Sunday, May 9, 2010

Krugman, Mankiw, and Beckworth on OCAs and Optimal Government Size

David Beckworth has an interesting post on his blog (Macro and Other Market Musings) about what optimal currency areas (OCA) are, titled “Krugman, Mankiw, and the U.S. as an OCA”.

He quotes Mankiw’s comment on Krugman’s claim that the euro is doomed because the area lacks a central government capable of redistributing a large amount of a centralized tax:

« A large part of [Krugman's] argument is that Europe is not an optimal currency area because it lacks a large central government enacting transfer payments among the various regions... Is that right? I am not so sure. The United States in the 19th century had a common currency, but it did not have a large, centralized fiscal authority. The federal government was much smaller than it is today. In some ways, the U.S. then looks like Europe today. Yet the common currency among the states worked out fine.”

And Beckworth goes on:

« Mankiw attributes the success of the U.S. currency union in the 19th century to wage flexibility and labor mobility. He notes, though, that Greece and much of the Eurozone lack these and thus the Euro experiment may be doomed. I agree with Mankiw that the Eurozone problems are more than just the lack of a centralized fiscal authority. »

The discussion becomes, at this point, a bit confusing. Let’s try to clarify it.

The original OCA theory did not include a central, federal government as a condition for an OCA to exist. The conditions as formulated by Mundell, McKinnon and Kenen were that trade between would-be members of the currency area should be large as a percentage of GDP, concentrated primarily among these partners, that the countries size would be quite different, factors of production mobile between them, business cycles and inflation rates correlated, and the structure of production relatively similar so that macroeconomic shocks would not be asymmetric from one country to another.

A large federal government appropriating a large fraction of GDP through taxes, and thus able to transfer income from high growth regions (or states) to depressed ones, was not considered a basic condition for an OCA, but rather as a compensating shock absorber for a non optimal currency area. The shock absorber is what enables such a non-optimal monetary zone to function almost as an OCA. But nevertheless a non-optimal zone with the right (centralized) type of government is less efficient economically than an OCA because the high level of central taxes necessary for its working is heavily distorting.

That the eurozone is not an optimal monetary area derives not from the lack of a federal European government similar to the U.S. one, but from basic characteristics of the European economies, particularly the low international mobility of labour, the low correlation of business cycles and inflation (mostly), and the different structures of production and specialization, as well as the rather similar dimensions of several of the member countries. Thus, even if European countries were able to build up a federal government right now (which is rather unlikely for reasons I develop below) the eurozone would not be an OCA and would remain economically dominated by a regime of independent national currencies.

Now, what about history? Is it true, as Mankiw says, that wage flexibility and labour mobility in the 19th century made the U.S. an OCA? I tend to agree since these are two basic conditions of OCAs in theory. And I would add as an illustration that the large international mobility of labour in that century (the “age of mass migration”) could explain, in the OCA theory perspective, why the gold standard prevailed in that period: the gold standard was in fact a common currency area, and it could be established, without a common government, among countries among which capital and labour moved rather freely, as well as goods and services, and where business cycles and inflation accordingly tended to move together.

Large centralized governments, and large central banks, came later, with the second industrial revolution by the end of the 19th and beginning of the 20th century, and with them large empires evolving into separate economic and monetary areas during the “first 20th century” (see my book referenced below). The Federal Reserve System was not completely established before 1913, and the European “gold bloc” disappeared shortly after WWI.

All this means that the non-optimal U.S. currency area could become workable – even though suboptimal - due to the growth of the federal government by the end of the 19th century. A larger government and larger federal taxes provided a growing “asymmetric shock absorber” in a currency area that was still not optimal. It followed that, from that point, monetary policy could be centrally managed, and the Federal Reserve established for that purpose. As mentioned by Beckworth, Hugh Rockoff makes the case that the U.S. economy did not become an OCA until the 1930s, but in my view, and as several U.S. economists claim, the U.S. is still not and OCA today! What changed during the 1930s was that the “shock absorber” became big enough to be effective.

In the case of Europe, the promoters of the euro hoped indeed that its creation was a first step towards the centralization of government at the continental level.

But governments do not grow at will, in spite of the experience of their continuous growth that characterized the last century. As I showed in “The Second Twentieth Century” (Grasset 2000, Hoover Press 2006, and see also my paper “The Shrinking Hand” on my SSRN page) large hierarchies grow when there is a growing scarcity of information, relative to total output. When information grows faster than production, the contrary result obtains: hierarchies, public as well as private, shrink and many disintegrate. Large and/or heterogeneous states implode: the USSR, Czechoslovakia, Yugoslavia …

These are not good conditions to integrate several independent nations into a large federal state as the U.S did, but in 1866, on the eve of the second industrial revolution. In the still evolving information revolution, it is quite unlikely that European countries can coalesce into a single federal state. As a warning of that impossibility, the project of an “even closer union” has been repeatedly rejected by voters in several countries during the last several years.

This is why the European non optimal monetary area will not survive, contrary to the U.S. one (probably, even though economists come to realize that the U.S. is not an optimal currency area, and is thus submitted to a welfare loss by sustaining it), because European nations cannot build, in today’s conditions, a large enough shock absorber.

No comments: