The Impossible Trinity Revisited

The Economist's brief is available here (subscription required). I have discussed the limitations of the Mundell-Fleming model in some posts (here and here, for example). But I have not discussed the limits to the trilemma fully. In this paper, I argue that under certain circumstances, associated to what Ben Cohen calls the geography of money, the trilemma might not hold. The countries at the top of the pyramid with convertible currencies do not face the harsh trade offs of the countries at the base.

As I said back then:

"The United States during the current crises is an example of the reduced impact of the trinitarian trade-offs for countries with convertible currencies. In fact, even though the economy was in a deep recession by the last quarter of 2008, the dollar, which had depreciated considerably over the previous years, started to appreciate as investors increasingly fled to Treasury bonds for safety. In other words, even though rates of interest were reduced to deal with the recession and the financial crisis, and capital mobility was preserved, a certain degree of exchange rate stability was maintained."

If you're not at the top, I would recommend to give up the complete free mobility of capital, which should be no surprise.

Capital controls and exchange rates

This was the topic of the RBI/ADB conference in Mumbai. No particular surprises. The consensus is that capital controls affect the composition of flows, but not their volume, and even the IMF, represented by Jonathan Ostry, suggested that capital controls should be part of the tool kit used by central banks. Also, some skepticism on the efficiency of short term (or episodical controls, such as the ones used by Brazil) was raised. Of course there are still differences on what circumstances capital controls are actually necessary.

Most of the discussion was related to the use of capital controls to reduce the risk of appreciation, since in the last decade developing countries have had to deal with inflows and a depreciating dollar. Note, however, that capital controls were not thought when originally defended by Keynes and White at Bretton Woods to be necessary for reducing appreciating tendencies, but to limit capital flight (and avoid and external crisis) and provide monetary autonomy (Impossible Trinity or Trilemma).

Further, in historical perspective, exchange controls (capital controls on quantities not prices) have been used as an instrument for industrial policy, to determine that the use of dollars is prioritized for capital equipment imports particularly in the periphery. In that sense their use is in fact an essential an permanent tool in the developing economy set of instruments, and that's why signing FTAs or BITs that reduce the ability to deploy capital controls is dangerous.

PS: On the long standing issues related to capital controls and exchange rate regimes see this paper.