Robert Mundell, father of the Euro dies
His "Theory of Optimum Currency Areas" was published in the AER in 1961
It´s taken as gospel that the US is a better OCA (Optimal Currency Area) than the Eurozone. But is this really true? (This might make an interesting post)”.
Mark´s reaction was triggered by a post by Kash, which, in turn, picked up on a post by Gavyn Davies.
I must say that the OCA discussion was revived last year when the Eurozone crisis flared up with the Greek “blow-up”. This post by David Beckworth – and the links within – is a good reference to the debate. At the end of the post he writes:
I will go one further in this debate. It is not clear to me even now that all of the United States is an OCA. Do we really think Michigan and Texas over the past decade or so benefited from the same monetary policy?
And do we think both states had an adequate amount of economic shock absorbers? Given the vast differences between these two states in their business cycles, diversification of industry, union influence, and wage stickiness it easy to wonder whether these states should belong to the same currency union. Yes, they have access to fiscal transfers, labor mobility is great (I myself left a job in Michigan for this one in Texas), culturally they are similar, and politically there is will for the dollar union.
Still, given the disparate impact of U.S. monetary policy on different regions of the country one does wonder whether all the United States is truly an OCA.
In singling out Michigan and Texas, DB is following up on a much earlier post of his that discussed differential impacts of Monetary Policy:
The differential responses of these two states to the same monetary policy shock are striking. Texas is hardly affected relative to the steep downturn in Michigan. As noted above, these patterns fall more broadly into the regions of the United States with different sensitivities to the federal funds rate shocks.
Our research confirms early studies that show these regional differences can be partly explained by the composition of output: those states with a relatively high share in manufacturing get hammered by a monetary policy shock while those states with relatively high shares in extractive industries fare much better.
We also find that states with a relatively high share in the financial sector fare better as well. Finally, we find that states that have (1) a relatively high share of labor income compared to capital income and (2) a relatively high rate of unionization also get hammered by monetary policy shocks.
Conventional wisdom has it that to “qualify” as a OCA, there should be high business cycle correlation (symmetry of shocks) among members and individual members should possess “economic shock absorbers” – flexible wages & prices, labor market flexibility, labor mobility and a federal fiscal authority that coordinates fiscal transfers.
Figure 1 depicts this set-up for the Eurozone countries. The correlation between member counties and eurozone RGDP is straightforward. The shock absorber index is constructed from a simple average of an index of protection (rigidity) in the labor market (EPS index constructed by the OECD) and the inflation rate (as proxy for downward price rigidity).
The outcome is what was expected. Portugal and Greece are well within the OCA frontier. Spain and Ireland are polar opposites. Spain has a high business cycle correlation and low shock absorbing capabilities while Ireland shows the opposite characteristics.
Figure 2 shows the business cycle correlation of 8 US regions. Even though, as mentioned by DB above, there are differences in the shock absorbers of different states (maybe not so much when they are aggregated in regions), an index that quantifies these differences is not available. Nevertheless it is noticeable that the range of the variation in business cycle correlation is about the same among Eurozone countries and US regions.
Mark chose to compare the difference relative to trend among states. To keep the analysis consistent, I compare the business cycle correlation of individual states. This is depicted in figure 3, which classifies correlations as high (H,>0.8), medium (M, 0.50 – 0.79), low (L, 0 – 0.49) and negative (N). Here the differences are enormous, with five states showing negative BC correlations and another four registering low correlation!
Michigan (corr=0.85) and Texas (corr=0.69) are singled out in the picture. So, if as in DB´s example, Michigan and Texas make him wonder if the US is an OCA, what is implied by the differential between states above the 0.8 line and those below the 0.50 cut-off point? If the OCA “conditions” were in fact “necessary conditions”, the US should have “blown-up” as a Union a long time ago!
But we know it hasn´t. So why not? Maybe shared language, custom and culture in addition to a central government are more important than some economic indicator conditions. Those are exactly the conditions that the Eurozone countries do not possess.
Figure 4 shows another aspect that might be relevant for “binding the Union” together. The 6 largest states have about 42% of income (RGDP) and 41% of the population while the 6 smaller states have about 1.3% of income and 1.1% of the population. Pretty equitable, no?