کارکرد بانک مرکزی مقاوم تحت چانه زنی دستمزد: آیا شفافیت سیاست های پولی مفید است؟

We study the role of transparency in an environment of robust monetary policy under wage bargaining. The standard view from the game-theoretical literature is that, with unionised labour markets, monetary policy transparency is unambiguously “bad” (it induces increases in wage and price inflation, unemployment and may lead to higher inflation uncertainty). The empirical literature is instead ambiguous about the macroeconomic effects of transparency. By recasting the earlier theory into a robust monetary policy environment, and focusing transparency on the uncertainty about the preference for price stability, we show that the macroeconomic effects of transparency are more favourable than normally found. The impact on nominal wages, inflation and real variables (real wages and unemployment) is not parameter-free but depends on the public's informedness about this coefficient. The impact on real variables is found to disappear in case unions do not internalise the effect of wage decisions on the economy (i.e. in the case of atomistic unions). Finally, we find that the effect of transparency on inflation uncertainty is more complex than in the standard approach. We show that transparency may have the beneficial effect of reducing inflation variability not only when monetary uncertainty is low (as previously reported), but also when monetary uncertainty exceeds an upper threshold.

Central banks make their decisions in the presence of considerable uncertainty about the structure of the economy and the impact of policy on private behaviour. Simple policy rules may give rise to the risk of disappointing outcomes when the reference model being employed fails to represent the economy well. This is the view recently conveyed by an increasing body of monetary policy research that delves into the formulation of robust control problems. According to this approach, central banks could be characterised as pursuing policies that trade ex post performance for greater certainty in the aim to avoid particularly poor economic outcomes. The objective of this paper is to examine the effect of transparency on macroeconomic developments when a robust central bank interacts with wage setters. To do so, we set up a simple robust control model, revisiting the formulation of robust policies proposed by Giordani and Söderlind, 2004 and Hansen and Sargent, 2007.1 The standard theory about monetary policy transparency under wage bargaining was developed by Grüner and Hefeker, 1999 and Grüner, 2002 in a context of “conventional” (i.e. not robust) central banking. The message from this game-theoretical literature is that, with a highly centralised labour market, monetary policy transparency leads to rises in wage and price inflation, unemployment and inflation uncertainty.2 In short, transparency is necessarily “bad”.3 In this literature, transparency is given by an inverse measure of the uncertainty about the monetary policy rule which is assumed to capture uncertainty about the central bank preference for price stability (relative to stabilising activity); that is, the type of transparency involved refers to a so-called “political” concept of transparency. Here we recast the existing theory into a robust monetary policy environment, reassessing the effects of (political) transparency on the level of nominal and real variables as well as inflation uncertainty. Moreover, instead of assuming a connection between uncertainty about the policy rule and uncertainty about the central bank preference for price stability, we directly use the latter type of uncertainty to characterise monetary policy. We show that the macroeconomic impact of transparency is not parameter-free but depends on the (initial) degrees of transparency and conservativeness of the central bank. The unambiguously negative assessment of transparency advanced by the standard approach is inconsistent with the empirical literature. In fact, the latter provides conflicting findings about the macroeconomic effects of transparency, in particular uncovering in some cases favourable effects from disclosing more information to the public. With regard to nominal wages, Grüner et al. (2009) find that a higher value of their index of monetary policy uncertainty is associated with lower wage inflation in France, Germany and Japan, but plays an insignificant role in the UK and the US. Turning to price inflation, its level has been found to be lower with higher (political) transparency about the target (Kuttner and Posen, 1999 and Fatás et al., 2007). Chortareas et al. (2002a) look into the effects of transparency about policy decisions and forecasts, reporting that higher transparency reduces average inflation. Chortareas et al. (2002b) show evidence that more forecast detail provided by central banks leads to lower average inflation in countries where the domestic nominal anchor is based on an inflation or money target but not when an exchange rate target is instead in place. In contrast, Demertzis and Hughes Hallett (2007) find that, among nine major central banks in the 1990s, transparency is not correlated with average inflation. The authors also find that transparency is negatively correlated with inflation variability, but this result is not so relevant here because it appears to be driven not by the “political” but by the “economic” and “operational” components of the transparency index. Using pooled regressions for 100 countries over the period 1998–2006, Dincer and Eichengreen (2009) show that higher monetary policy transparency (instrumented using political determinants) is associated with less inflation variability; however, this result depends on the controls used, with only half of the results being statistically significant (see their Table 7). Finally, the literature on the macroeconomic effects of inflation targeting (IT) could also be seen as indirectly providing information about the role of transparency, given that this monetary regime aims at increasing central bank transparency and accountability. For countries that target inflation explicitly, the international evidence on this issue is mixed. Experts often conclude that countries that adopt IT manage to lower the level and variability of inflation (see e.g. Corbo et al., 2001 and Neumann and von Hagen, 2002). However, Ball and Sheridan (2005) question whether this is the case among advanced countries. In their study of 20 OECD countries (7 of which are inflation targeters), they show that, after controlling for the effect of the regression to the mean, IT countries fail to display a better performance than non-inflation targeters in terms of inflation. The rest of the paper proceeds as follows. For the case where wage setters are organised as a monopoly union, Section 2 sets up the model, while Section 3 provides the results to be discussed in Section 4. Section 5 extends the results to the case of atomistic unions. Section 6 concludes.

This paper extends the literature by applying the robust control approach to a model of interaction between wage setters and the central bank. This is meant to introduce more realism into the analysis, in light of real-world central bankers appearing to be confronted with considerable uncertainty as to which economic model will prevail. More precisely, our setup postulates that the central bank is unsure about unemployment developments, which are potentially affected by model misspecification. In addition, we allow for a second source of uncertainty whereby wage setters do not exactly know the central bank's preference for price stability. This influences the private sector's inflation expectations and therefore wages, actual inflation, unemployment and inflation uncertainty. Another way in which we deviate from the standard literature is by focusing transparency on the uncertainty about the preference for price stability, rather than—as normally done—on some policy rule coefficient. Our theoretical results challenge the “conventional wisdom” that, with centralised labour markets, transparency will always entail adverse macroeconomic consequences, as given by high levels of wage and price inflation, unemployment. We also qualify the standard approach's result that transparency may lead to higher inflation uncertainty. Overall, thus, by recasting the earlier theory into a robust monetary policy environment, and focusing transparency on the uncertainty about the preference for price stability, we are able to show that the macroeconomic effects of transparency are more favourable than normally found. This is more inline with the trend towards heightened dissemination of information by central banks worldwide (see e.g. Dincer and Eichengreen, 2009). More concretely, concerning nominal wages, inflation and real variables (real wages and unemployment), we find that the impact of transparency is not parameter-free. The impact on nominal wages depends on the expected value of the preference for price stability, while the impact of transparency on inflation and real variables depends on the public's informedness about that coefficient.16 In particular, our finding concerning the effect on nominal wages stands in contrast with the standard view that opacity necessarily helps discipline wage demands in a centralised labour market environment. We find that this is true only if initially the public's perception of the central bank preference for price stability is low enough. Our study indicates that opacity only works effectively for a “liberal” central bank, i.e. as a substitute for price stability concerns. In contrast, transparency when accompanied by conservativeness is able to moderate wage inflation. All in all, our findings that macroeconomic effects of transparency are in general parameter-dependent align better than the standard approach with the empirical literature, which suggests that such effects are ambiguous. Finally, the effect of transparency on inflation uncertainty is here found to be more complex than in the standard approach. In particular, we show that transparency may have the beneficial effect of reducing inflation variability not only when monetary uncertainty is low (as previously reported), but also when monetary uncertainty exceeds an upper threshold. We uncover this general result for both the cases of monopoly union and atomistic unions. Our finding implies that, even if there is a lower bound to how much transparency—and thus inflation predictability—can be achieved (as discussed by Grüner, 2002), enhancing central bank communication may still play a favourable role for cases where policy uncertainty is high enough. In any case, it must be borne in mind that the understanding of robust monetary policy, including in a context of imperfect transparency, is the subject of extensive ongoing investigation, so the problems addressed in the present paper will likely have to be revisited in further research.