Stephen Laniel’s Unspecified Bunker

Attention conservation notice: 1300 or so words on the evolution of international currency arrangements from the 1700’s to now. Contains more than my standard share of confusion, but it felt important to get the ideas down and let other people point out my errors.

This book has two central premises:

1. It’s easier for national governments to keep their currency’s exchange rate in check when they don’t have to worry about voters.

2. International monetary arrangements have a way of coming unraveled if any of the parties has an incentive to back out of the arrangement; this is an instance of what’s called a “coordination problem” (the prisoner’s dilemma is the classical coordination problem). Consequently, the only way to keep such arrangements working is to tie nations together, in such a way that any nation’s backing out of the arrangement would harm it.

The first point gets us from the start of the gold standard until its end in the early part of the 20th century. (Depending upon how it’s phrased, it seems like you could make a case that the gold standard ended in the 30’s when Britain left it, or in the 70’s when Nixon let the U.S. dollar float. I don’t follow the details well enough to explain why you’d pick the one date over the other. More on this ignorance below.) And actually, the start of the gold standard in Britain was a pure historical accident, which takes up one of the more interesting paragraphs I’ve ever read.

It’s an application of Gresham’s Law, and involves Sir Isaac Newton. Newton was the warden of the Royal Mint when Britain was on a bimetallic standard; it was using both silver and gold for its currency. A bimetallic standard is tricky to manage, because you have to set the relative price of the coins properly. For instance, suppose that on the open market, an ounce of gold is worth as much as 16 ounces of silver; gold and silver coins should then stand in approximately the same ratio. Suppose that the market price of silver then rises to 1/10th the price of gold. If the currency still stands at the old 16-to-1 ratio, then silver coins are worth more melted down into bars and traded for gold than they are as coins. Silver coins will systematically disappear and be melted down. If this ill-chosen exchange rate persists for too long, silver coins will disappear altogether from the market.

I should note something up front here: I need to think more about the economic details of this argument. Why, for instance, wouldn’t the melting-down of silver coins eventually self-correct? By melting down coins and taking them out of circulation, we reduce the supply of currency, so the value of each coin goes up; likewise, the rising supply of silver bars makes the value of each bar go down. Eventually, wouldn’t the two balance out again?

The answer is probably that it depends. If the difference between the commodity value and the coin value of silver is too large, all the coins could well disappear before the market has had time to equalize. One might then ask why the British government didn’t continue producing more coins until the market worked it out — or quickly realize their error and return to more sensible relative prices for gold and silver. As it happens, in any case, they didn’t: they ended up accidentally with a monometallic standard. And that’s how we got the gold standard: the lucky (?) confluence of a Newtonian mistake and British dominance in the 18th century.

Another historical accident kept the gold standard working for a while: most people didn’t have the right to vote, and those who did were wealthier than those who didn’t. In a world run by (to modern eyes) undemocratic governments, it’s easy to defend your currency’s gold value: if an ounce of gold buys you $35, say, and the currency devalues so that it now buys you $40, the U.S. government can just spend as many dollars as necessary to bring the currency back into line. In a more democratic world, you need to use those dollars to support your people. In a less-democratic world, the dollars can all go toward gold.

There’s an important supporting fact about 19th-century less-democratic governments that makes this all work out: currency traders know exactly how the government is going to behave. Hence they have every reason to believe that the American government will keep the dollar right where it’s always been. If the currency drops from $35/oz. to $40/oz., traders know that the government will make every effort to bring it back to the old level. They have no reason to bet on future currency drops; the only rational bet they could make is that the dollar would return to its old value. Consequently, the dollar does return to its old value.

What if, instead, traders know that the government has other obligations toward its people? If it would cost a prohibitive amount to defend the gold standard, traders rationally believe that the government will let the depreciation stand, and they will bid the value of the currency down. Now there’s an even wider gap for the government to make up, so traders know it’s even less likely that the government will fix it. And so on down the drain. This is a “speculative attack.” In fact Eichengreen gives an example of a self-fulfilling speculative attack, where countries with a perfectly healthy economy (low inflation, healthy balance sheet, etc.) nonetheless find themselves with a quickly depreciating currency. There’s a paper on the topic, “Self-Fulfilling Expectations, Speculative Attack, and Capital Controls,” that seems like it might be fascinating.

Governments can try other options to keep their currency under control. They can limit the flow of money: charge a hefty fee for every large bundle of dollars or gold that leaves the country, for instance. Other countries have to get involved to make this work: as capital gets more mobile, any country that locked down currency transfers would find itself with a massively depreciating currency as investors sold it and bought other, freer currencies. You can see why multinational agreements to lock down currency transfers would have a hard time sticking around: if I defect — you lock down and I don’t — money flows into my country and out of yours. We need some way to tie our fates together if we’re going to make this work. My understanding is that European monetary union — the euro — is precisely such a fate-tying mechanism: we give up flexibility in some areas of monetary policy in order to solve a larger coordination problem.

Again a question of detail comes up: suppose the currency flows from a less-open to a more-open economy. For the sake of concreteness, let’s say those economies are Germany and France, respectively. Money flows from Germany to France, so now one franc buys more deutsch marks. France doesn’t want this exchange rate to rise too far, because now Germans can’t buy as many French products; export-intensive French industries are harmed by Germany’s capital constraints. So it would seem that France and Germany are already quite strongly linked by self-interest, even in the absence of any agreements on capital controls.

I’m sure the answer to this, as well as to all my other questions, is within Globalizing Capital. My sense is that it is a book which can never be read, only reread. I’m really looking forward to rereading it.

(Thanks to Cosma Shalizi for recommending this book in an email. He wrote a review of Globalizing Capital, which I forbade myself from reading before I wrote this one.)