The IMF, that has been praised from some corners for its willingness to change and adjust its stance as the situation demands, had an institutional paper out recently on the use of capital controls. While history tells a different story, this time is different and Blanchard/Ostry, the Director and Deputy of the research department, have a brief, explanatory follow-up note out at voxeu.
What the essentially elaborate upon is the main tenet of the paper - i.e: the acknowledgement that complete capital account liberalisation, depending on the macro and financial environment (among others), may not be the right goal for a particular country.
Harry Dexter White and Keynes, during the Bretton-Woods era, essentially laid out a fairly obvious but crucial point that holds enormous relevance even today - that CFM (capital flow management) as a toolkit is severely limited when initiated unilaterally. The restriction of movements of capital would be most effective when "controlled at both ends". This is akin to asking for cooperation among different countries.
This emphasis on multilateral actions is backed by four instances according to the authors. Basically:
1) If an economy is in a state where an adjustment in the external balance is warranted, it could look to capital controls as means of bypass. Essentially, this is a beggar-thy-neighbour attitude of using controls to sustain an under-valued currency. In the note, Blanchard and Ostry state outrightly, "This is why one normally thinks that capital controls whose purpose is to frustrate external adjustment are multilaterally aberrant". Strong words with room for a benefit-of-doubt in the sense that the reason for controls could also be directly linked to the stability of the financial system in an economy (think foreign borrowing reliance etc.)
2) A more insidious case of controls would be related to an indrect attempt at manipulation of the intertemporal TOT for an economy. This is essentially the same as using tarriffs and subsidies to tinker with terms-of-trade but would rarely be something seen in practice blatantly. The authors have a charming way of putting this point across - "it is not beyond credulity to think that some policies that affect capital flows can materially move world interest rates in a direction that benefits the country. To see this, ask yourself whether there are any large creditor countries that maintain restrictions on capital outflows; or whether any large debtor countries pursue policies that push down world interest rates." Ahem!
3) Controls that are put in place to deal with externalities (production-side) in the export sector. This goes back, in a way, to parts of the first instance, i.e: production taxes/subsidies could be a response to externalities but are more difficult to put into practice due to budgetary constraints etc. Theoretically, a situation such as this could be alternatively dealt with by a control-supported currency devaluation which would have negative consequences (distorting the consumption/production effects and decisions), As one can see, there's asymmetry due to the impact on consumption.
4) This goes back to the latter half of the first instance where measures are undertaken for the purpose of financial stability. Think foreign borrowing risks if not internalized by the borrower - in such a case, capital controls can act as a Pigouvian tax (tax on an activity that generates externalities) and internalize these external effects.
What capital controls come down to in a multilateral scenario is the existence of spill-overs and subsequent financial stability issues. What the authors try to emphasize is that if capital controls had no costs for a country, there would exist a Nash equilibrium where any country chose controls at a certain level to check its own financial stability risks.
And that's the problem. Because imposing controls creates costs and this is where coordination is required (risk of capital control wars, excessive flows etc.) between both the source and the recipient. In such a case, naturally, the equilibrium ceases to be efficient because each country still has one instrument but now two targets (the externality domestically and the cost of the measure).
The note concludes by admitting the difficulty of this possible paradox. In certain cases, some countries might have no interest whatsoever in bearing the cost of financial stability in another country. Coordination however, should not be thought of as a pursuit of interests contrary to domestic policy but rather, to ensure that domestic objectives are met while the spill-over damage done is minimized.