we found little support for the proposition that a flexible exchange rate is a useful policy instrument. ...the existence of an exchange rate makes surprisingly little difference to the extent of real devaluation. This is not to say that it is easy or even preferable to achieve real devaluation by other means (basically, internal contraction) but it is hard to sustain the argument that only nominal devaluations succeed. Furthermore, advocates of nominal devaluation may underestimate the real costs imposed on actors in the domestic economy: a particularly salient issue in Hungary and Latvia, where many households and firms had taken out foreign currency-denominated loans.I think this is too strong, at least for Hungary--I think the evidence is that nominal devaluation has been a significant advantage, even despite the foreign loans issue, and the differences in internal contraction are quite significant. At the chart below suggests, Hungary has seen much, much less employment loss in the course of the crisis than Greece or Spain. At the end of 2012, all three countries had seen nearly equivalent changes in real exchange rate, but Hungary's was almost entirely down to nominal exchange rate changes while those in Greece and Spain were purchased at the cost of huge increases in unemployment. Even recently, as Hungarian real exchange rates are moving up (hopefully because workers are getting paid more as employment recovers), the nominal exchange rate has played a significant buffering role. Obviously, there are other forces in play, etc., but I still think this is pretty convincing on the backdrop of clear theoretical reasons why we expect real devaluations to be costly.
|These are unit-labour-cost real exchange rates versus the Eurozone, if you care--it's not much different if you use the whole world. Sources: real exchange rates, employment|