Monetary Unions and Endogeneity of the OCA Criteria

Abstract

This article examines an endogeneity issue within the Optimum Currency Area (OCA) theory. According to the cost-benefit analysis, we found that there are the upper and the lower bounds in the degree of monetary integration for a monetary union to be created. We also found that a country may secede from the monetary union, depending on its degree of integration. A country may also secede when production specialization is facilitated with monetary integration within a framework of the “OCA line”. We also consider the endogeneity of the “OCA Index”, and applied our analysis to the optimum number of world currencies.

Keywords:

• Monetary Union
• OCA
• endogeneity

Notes

1. For extensive discussions on the OCA criteria, see, e.g. Lafrance and St-Amant (1999) or Mongelli (2002). For monetary union in general, see e.g. Jovanović (2006).

2. For detailed examinations of the Rose effect from both theoretical and empirical considerations, see e.g. Rose (A. K. Rose, A meta-analysis of the effect of common currencies on international trade, unpublished manuscript, 5 March 2004) and Rose and Stanley (2005) who summarized that a currency union has a positive effect on trade. See also Baldwin and Taglioni (2004), Baldwin et al. (2005), and Baldwin (2006a,b).

3. Chaplygin et al. (2006), p. 56) mention that “a necessary and sufficient condition for low costs is both a high positive correlation between the shocks..., and shocks of similar size.”

4. Mongelli (2002, pp. 28–29) calls this hypothesis the “endogeneity of OCA hypothesis”.

5. Dixit (2001, p. 593) also summarizes succinctly that “more trade can create more specialization and therefore more asymmetric econometric structures and different susceptibility to supply shocks.” Mongelli (2002), pp. 28–29) terms this hypothesis the “Krugman specialization hypothesis”.

6. In the empirical part of Kalemli-Ozcan et al. (2001), they demonstrated that the Organization for Economic Cooperation and Development (OECD) countries and US states with higher industrial specialization exhibit output shocks that are less correlated on average with aggregate OECD output and US output, respectively. Fidrmuc and Korhonen (2006) present empirical assessment of the literature on the business cycle correlation between the Euroland and the Central and East European Countries (CEECs).

7. A notable theoretical approach is Corsetti and Pesenti (2002) who considered the issue of the “endogeneity of OCA hypothesis” in a NOEM model.

8. However, Shin and Sohn (2006) and Sato and Zhang (2006) both found weak evidence of the pro-synchronization hypothesis for some East Asian countries.

9. In other words, beyond the threshold level κ*, the increased loss from specialization dominates the decreased loss by integration. Artis and Fellow (2006) also briefly mention this possibility due to specialization.

10. A member country with the degree of integration beyond κ ₂ or below κ ₁ will secede the union. Alesina and Barro's (2001) case may correspond to the former. See also Hochreiter et al (2002).

11. Under the pro-synchronization hypothesis, the curvature of the OCA line is shown to be in general concave. However, if the sign of H _TY ≡ ∂² H/∂Y∂T is positive and large enough to make d² Y/dT ²>0, then the OCA line is convex to the origin.

12. A positive sign is expected for the coefficient of the first variable, since the smaller this variable, the more synchronized the business cycle is and thus the lower variability exchange rates exhibit. A slightly different but similar reasoning applies to a positive (the larger this is, the higher the variability of exchange rates), negative (the larger this is, the more stable the exchange rates), and positive coefficients (the smaller a country is, the more gains from a common currency, i.e. the more stable exchange rates) are expected for the second, the third, and the fourth variable, respectively.

13. It should be stressed that they did never calculate (the present discounted value of) the net benefits from monetary unions.

14. For other empirical literature, see the surveys of recent empirical investigation on the OCA, e.g. Mongelli (2002) or Lafrance and St-Amant (1999), Fabella (2002, for Asian countries), and literature cited therein. See Bénnasy-Quéré and Lahrèche-Révil (2000) for CEEC countries and Horváth (2005) for 20 developed countries.

15. Aminian (2005), p. 95) mentions that, in East Asia, contrary to the traditional trend, “the financial proposals are moving faster than any serious intention of trade and investment cooperation.” Thus, as Gounder and Sen (2002) imply, if the Asian foreign exchange markets have long been integrated, specialization in production may not be a necessary consequence for Asia.

16. McCallum (2003) also pointed out that the intersection point may not be unique, since BB and CC curves take only integer values, and thus both are essentially step functions. Those curves depicted in Figure 3 are therefore linear approximations.

17. See also Alesina and Barro (2002) who conclude that the number of independent monies may be reduced with an increase in the number of independent countries.

18. If the CC and BB lines happen to intersect at point A on the vertical line starting from one (the single currency case), the benefits are equal to the costs. But, as we have discussed, such a point is unstable. If the CC line happens to be tangent to the BB line, the tangential point determines the (unique) optimal number of the world currencies with zero net benefits.

19. There are no explicit discussions so far about network externalities, although the externalities suggests there should only be one international currency (the euro or the US dollar?) (Eichengreen & Razo-Garcia, 2006, panel discussion section).

20. A recent example is Furceri and Karras (2006) who consider only “two particular determinants of benefits and costs” (p. 26), i.e. the business-cycle correlation and the inflation bias for the sample of 30 European countries. They conclude that “the evaluation of net benefits” is particularly difficult (p. 37), because of the two particular determinants indicating a “positive relation” (p. 36). Another example is Hughes Hallet and Weymark (2006) who considered the costs of heterogeneity in (1) asymmetries in transmission of monetary and fiscal policies, and (2) difference in national preference for price stability, output growth, and income redistribution for 16 European countries. They found that the welfare costs can “be considered to be quite large” (p. 143).