رشد بهره وری، شفافیت، و سیاست های پولی

This study examines whether central bank transparency about views of future productivity growth contributes to stabilizing macroeconomic fluctuations. In a standard New Keynesian model, the central bank and private agents make their subjective estimates on the persistence of productivity growth. In this situation, if private agents believe that the central bank's projections include forecast errors on future productivity growth, these beliefs can destabilize private agents' own expectations because the central bank's forecast errors may lead to policy mistakes in the future. Consequently, central bank transparency does not necessarily stabilize the variations of the output gap and inflation rate. The central bank should respond strongly to the inflation rate, if the impact of transparency is uncertain.

In recent years, the implications of central bank transparency have been actively investigated in monetary economics.1 Among the various aspects of central bank transparency, this study focuses on “economic transparency” in the terminology of Geraats (2002). Economic transparency concerns the economic information that is used for monetary policy, including economic data, policy models, and the central bank's forecasts. Economic transparency is distinct from other kinds of transparency (such as political, procedural, policy, and operational transparency) in that it does not deal with the behavior of the central bank itself. Rather, it concerns the central bank's views of economic conditions, which are determined mainly by the activities of private agents. In the case of economic transparency, it is arguable whether the central bank should seek to be perfectly transparent, because the central bank usually faces considerable uncertainty as to economic conditions or economic structures. If we take account of this kind of uncertainty, it is not a straightforward task to evaluate the value of central bank transparency, because the information provided by the central bank to private agents might be inaccurate. The problem of uncertainty becomes particularly serious with respect to the trend growth of aggregate productivity. Trend productivity growth is a key variable for monetary policymaking, because it is the crucial determinant of potential GDP and the equilibrium level of the real interest rate. However, it is widely recognized as difficult to obtain an accurate estimate of the trend growth of aggregate productivity, especially in real time. An important issue for monetary policymakers is whether the central bank should be transparent even when they are uncertain about future productivity growth. As noted by Bernanke, the issue is complicated because not only the central bank but also private agents face uncertainty regarding the persistence of productivity growth. In such a case, the impact of central bank transparency is likely to depend on private agents' forecast of future productivity growth. Because central bank transparency mainly influences private agents' perception about the central bank's forecast, the impact of transparency should depend on private agents' conjecture about the central bank's forecast. We examine whether central bank transparency about views of future productivity growth contributes to stabilizing macroeconomic fluctuations. In a simple version of a New Keynesian model, which is very close to the models of Galí et al. (2003) and Ireland (2004), we assume that the central bank and private agents cannot fully identify the transitory and persistent components of productivity growth and that they solve a filtering problem in order to estimate the persistence of productivity growth. This kind of filtering process has already been introduced in some previous studies, such as Tambalotti (2003), Edge et al. (2007), and Gilchrist and Saito (2008). These studies have shown that private agents' gradual recognition of the persistence of productivity growth can replicate the persistent movements of major macroeconomic variables, which are usually found in vector autoregression (VAR) analysis. This paper focuses on the heterogeneity of the forecasting mechanisms used by the central bank and private agents. Fig. 1 presents forecasts of real output growth made by the Federal Reserve Board (FRB) and economists in the private sector. This figure shows that the difference of the forecasts between the FRB and the private sector reached nearly or more than 1% in some years, such as 1982, 1991, 1998, 2000, and 2004. The standard deviation of the forecasts between 1982 and 2004 is 0.682 for private sector and 0.854 for the FRB. The first order autocorrelation coefficient is 0.781 for private sector and 0.587 for the FRB. Thus, the data suggest that the forecasts are fairly heterogeneous between the FRB and private agents, and that the FRB changed the forecast more sharply (or largely) than the average of private forecasts at least during this period.2 Although output growth does not directly correspond to productivity growth, it is fair to judge that at least some portion of the different forecasts on output growth comes from the heterogeneity in the views on future productivity growth.We assume that the heterogeneous forecasts arise because the central bank and private agents use different gain parameters for the filtering problem.3 There are two potential sources for which the central bank and private agents use different filtering gains. First, as some empirical studies (Stock and Watson, 1998 and Roberts, 2001) show, there is considerable uncertainty about the variances of transitory and persistent productivity shocks. Under this situation, it is natural for the two agents to have different priors on the signal to noise ratio of the productivity shock. Especially, the recent analysis of Justiniano and Primiceri (2008) shows that there have been large structural changes in shock variances in the U.S. economy. Given this feature, it is plausible that the two agents assess the possibility of structural changes differently. Second, as the recent literature of “Rational Inattention” (Sims, 2003) suggests, there should be some information processing costs for the central bank and private agents. These costs can be asymmetric for these two agents because the possibilities and the necessities of accessing to the aggregate productivity data could be generally different between the central bank and private agents.4 If we take account of these realistic factors, the filtering gains can be heterogeneous even if they use the same information set concerning the historical movements of productivity growth. We define the central bank as transparent (or the central bank as adopting a transparent regime) if the central bank announces its forecast of future productivity growth, and we also define the central bank as opaque (or the central bank as adopting an opaque regime) if the central bank does not announce the forecast. Private agents must conjecture the central bank's forecast of future productivity growth, since the forecast is closely linked with the central bank's estimate of the efficient level of the real interest rate (efficient interest rate), which influences the future interest rate. If the central bank is transparent, private agents accurately recognize the central bank's estimate of the efficient interest rate. However, if the central bank is opaque, private agents have misperceptions about the central bank's estimate of the efficient interest rate. Our results show that central bank transparency about views of future productivity growth does not necessarily contribute to stabilizing macroeconomic fluctuations. It can potentially destabilize the economy, depending on the combinations of (i) the forecasts made by the central bank and private agents and (ii) private agents' conjecture about the central bank's forecast. We find that the central bank transparency can have a destabilizing effect especially when private agents conjecture that the central bank's forecast is not largely different from their own forecast. This result can be intuitively explained as follows. Suppose that in an opaque regime, private agents believe that the central bank has the same forecast and the same model as theirs. Then private agents think that the central bank's estimate of the efficient interest rate has to be accurate, because private agents do not harbor any doubts about their estimate of the efficient interest rate. In this situation, if the central bank reveals that the bank's forecast is actually very different from private agents' forecast, private agents realize that the central bank will not adjust interest rate enough (or too much) in future due to the mismeasurement of the efficient interest rate. Private agents' expectations for future policy mistakes have a destabilizing impact on their expectations for the output gap and inflation rate. This mechanism is the key for our main findings. Our study is distinct from previous studies of central bank transparency (such as Amato and Shin, 2003; Morris and Shin, 2005; Walsh, 2007), in that the mechanism through which central bank transparency can destabilize the economy differs fundamentally from those of the previous studies. In the previous studies, central bank transparency can destabilize the economy mainly because private agents might overreact to the information provided by the central bank, in an environment where the central bank or private agents have private information about current economic conditions.5 In our setup, central bank transparency can magnify the volatilities of the output gap and inflation rate because it can destabilize private agents' forward-looking expectations by exposing the central bank's mismeasurement concerning the efficient interest rate, which lead to expected policy mistakes in future.6 The rest of this paper is organized as follows. In Section 2, we present our model. In Section 3, we explain the influence of central bank transparency on economic dynamics. In Section 4, we investigate whether central bank transparency contributes to stabilizing macroeconomic fluctuations and what kind of monetary policy actions is desirable if the impact of transparency is uncertain. In Section 5, we present concluding remarks.

In this study, we have investigated whether central bank transparency about views of future productivity growth contributes to stabilizing macroeconomic fluctuations. To this end, we have used a standard New Keynesian framework in which the central bank and private agents make their subjective estimates on the persistence of productivity growth. In this situation, if private agents believe that the central bank's projections include forecast errors on future productivity growth, these beliefs can destabilize private agents' own expectations because the central bank's forecast errors may lead to policy mistakes in the future. Consequently, central bank transparency does not necessarily stabilize the variations of the output gap and inflation rate. We find that transparency has a destabilizing effect especially when private agents conjecture that the central bank's forecast is not largely different from their own forecast. Because this situation is not entirely unrealistic, the issue analyzed in this study could be viewed as practically relevant. In addition, it is possible that the central bank faces the uncertainty about the desirability of transparency because the central bank hardly observes private agents' conjecture about the central bank's forecast. Our study suggests that in the face of this uncertainty, it is sensible for the central bank to respond strongly to the variations in the inflation rate.