While gold is widely held as a store of purchasing power, and while it is possible to use gold as a unit of account (by converting its floating value to denominate anything in gold terms), gold is no longer widely used as a medium of exchange.

Noah Smith says that gold will never be a widespread medium of exchange again:

In the days when people carried around gold doubloons and whatnot as money, you had a global political system characterized by pockets of stability (the Spanish Empire, or the Chinese Empire, or whatever) scattered among large areas of anarchy. Those stable centers minted and gave out the gold coins. But in the event of a massive modern global catastrophe that brought widespread anarchy, the gold bars buried in your backyard would not be swappable for eggs or butter at the corner store. You’d need some big organization to turn the gold bars into coins of standard weights and purity. And that big organization is not going to do that for you as a free service. More likely, that big organization will simply kill you and take your gold bars, Dungeons and Dragons style.

In other words, I think gold is never coming back as a medium of exchange, under any circumstances. It is no more likely than a return of the Holy Roman Empire. Say goodbye forever to gold money.

Well, forever is a very long time. Human history stretches back just six million years. Recorded history suggests that gold has only been used as a medium of exchange for five or six thousand years. But for that tiny sliver of human history, gold became for many cultures entirely synonymous with money, and largely synonymous with wealth. So I think Noah is over egging his case by using the word forever. Societies have drastically changed in the last six thousand years, let alone the last one hundred. We don’t know how human culture and technology and societies will progress in the future. As humans colonise space, we may see a great deal of cultural and social fragmentation; deeper into the future, believers in gold as money may set up their own planetary colonies or space stations.

But what about the near future? Well, central banks are still using gold as a reserve. In the medium term, it is a hedge against the counter-party risks of a global fiat reserve system in flux. But central banks buying and acquiring gold is not the same thing as gold being used as a medium of exchange. Gold as a reserve never went away, and even in the most Keynesian of futures may not fully die for a long time yet.

And what about this great hypothetical scenario that many are obsessed with where the fragile interconnective structure of modern society — including electronics — briefly or not-so-briefly collapses? Such an event could result from a natural disaster like a megatsunami, or extreme climate change, or a solar flare, or from a global war. Well, again, we can’t really say what will or won’t be useful as a medium of exchange under such circumstances. My intuition is that we would experience massive decentralisation, and trade would be conducted predominantly either in terms of barter and theft. If you have gold coins or bars, and want to engage in trade using them — and have a means to protect yourself from theft, like guns and ammunition — then it is foreseeable that these could be bartered. But so too could whiskey, cigarettes, beer, canned food, fuel, water, IOUs and indeed state fiat currencies. If any dominant media of exchange emerges, it is likely to be localised and ad hoc. In the longer run, if modern civilisation does not return swiftly but instead has to be rebuilt from the ground up over generations then it is foreseeable that physical gold (and other precious metals, including silver) could emerge as the de facto medium of exchange, simply because such things are nonperishable, fungible, and relatively difficult to fake. On the other hand, if modern civilisation is swiftly rebuilt, then it is much more foreseeable that precious metal-based media of exchange will not have the time to get off the ground on anything more than the most localised and ad hoc of bases.

Noah concludes:

So when does gold actually pay off? Well, remember that stories do not have to be true for people to believe them. Lots and lots of people believe that gold or gold-backed money in the event of a global social disruption. And so when this story becomes more popular (possibly with the launching of websites like Zero Hedge?), or when large-scale social disruption seems more likely while holding the popularity of the story constant, gold pays off. Gold is like a credit default swap backed by an insolvent counterparty – it has no hope of actually being redeemed, but you can keep it around forever, and it goes up in price whenever people get scared.

In other words, gold pays off when there is an outbreak of goldbug-ism. Gold is a bet that there will be more goldbugs in the future than there are now. And since the “gold will be money again” story is very deep and powerful, based as it is on thousands of years of (no longer applicable) historical experience, it is highly likely that goldbug-ism will break out again someday. So if you’re the gambling type, or if you plan to start the next Zero Hedge, or if your income for some reason goes down when goldbug-ism breaks out, well, go ahead and place a one-way bet on gold.

Noah, of course, is right that gold is valuable when other people are willing to pay for it. The reason why gold became money in the first place was because people chose to use it as a medium of exchange. They liked it, and they used it, and that created demand for it. If that happens again, then gold will be an in-demand medium of exchange again. But for many reasons — including that governments want monetary flexibility — most of the world today has rejected gold as a medium of exchange.

But there is another pathway for gold to pay off. Noah is overlooking the small possibility that gold may at some point become more than a speculative investment based on the future possibility that gold may at some point return as a monetary media. In 2010, scientists from the Brookhaven National Laboratory on Long Island, using their Relativistic Heavy Ion Collider (RHIC) collided some gold nuclei, traveling at 99.999% of the speed of light. The plasma that resulted was so energetic that a tiny cube of it with sides measuring about a quarter of the width of a human hair would contain enough energy to power the entire United States for a year. So there exists a possibility that gold could be used at some date in the future as an energy source — completely obliterating any possibility of gold becoming a medium of exchange again. Of course, capturing and storing that energy is another matter entirely, and may prove impossible. In that case — if gold does not become a valuable energy source — it is almost inevitable that some society somewhere at some stage will experiment again with gold as a medium of exchange.