دوفاکتوی و رژیم نرخ رسمی ارز در اقتصادهای در حال گذار

The choice of an appropriate exchange rate regime is a venerable topic in international macroeconomics and finance. Theoretical literature discusses the choice of exchange rate regimes from various perspectives, such as optimum currency areas (OCA), optimal stabilization, monetary policy credibility, and currency crises. Empirical research applies theoretical guidelines to explain the observed choices of exchange rate regimes.1 Most of these studies, however, focus on the official regimes, i.e., those that the governments declare rather than the de facto regimes they actually pursue.2 The relation between official and de facto exchange rate regimes is best characterized as one between words and deeds (Levy-Yeyati and Sturzenegger, 2003). Official regimes are those that national authorities annually declare to the IMF as the regimes that they adopt. De facto regimes are those that are actually practiced by the authorities. Until recently, the distinction between official and de facto regimes has largely been ignored in the literature. However, the studies by Ghosh et al. (1997) and Calvo and Reinhart (2002) show that regime discrepancies are common in practice and these discrepancies may last for substantial periods of time. While Ghosh et al. (1997) find that frequent adjustments of the central parity can make an officially pegged exchange rate quite flexible, Calvo and Reinhart (2002) show that heavy foreign exchange interventions can make an officially floating exchange rate very stable.3 Why countries choose to practice exchange rate regimes different from their official announcements remains a puzzle in the literature. Calvo and Reinhart (2002) argue that low credibility of the monetary authority is the main reason for the fear of floating. By providing the economy with a transparent and easily verifiable nominal anchor for inflation expectations, stable exchange rates can help weak central banks improve the credibility of their commitment to price stability. Lahiri and Végh (2001) suggest that regime discrepancies result from a trade-off between the cost of foreign exchange market intervention and real output losses due to exchange rate volatility. Their analysis predicts that central banks allow the exchange rate to adjust to small shocks but intervene in the presence of large shocks to avoid excessive exchange rate volatility. Hausmann et al. (2000) explain regime discrepancies by the desire to avoid large exchange rate volatility, a desire which increases with a country's borrowing in foreign currency. At a closer look, however, these arguments contribute more to explaining the desirable degree of exchange rate flexibility and, therefore, the choice of official exchange rate regimes than to explaining discrepancies between official and de facto regimes. The present paper takes a positive approach to explaining exchange rate regime discrepancies in 25 transition economies during the 1990s.4 This is an interesting sample. Not withstanding their economic heterogeneity, these countries have many features in common: They all emerged from socialist regimes largely isolated from the world economy at the end of the 1980s; they all faced large macroeconomic imbalances and stabilization problems initially; they all became gradually integrated into international trade and financial markets during the period we consider. Yet, there is quite a variety of exchange rate regimes, both official and de facto, among these countries. von Hagen and Zhou (in press) show that the official regime choices in transition economies can be explained empirically by standard arguments from international macroeconomics. This paper focuses on the choices of de facto exchange rate regimes and their deviation from the official arrangements. We extend Levy-Yeyati and Sturzenegger's (2003) analysis to our sample of transition economies. We show that, similar to their findings in a sample of industrialized and developing countries, regime discrepancies are quite common among the countries we consider. A phenomenon closely related to regime discrepancies is the dissynchronization in regime adjustment. Official regimes exhibit less frequent but larger changes than de facto regimes. One rationale for this difference is that changing an officially announced exchange rate regime involves costs, such as the loss of credibility, while changing a de facto regime to which the authority makes no commitment is less costly. Therefore, official regimes are changed only when the benefit of doing so exceeds the related cost, while de facto regimes are frequently adjusted to “fine tune” the official regime. We develop a univariate probit model to explain the determination of the discrepancy between official and de facto exchange rate regimes. The model is derived from a bivariate one that allows correlation in the choices of official and de facto regimes. This model also incorporates the idea that de facto regimes may deviate from official ones in an attempt to correct “errors” in the choice of official regimes. Moreover, the model relates regime discrepancies to macroeconomic factors that may impact official and de facto regimes differently. The rest of the paper is organized as follows. Section 2 discusses the classification of de facto exchange rate regimes and the identification of discrepancies between de facto and official regimes. Section 3 derives a univariate probit model for regime discrepancies. Section 4 explains the empirical specification and discusses the estimation results. Conclusions are collected in Section 5.

Several recent studies have pointed out that official and de facto exchange rate regimes often differ in practice. While documenting these discrepancies, the existing studies do not attempt to provide an empirical answer to the obvious question, why countries choose different official and de facto regimes. This is the purpose of the current paper. We have presented an empirical study of the choices of official and de facto exchange rate regimes of 25 transition countries during the 1990s. We have shown, first, that regime discrepancies are frequent among these countries. Furthermore, official regimes are more persistent and change in less frequent but larger steps than de facto regimes. This is consistent with the notion that official regime changes carry a cost that exceeds the cost of changing the de facto regime, and that countries use the latter as a policy instrument to adjust their exchange rate policy to macroeconomic developments earlier and faster than they respond with their official regime. We develop a discrete choice model of regime discrepancies to explain deviations between the official and the de facto regimes. A first result of interest here is that regime discrepancies respond to the inappropriateness of the observed official regime. Specifically, if a country's official regime is too rigid as suggested by our discrete choice model, its de facto regime is likely to be more flexible than its official one. Conversely, if a country's official regime is too flexible, its de facto regime is likely to be more rigid than the official one. Thus, the widespread observation of fear of floating, where countries run de facto pegs although they declare themselves officially to have floating exchange rates, reflects cases where the underlying fundamentals contained in our model favor a more rigid official exchange rate regime than that actually declared. Such a pattern could rise, if the underlying fundamentals favour fixed exchange rates, or if countries preferring exchange rate pegs are reluctant to declare official pegs because of reputational constraints and the fear of being exposed to speculative attacks. We find that countries less diversified in the commodity structure of foreign trade, with more developed financial market, or having complete exchange-rate pass-through tend to exhibit fear of floating, i.e. with flexible official regimes combined with de facto stable exchange rates. In contrast, countries with larger economic sizes, more sufficient international reserves, or larger budget deficits are more likely to show fear of pegging. Their de facto exchange rate regimes are usually more flexible than the official exchange arrangements.