that when we talk about economics, we're talking about the social behavior of a specific animal. Economists concede today that Homo sapiens is not the perfectly rational self-interested and profit-maximizing animal it was sometimes previously thought to be. Having said that, in asking "why gold?", it seems to me the principle factor that's overlooked is entirely irrational and human-dependent, and it best explains why modern humans might be predisposed to maintain gold as a store of value. That factor is that we're primates, and we like shiny things.
I mean that to be taken literally and seriously. Though gold does have industrial applications, we still use it primarily for decoration. On the other hand, I've never seen a (dull silverish) palladium necklace being sold by the primates at Tiffany's. Males like to have sex with females. Like other primates, females like shiny things, and especially like them if their culture equates certain types of shininess with sexual desirability, and especially of their culture teaches them to value what other female primates think of the shiny things. Therefore, sometimes males go to Tiffany's to get them shiny things. Therefore, gold is valuable.
Now add several thousand years on top of our initial biologically-determined primate-fascination with gold (which keeps it from being like, say, tulips) and you have considerable historical inertia. By this I mean many (but not all) civilizations independently adopted gold as their exchange standard and store of value, and now that billions of us are all trading, it would be even more impossible to change. To do so, you'd somehow have to simultaneously convince all humans who now recognize gold as inherently more valuable than most other metals (the majority of humans, including wives) that another medium was more valuable; this would be true even if there were no other reasons to use gold as a standard. Even the effective removal of monetary gold from circulation in the U.S. in the 1930s had little impact. It would be very difficult to convince people to switch to a different metal, because everyone already is programmed by history to value gold, and would you trust your neighbor to stop considering it valuable at the same moment you do? It's for this same reason that Ithaca hours, among other quirky currencies, have largely died out. When people in your town take Ithaca hours and American dollars, and no one in the town next door takes Ithaca hours, you get pretty nervous about, say, opening your 401K in Ithaca dollars.
One frequent objection is that if the use of gold were really just an economic legacy system or in game theoretic terms a massive and long-running coordination game, then there should have been economies that assigned higher values to shiny metals other than gold. And of course there were. For centuries China valued silver more highly than gold (whether you can say they were on a silver standard is another question, though I would argue that in a sloppy and imprecise way, yes, of course they were). Isaac Newton even wrote on the topic for the British government, and illustrated his concern over the trade imbalances that would result for the two ends of the Old World if economies on opposite sides of Eurasia used different standards. Note that he didn't state that one was better than the other, just that having two value systems operating simultaneously was unhealthy. Today the gold/silver price ratio in China is the same as in the rest of the world. China's system was Betamax to Europe and the Middle East's VHS; I make this analogy to stress that, like Newton, I don't think that gold is necessarily better either, just that more people were using gold, and those people happened to colonize the rest of the world, so it was a matter of time before China's commodity valuations would have to adjust if they were to keep trading with everyone else.
CONCLUSION
I'm no commodities expert which is why I took advantage of a friend's recent visit to pick his brain. He does trade commodities, and works as a geologist for a major U.S. government contractor. When I asked him "why gold", he made the usual jokes about going on a selenium standard, and then said, "I don't know. Historical reasons I guess."
I'm not interested in convincing anybody to stop using gold. If you're going to use a precious metal as a store of value, everybody should use the same one, and gold is the one that humans have settled on at this point in history; it's still a kind of stable value reserve, even if it doesn't anchor currencies anymore. Some economists have predicted that it will eventually lose its value, as matter becomes smarter and labor continues to get more valuable, but we'll see. I asked the question "why gold" purely out of curiosity, but it's interesting to note that no anti-fiat-currency websites have begun calling for a rhodium standard. Even more telling is that no one can provide an argument that explains the tenacity of gold as well better than observing that we primates are creatures of habit, and we like shiny things.
