نرخ تورم، باز بودن و رژیم نرخ ارز : تلاش برای تعهد کوتاه مدت

This paper further tests Romer's [Romer, D., 1993. Openness and inflation: theory and evidence. Quarterly Journal of Economics 58, 869–903] extension of Kydland and Prescott's [Kydland, F., Prescott, E., 1977. Rules rather than discretion: the inconsistency of optimal plans. Journal of Political Economy 85, 473–491] predictions for dynamic inconsistency problems in open economies. In a panel data set of developed and developing countries from 1973 to 1998, I find that openness does not play a role in restricting inflation in the short run. On the other hand, a fixed exchange-rate regime plays a significant role. The results are robust to controlling for other variables that determine inflation, performing sensitivity analysis, and using a de facto exchange-rate regime classification.

The 1990s probably will be remembered, among other things, as the decade in which average inflation came under control around the world. According to the IMF, average inflation in industrialized economies between 1982 and 1991 was 4.9%; at the end of 1999, it was 0.8% as measured by the GDP deflator (Table 1a). More remarkably, average inflation in developing countries went from 45.1% between 1982 and 1991 to 6.9% in 1999. The 1990s also will be remembered as the “globalization” decade. Trade of goods as a percentage of PPP GDP went from 21.2% in 1988 to 28.3% in 1998 (Table 1b). This increase is even more striking when analyzed as a percentage of goods production, which increased from 71.9% in 1988 to 92.1% in 1998. Although concerns remain about lingering tariffs, nontariff barriers, and other protectionist practices, it is hard to deny that the global economy has become more integrated.The obvious question is then: Are these events related? Romer's (1993) work shows that inflation and openness are negatively and significantly correlated. Using Kydland and Prescott (1977) and Barro and Gordon (1983) type models, Romer argues that the negative openness–inflation relationship arises from the dynamic inconsistency of discretionary monetary policy. Romer's empirical work, using cross-country averages spanning more than a decade (beginning in 1973), tests the long-run commitment effect of openness on restricting the usefulness of discretionary monetary policy. The question remains, however, whether openness or other mechanisms bind in a short-term horizon. In this paper, I test whether openness serves as a commitment mechanism for restraining inflation in the short run. In addition, I consider the role of other variables that affect the short-term dynamics that drives inflation and the time inconsistency problem, in particular the exchange-rate regime. One of the central issues underlying the time inconsistency problem is a government's inability to make binding commitments to future policies. Due to the sequential nature of policy-making, it is generally optimal ex-post for governments to deviate from earlier, ex-ante optimal plans or preannounced policy rules. Private agents, aware that policy-makers have incentives to deviate in the future, base their current actions on the expectation that deviations will occur. The final outcome is a suboptimal high inflation rate. Without binding commitments that impose high costs for reversing policy, a plan with desirable properties might encounter serious credibility problems.1 These considerations give rise to important empirical puzzles. Although we do observe high periods of inflation, we do not see them in all countries at all times. What does this say about the ability to commit? Does it change, or correlate to other policy instruments? What mechanisms enable countries to relax incentive constraints and mitigate the credibility problem? Judging from observed outcomes, the efficiency of such mechanisms varies across societies and time. Unfortunately, there is little empirical evidence that the time inconsistency problem identified by Kydland and Prescott (1977) and Barro and Gordon (1983) is important to actual inflation. Romer's (1993) worthy contribution is to test the prediction that the absence of precommitment in monetary policy, given by the degree of openness, leads to inefficiently high inflation. The theoretical reasoning follows from Rogoff's (1985) model, which shows that more open economies gain less from surprise inflation. Surprise monetary expansions cause the real exchange rate to depreciate, leading to a negative terms-of-trade effect. The more open the economy, the more the real exchange depreciates, thus reducing incentives to undertake expansion. Using cross-country data, Romer finds a robust negative link between openness and inflation.2 I further explore the robustness of the cross-section results to adding more than a decade of observations and controlling for other variables that might determine inflation. The cross-section empirical work, however, tests the long-run commitment effect of openness on restricting the usefulness of discretionary monetary policy. A more relevant exercise, given the very nature of the time inconsistency problem, is to test the effectiveness of openness on restricting discretionary monetary policy over shorter-time horizons. Accordingly, and in contrast to previous work, I also use a panel data set of developing and developed countries from 1973 to 1998. I take advantage of the time dimension of the data to analyze whether openness serves as a commitment mechanism for restraining inflation in the short run. It could be argued that the correlation in the cross-section analysis might be driven by time-invariant omitted variables that often are difficult to measure. As a result, one could find evidence of a negative effect of openness on inflation where no such restraint on inflationary policy takes place. In effect, once time and country dummies are considered to capture this difference, I do not find a negative relationship between openness and inflation. In the short run, there is no robust evidence that openness has restrained inflation. I further explore which other variables might affect the short-term dynamics that drive inflation. In particular, I analyze the relationship between inflation, openness, and the exchange-rate regime as a commitment device. Romer (1993) argues that the choice of exchange-rate regime is not an important determinant of inflation.3 But, as Frankel (1999) observes, fixing the exchange rate has the advantage of providing an observable commitment to monetary policy. Calvo and Vegh's (1999) work on stabilization programs shows that exchange-rate pegs, albeit usually ending in balance of payment crises, enable countries to reduce inflation temporarily. The other often mentioned advantage of fixing the exchange rate is that it reduces transaction costs and exchange-rate risks that might discourage trade (Frankel, 1999). This argument implies that fixed exchange-rate regimes could be associated with higher levels of trade. Excluding a fixed exchange-rate variable from an analysis that considers the relationship between inflation and openness can thus bias the results. I find a significant negative relationship between a fixed exchange-rate regime and inflation that is robust to the inclusion of other control variables used in the literature. Ghosh et al. (1997) use a similar approach when studying the relation between inflation and different exchange-rate regimes across a panel of countries from 1960 to 1990. A main difference from previous work is that this paper focuses on the role time-invariant omitted variables play in driving the results for which I control for country and year fixed effects. In addition, I use a new exchange-rate regime classification constructed by Rogoff and Reinhart (2004), which considers parallel exchange-rate data when assessing whether a country has, de facto, maintained a pegged or a flexible regime. The main results are robust to using this classification. The rest of the paper is organized as follows. The data are defined in Section 2. Section 3 examines the cross-section evidence on openness and inflation. Section 4 discusses the panel empirical approach and the main results. Section 5 concludes.

This paper demonstrates empirically the non-significant role openness seems to play as a commitment mechanism in the short term. Pegged exchange-rate regimes, however, have been associated with significantly better inflation performance. These results hold after controlling for country and time-fixed effects and using Rogoff and Reinhart's (2004) de facto exchange-rate regime classification. The negative and strong relationship between inflation and a fixed exchange-rate regime is consistent with the classical literature, which argues that fixed exchange-rate regimes impose “discipline” on individual countries, reflecting the greater observability, accountability, and transparency of the exchange rate over openness. The relationship is consistent as well with the literature on dynamic inconsistency problems, which suggest that the absence of precommitment in monetary policy leads to inefficiently high levels of inflation. A government that decides to pursue an inflationary policy risks undermining the viability of the fixed exchange-rate regime, which can constrain its intentions. There being a stronger link between the exchange rate and inflation than between inflation and openness in the short term, inflation might be more effectively restrained through economic cooperation and by integrating the design of macroeconomic policies consistent with maintaining fixed exchange rates. The message of the paper is not that a peg inconsistent with fiscal and monetary policies can achieve low, long-term, sustainable inflation. Rather, this paper argues that, in the short run, a fixed exchange rate has served as a commitment mechanism and thereby limited inflation