An Austrian Critique of Robert Mundell's "Impossible Trinity"Tags Protectionism and Free TradeMonetary TheoryOther Schools of Thought
Robert Mundell, winner of the Economics Nobel in 1999, recently passed away at the age of eighty-eight. As many of the obituaries have explained, Mundell is considered the intellectual father of the euro and was associated with the supply-side revolution in economics in the early 1980s. However, in this article I will focus on his earlier, seminal work on monetary and trade theory—which won him the prize—and critique it from an Austrian perspective. Specifically, I will explain what is meant by Mundell’s “impossible trinity,” and then show why a hard-money Misesian would reject its apparent insight.
Mundell’s Impossible Trinity
Although I often take shots at him, Paul Krugman actually had a very good explanation in Slate of Mundell’s contributions back in 1999 when he (Mundell) was awarded the Nobel. For our purposes, I’m reproducing here the relevant portion of Krugman’s column. The quotation is somewhat lengthy but it’s very good at first setting the historical context and then explaining (in layman’s terms) what Mundell demonstrated:
Here’s what the world looked like in 1960: Almost all countries had fixed exchange rates with their currencies pegged to the U.S. dollar. International movements of capital were sharply limited, partly by government regulations, partly by the memory of defaults and expropriations in the ’30s. And most economists … took it for granted … that this was the way things would continue to work for the foreseeable future.
But Canada was different. Controlling the movement of capital across that long border with the United States had never been practical; and U.S. investors felt less nervous about putting their money in Canada than anywhere else. Given those uncontrolled movements of capital, Canada could not fix its exchange rate without giving up all control over its own monetary policy. Unwilling to become a monetary ward of the Federal Reserve, from 1949 to 1962 Canada made the almost unique decision to let its currency float against the U.S. dollar. These days, high capital mobility and a fluctuating exchange rate are the norm, but in those days they seemed outrageous—or would have seemed outrageous, if anyone but the Canadians had been involved.
And so perhaps it was the Canadian case that led Mundell to ask, in one of his three most famous contributions, how monetary and fiscal policy would work in an economy in which capital flowed freely in and out in response to any difference between interest rates at home and abroad. His answer was that it depended on what that country did with the exchange rate. If the country insisted on keeping the value of its currency in terms of other nations’ monies constant, monetary policy would become entirely impotent. Only by letting the exchange rate float would monetary policy regain its effectiveness.
Later Mundell would broaden this initial insight by proposing the concept of the “impossible trinity”; free capital movement, a fixed exchange rate, and an effective monetary policy. The point is that you can’t have it all: A country must pick two out of three. It can fix its exchange rate without emasculating its central bank, but only by maintaining controls on capital flows (like China today); it can leave capital movement free but retain monetary autonomy, but only by letting the exchange rate fluctuate (like Britain—or Canada); or it can choose to leave capital free and stabilize the currency, but only by abandoning any ability to adjust interest rates to fight inflation or recession (like Argentina today, or for that matter most of Europe).
Elaborating on the Impossible Trinity
Although Krugman’s summary was written in plain English—readers who want a more technical account with citations to the literature should check out the Wikipedia entry on the “Mundell-Fleming model”—let me elaborate a bit, to make sure the reader understands the dynamics involved.
Imagine the US is initially in equilibrium with Japan, and it takes $1 to buy ¥100 in the foreign exchange markets. Now suppose that at current prices, interest rates, exchange rates, etc., American consumers suddenly want to spend $1 billion more on Japanese cars, and that this is not counterbalanced by any desire from Japanese consumers to import more American-made goods. What happens? We will analyze the three possibilities from a Mundellian perspective, letting the authorities pin down each of the elements in the “impossible trinity” in the respective options. (Note that our discussion of the first option will be the longest, while we set up the framework, so that our discussions of options 2 and 3 can be brief.)
Mundell Option #1: Choose free capital movement and effective/independent monetary policy, while giving up a fixed exchange rate (i.e., allowing flexible exchange rates).
In this first scenario—which corresponds to the policy suite adopted by the United States and Japan today, incidentally—the authorities would allow the desired transactions to go through; the US consumers would get to buy their $1 billion in additional cars. This increase in the trade deficit would be matched by a corresponding Japanese investment of financial capital in the (net) accumulation of American assets.
In principle, the Americans could literally send hundred-dollar bills across the ocean that the Japanese would then add to their safe-deposit boxes; this would show the Japanese increasing their portfolios by $1 billion worth of American-issued financial assets (specifically, currency). However, a more typical outcome would be that the American importers would first enter the foreign exchange markets with their $1 billion and use them to buy Japanese yen. They would then use the yen to buy the cars from the Japanese sellers, who (of course) typically conduct their business in their own domestic currency. Meanwhile, those in the forex market who had sold their yen for dollars would then take the new $1 billion in US money to invest in American assets, such as Treasury debt, US stocks, real estate in Miami, etc.
Now here’s where things get tricky. In this scenario, we are supposing that the thing that upset the original equilibrium is American car buyers suddenly deciding to import additional vehicles from Japan. The only way to make that increase in the trade deficit “balance” is for Japanese investors to increase their investment in American assets—which remember, can be debt claims—by a comparable amount, measured in dollars.1
But if we were originally in equilibrium, where Japanese investors were happy with their holdings of American assets, then to induce them to hold an additional $1 billion worth, the American assets have to become more attractive. Specifically, their expected yield has to increase. In the case of Treasury securities, that means the interest rate on US government debt has to increase, so that the Japanese investor is willing to add more of it to his portfolio than before.
Yet hold on. In this Mundell Option #1, the central bank is still allowed to practice “effective monetary policy.” Presumably the Federal Reserve doesn’t want to let US interest rates rise above its original target level just because American car buyers want more Japanese vehicles. So if the American and Japanese authorities are going to allow the desired transactions to go through—thereby maintaining free capital movement—and the central bank is going to maintain control over its monetary policy decisions, then the Fed will act to combat the upward pressure on interest rates.
Typically, the Fed will engage in open market operations (which I explain in this chapter), buying assets from the market while creating additional dollars “out of thin air.” By flooding the credit markets with new reserves and by taking Treasury debt onto its own balance sheet, the Fed’s actions will push down US interest rates back to the original target level.2
All of the extra money printing will move exchange rates, however. (Remember, in Mundell Option #1, the element of the trinity that couldn’t be maintained was a fixed exchange rate.) In order to restore equilibrium, the US dollar will fall against the yen, so that instead of buying the original hundred, perhaps now a dollar only trades for ninety-nine yen. Other things equal, as the dollar weakens against the yen, it makes Japanese cars appear more expensive to American importers, and so this movement in exchange rates will stanch the new flows of cars and capital.
Now that we’ve carefully walked through the details of one leg of the “impossible trinity,” we can quickly cover the other two.
Mundell Option #2: Choose a fixed exchange rate and effective/independent monetary policy, while giving up free movement of capital.
This second scenario is quite easy to explain: when our hypothetical American consumers want to import more Japanese cars—or if Japanese investors want to buy more American assets—the respective authorities just say no. By strictly limiting international transactions to only approved channels and amounts, the authorities retain a free hand to set whatever targets they want for domestic interest rates and the foreign exchange rates on their currencies. If their citizens complain that they see a better car deal or bond available in a foreign country, their political officials tell them, “Tough! If you’re so unhappy here, move to Somalia.”
Mundell Option #3: Choose free capital movement and a fixed exchange rate, while giving up control over monetary policy.
In this third and final scenario, we suppose that the American and Japanese authorities want to allow their citizens to import cars and export financial capital however they’d like. However, they don’t want the vagaries of consumer and investor demand to affect the USD-JPY exchange rate, which (we suppose) is permanently locked in at 1:100.
Therefore, when American importers want to buy more Japanese cars, thereby putting upward pressure on US interest rates (to induce Japanese investors to hold the corresponding extra amount of American assets), the Federal Reserve has to allow it. This is the case even if the Fed’s own views on unemployment and inflation say that a “rate hike” would be bad for the US economy. Indeed, not only would the Fed not engage in looser policy (as in option 1), but depending on the specifics it might actually have to sell some of its own assets and soak up dollars from the system in order to keep the dollar from falling against the yen. Other things equal, Fed officials might worry that such tightening would slow US economic growth and keep US unemployment higher than they desired, but such is the corner into which Robert Mundell has painted them.
An Austrian Critique of Mundell’s Impossible Trinity
Admittedly, my narrative above was somewhat loosey-goosey; perhaps Krugman would have explained the cause and effect differently (see for example this Wikipedia entry), and Mundell in his published papers of course spelled out a formal model with all the i’s dotted and t’s crossed. Even so, I’ve given a fair flavor of what Mundell meant by his claimed “impossible trinity” or what is also called a “policy trilemma.”
Yet from an Austrian perspective, this apparent tradeoff is spurious, particularly in the long run. Under the classical gold standard (which prevailed among the advanced nations up to the eve of the First World War), the participating countries enjoyed free movement of capital, fixed exchange rates, and their central banks were practicing the most “effective” monetary policy in fighting recessions that was available to them. In this respect the situation is analogous to the alleged tradeoff between freedom and security: when citizens give up their liberties so that the authorities can (allegedly) protect them, they end up exposed to more danger.
I explain the operation of the classical gold standard in great detail in this chapter. For our purposes here, I will highlight two key points:
First, under the classical gold standard, each sovereign nation defined its own currency in terms of a specific weight of gold. This implied a fixed exchange rate between these sovereign currencies. For example, in the year 1913, the British government stood ready to redeem its currency at the rate of £4.25 per ounce of gold, while the US government would redeem its currency at the rate of (approximately) $20.67 per ounce of gold. These respective policies implied—using simple arithmetic—that the exchange rate between the currencies was fixed at about $4.86 per British pound.
Yet this exchange rate between the dollar and the pound wasn’t “fixed” by coercion, in the sense of literal price controls; there was still a free market in foreign exchange. What happened in practice was that if the actual market exchange rate of dollars for pounds deviated too far from the anchor point of $4.86, it would become profitable for currency speculators to ship gold from one country to the other in a series of trades that would push the market exchange rate back toward the “fixed” anchor point. (See my chapter for more details on how this worked.)
The benefit of a fixed exchange rate is that it allowed individuals to make economic plans involving foreign commerce much more confidently, because they wouldn’t have to worry about movements in each currency. Imagine the difficulty of being an American entrepreneur if each of the fifty states issued its own fiat currency that “floated” against all the others.
The second point to highlight is that yes, Mundell is right that under the classical gold standard, the central bank had to subordinate its other policy goals to maintaining adequate gold reserves. If changes in consumer and investor attitudes put downward pressure on a nation’s currency, this would lead to gold outflows. If the central bank were to maintain the fixed redemption rate of its own currency in gold, it would need to reverse this outflow of gold, and so the central bank would have no choice but to tighten its monetary policy, by slowing the creation of new currency (or even absorbing some of the outstanding stock), and/or allowing domestic interest rates to rise.
Yet if the Austrian theory of the business cycle (explained here) is correct, then letting markets—rather than central banks—determine interest rates is precisely the way to avoid the boom-bust cycle that so plagues market economies. In other words, what Robert Mundell or Paul Krugman refer to as “independent” or “effective” monetary policy should actually be translated as “the tool with which central banks fuel an unsustainable boom that leads inevitably to a crash.”
Now it’s true that even under the classical gold standard, the market economies suffered from periodic panics and depressions. But even here—if Mises is correct in his diagnosis—the problem is that even the commercial banks have too much “discretion” over interest rates, derived from their ability to create and destroy money (broadly defined). So to be clear, I am not in this essay claiming that the classical gold standard by itself prevented recessions. Rather, I am saying that tying the central bank’s hands—in a way that a Mundellian might describe as “forfeiting effective monetary policy”—is a necessary (but insufficient) condition to eliminate the boom-bust cycle.
Fans of the Austrian school would benefit from reading more about the Mundell-Fleming model, at least intuitive explanations. I know that I personally understood more about international trade and capital flows after working for Arthur Laffer, who himself published in this framework. To be clear, the canonical Austrian treatments (such as in Mises and Rothbard) of these topics are correct, but there are subtleties that Human Action and Man, Economy, and State don’t cover.
Even so, what most economists conclude from Mundell’s theoretical analysis is nonetheless wrong. To wit, we don’t do any favors for economic growth or the labor market by giving central bankers a “free hand” to create money and set interest rates however they see fit. Later in his career, when he became associated with the supply-side revolution, Mundell himself may have embraced the virtues of hard money and eschewed macro “fine-tuning,” but his insights still pale in comparison to those of Ludwig von Mises.
- 1. More accurately, the trade balance is part of the “current account” balance. In terms of simple accounting, it is always the case that a current account deficit is counterbalanced by its mirror image in a “capital account” surplus. For a more comprehensive explanation, see my article from early 2007. Note that although the accounting explanation of the trade balance is correct, at the conclusion of that article I criticized Peter Schiff’s gloomy forecast—for which I wrote a mea culpa in this later piece.
- 2. Perceptive readers may wonder why Japanese investors would continue to hold more Treasury debt if the Fed’s actions undo the higher US yields that we originally supposed would be necessary to make them willing to do so. We will refrain from spelling out the answer, because the story we are telling in the text is somewhat ad hoc. In a more formal setting, we would have to carefully spell out all of our assumptions—including the possibility of other investors in different countries, who might rearrange their own portfolios when the Fed brings extra Treasurys onto its balance sheet. One of the loose ends in our own story is that we didn’t explain how the US consumers came up with an extra $1 billion to spend on Japanese cars.