Testing for nominal convergence in the Central American area: evidence from panel data unit-root tests

Abstract

Inflation convergence between the Central American Area and the United States is investigated using both recent homogeneous and heterogeneous panel data unit-root methods. Strong rejections of unit-root hypothesis are found, and therefore evidence of Purchasing Power Parity, in the Central American countries for the 1981:1–2005:4 period. Then by considering the nominal convergence criterion, the dollarization system seems to be suited to this region.

Notes

¹ Since the beginning of the 2000s, Ecuador and Bolivia implements a dollarization system and an intermediate system, respectively.

² Costa-Rica, Honduras and Nicaragua experiment an intermediate system. Guatemala has adopted a flexible exchange rate regime. But El Salvador and Panama have already chosen the dollarization regime.

³ Several arguments can be put forward as the better credibility of national monetary institutions, strong economic spillovers with the intensification of the partnership with the USA, and the presence of a spontaneous dollarization behaviour in the zone.

⁴ These bilateral real exchange rate series (against the US dollar) are obtained from ERS-USDA database at http://www.ers.usda.gov/data/macroeconomics. For Nicaragua, data are available over the period 1989:2–2005:4. Note that an increase in the real exchange rate means a real depreciation.

⁵ All unit-root tests consider a specification with a constant but without a time trend, because time trend in real exchange rates is not consistent with the long run PPP.

⁶ Both ADF and PP procedures are used when implementing the Choi test.

⁷ The results of tests are available from the authors upon request.