Dollarization and Foreign Shocks in Latin America
Dollarization and Foreign Shocks in Latin America
Foreign shocks affect dollarized economies more than countries that have their own central banks... or do they?
The original version of this post, in Spanish, can be found here.
Dollarization eliminates the possibility of freely conducting monetary policy. A common criticism of this reform is leaving the country unprotected from foreign nominal shocks. It is a bad idea to give away a well-behaved central bank that can minimize the impact of these shocks. If you have been reading my recent posts on dollarization, you won’t be surprised to find out I have a couple of issues with this line of argument (you can find a link to my previous post below). The first one is theoretical, and the second one is empirical.
Theoretical limitations
There is no dispute that in theory a central bank can accommodate its policy in the presence of foreign shocks. The problem with this argument is that this ideal central bank is not a feasible option for a country like Argentina. Referring to how the central bank can behave in theory is, once again, falling for the Nirvana fallacy.
In a recent tweet, Argentine Congress Representative Luciano Laspina (JxC) uses a well-known passage by Milton Friedman (p. 165) on the benefits of having a flexible exchange rate to question those who propose a dollarization reform.
There is nothing wrong with Friedman’s passage other than it assumes a well-behaved central bank, for which there is no clear evidence this is an option for Argentina.
But I digress. I mentioned two theoretical limitations I see in this line of argument. Here they are:
Central banks can themselves be a source of disruptive shocks. Critics should contrast external shocks with the reality of domestic shocks (inconsistent monetary policy).
Some external shocks are the indirect effect of inconsistent domestic monetary policies. Critics should also discount external shocks resulting from inconsistent domestic policies.
In addition, as Calvo (2002) points out, (a) in Latin America devaluations have been mostly contractionary and (b) the impact of the foreign shock can be augmented by said devaluation (because of debt denominated in US dollars). Because central banks can themselves be a source of significant shocks, it is not a priori true that dollarization is worse than having your own central bank. This should be evident to anyone somewhat familiar with Argentina’s economic history.
Empirical limitation: The blackboard meets the data
If dollarization leaves a country unprotected from shocks, then these countries will suffer more from external shocks than non-dollarized countries do. This unprotected dollarized economy is what we get on the blackboard. But, does it show up in the data? The 2008 crisis was a shock of historic proportions. Therefore, we should expect to see an economic crisis of historic proportions in dollarized economies.
We can compare real GDP percent change of three Latin America dollarized countries (Ecuador, El Salvador, and Panama) with that of Argentina. We should notice a sharper fall in the real GDP of dollarized countries than in Argentina. You can see by yourself that this is not the case.
Argentina’s real GDP fell 6.9%. El Salvador’s real GDP also fell, but much less than in Argentina, 2.1%. On the other hand, the other two dollarized countries maintained a positive change in real GDP. Their economies slowed down rather than going in reverse; not what we expect to see in the face of a shock as impactful as the 2008 crisis was. If anything, the graph shows the opposite of what critics worry about. This result is no surprise if we accept that central banks can be so poorly managed that they produce more damage than wearing the ‘straitjacket’ of dollarization.
Of course, a more precise analysis than casual observation would rely on a controlled regression with a better diagnosis of how much a external shock affects a dollarized economy compared to other variables. But, this type of analysis defeats the purpose of my point. The 2008 crisis was of such magnitude that the difference should be observable in plain data if critics are right about worrying so much about this issue.
Bonus: Covid-19 in Dollarized Countries
What about real shocks such as Covid-19? Admittedly information is still incomplete as 2021 data is not yet available. We can, however, look at some preliminary data for 2020.
Unless you are Panama, if you are dollarized your real GDP suffered less than that of Argentina. A few clarifications are relevant. First, in 2020, Ecuador was not only hit by Covid-19; it also declared default on its sovereign debt, yet its economy suffered less than Argentina’s. All three dollarized economies saw a fall in real GDP but maintained a stable price level compared to Argentina, whose inflation rate remains out of control. Third, Panama did not suffer from a financial crisis or bank runs despite its sharp economic fall. Neither did Ecuador or El Salvador for that matter.
Finally, real and nominal shocks affect output and inflation differently. A negative nominal shock makes output and the price level fall. But a negative real shock makes output fall and the price level rise. This poses a dilemma for central banks: (1) react to the fall in GDP, (2) great to the rise in inflation, or (3) do nothing?
There is not much monetary policy can do in the presence of a real shock. An expansionary monetary policy can minimize the fall of GDP in the short-run at the cost of increasing inflation further. Alternatively, reducing inflation comes at the expense of a deeper GDP fall.
Therefore, how dollarized countries react to real shocks is less relevant than how they respond to nominal shocks. Yet, whatever conclusions can be drawn from real shocks, that dollarization is a curse does not look like one of them.Previous posts on this topic:
This is easy to observe in a graphical exposition of an AD-AS model.