Nominal devaluation and real wages

(apologies for my English: constructive remarks, or even destructive, are welcome, if you dare...)

When going gets tough, the tough get going. As the perspective of a Grexit becomes more likely, thereby confuting the ludicrous thesis of a supposed "irreversibility" of the euro, the propaganda resorts to its weapons of mass destruction: Nobel Prizes. I used to say that only two human things are irreversible: stupidity and death (although in the second case we have a brilliant counterexample). I was mistaken: there is a third human thing which seems to be irreversible: the firm belief of economists in the fact that a nominal devaluation will bring about a disastrous fall in real wages. After showing you some data, I will leave to your prudent appreciation to evaluate whether this belief falls in one of the two above mentioned categories (Occam's razor).

This morning I received through Twitter this opinion of Prof. Pissarides: Grexit means biggest haircut of wages ever! A very bold statement, but... is it supported by any model, or any data?

We all know (or should know) that Grexit will not be the first end of a currency union, and certainly not the last. We all know (or should know) the relevant scientific literature, that features top authors such as Andrew Rose. As a consequence, over time we have gathered a lot of evidence, and of particularly authoritative interpretations. History never repeats itself exactly: so true! We all know that it repeats itself first as a tragedy, second as a farce. The behaviour of mainstream economists during the 1929 crash was tragical (think for instance of Irving Fisher "high plateau", so close to Alesina's dismissing "strong correction"). Nowadays the mainstreamers' behaviour is becoming farcical: most of them either assert unsupported claims, or reinvent the wheel.

It is reinventing the wheel to show now a commendable, but definitely not timely, concern about democracy in the Eurozone. When I wrote my book on this subject in 2012, I knew that I was the last one to express such a concern and I quoted all the authors that had expressed and supported with empirical evidence a similar worry: among them, Rose (see above). Dwarf on the shoulder of giants.

It is an unsupported claim that a fall in nominal exchange rate will bring about a one-to-one fall in real wages, i.e., Pissarides's claim that the necessary realignment of the new dracma will make the Greeks poorer. Anything similar did never happen (with an exception that we will study below). Moreover, since colleagues could (rightly) argue that the past does not always teach correct lessons about the future, I add that I do not know of any formal economic model where a nominal devaluation results in a one-to-one fall in real wages. This is likely to be a limit of my knowledge. However, I know of models where devaluations are an effective tool for fostering economic growth in the presence of real wage resistance (one is here). It is noteworthy that after two hours of inquiries to my mainstream colleagues through Twitter, I received no clear indication of any standard textbook model where a nominal devaluation causes a one-to-one fall in real wages.

There are two reasons for that: one is theoretical, the other is practical.

From the theoretical point of view, if you are a mainstreamer, you must believe (yes: believe) that real wages are tied to the marginal productivity of labour. In a world where money is a veil (the mainstreamers' world), it is dubious whether and how this relation between real variables (productivity and real wage) could be affected by a nominal variable (nominal exchange rate).

The practical reason why no mainstreamer was able to answer so far is that such a one-to-one connection between nominal devaluations and real wage falls never presented itself in practice. As a proof, I show below the patterns of the Nominal Effective Exchange Rate (NEER) and of the Real Wage (RW) indexes in the aftermath of some well-known major devaluation episodes, some of which implied the exit from a fixed exchange rate agreement (which is not a currency union, but is close enough to it). The data come from the International Financial Statistics:

XXX..NECZF...
XXX64...ZF...
XXX65...ZF...

(where XXX is the country code, NEC is the nominal effective exchange rate index, 64 and 65 are the consumers' price and the nominal wage index).

I plot the results below for your reference (country names in Italian: apologies, this is old stuff coming from this ancient post):

In no case did real wages fall as much as the nominal effective exchange rate, in some cases they fell by a smaller amount, and in some other cases they increased. The claim that a nominal devaluation is equivalent to a real wage haircut is disproved by the data.

I mentioned an exception. Here you are:

(Messico means Mexico, just in case you are a mainstreamer).

Mexico is the only country that seems to support Pissaride's informal model of a one-to-one long-run equivalence between nominal exchange rate fall and real wages "haircut". Is this a surprise? No, not at all. As every professional economist knows, there is a lot of evidence showing that the long-run pass-through from nominal devaluation to prices is above one in Latin America. The main reference is here (and it is not a negligible one: it has 384 quotes on IDEAS, and it is quoted in one of the most authoritative handbooks of international macroeconomics; btw, I studied with Prof. Gandolfo).

Roberto Frenkel (this one) once suggested me an interesting explanation of why this happens (i.e., of why, contrary to what happens in the other regions of the world, in Latin America inflation overshoots devaluation, with obvious negative consequences on real wages). In those countries the US dollar is seen as the typical safe-haven asset. In case of fear of devaluation, this feature sets out a self-fulfilling mechanism where everybody buys dollars, thereby pushing down the national currency, and you can imagine the rest (even if you are not a professional economist)!

But Greece is definitely not Mexico. Despite being relatively backward with respect to its unfriendly European partners, we can suppose that its financial market is relatively more developed in 2015 than Mexico's was in 1995. This self-fulfilling vicious circle is extremely unlikely to present itself in Greece (as it did in Argentina or in Mexico).

This is what a professional economist can say. But professionalism, in all evidence, is the slightest concern of my profession. Economists nowadays behave more like social media influencers than like researchers. Their claims are rarely fact based, and are always connected to the ongoing political debate in an extremely unpleasant and suspect propaganda-like way.

This is a really unfortunate outcome, because we are in an economic crisis, and it would be extremely important for all of us, and for our democracy, to consider economics as a credible science, and economists as experts who stick to the facts and try to explain them, rather than as influencers committed to deliver a storytelling increasingly remote from reality.

So is life.


(...in case someone has other counterexamples, please feel free to intervene. In case someone has models, please feel free to quote them in full, according to academic standards. Thanks for your attention.)