Economics Letters 77 (2002) 195198www.elsevier.com/ locate /econbase

H ow much should central banks talk?A new argument

a,b ,*Hans Peter Gruner

aCEPR, London, UKbUniversity of Mannheim, IZA, Bonn, Germany

Received 17 December 2001; accepted 18 March 2002

Abstract

More openness in central bank decision procedures may increase inflation uncertainty for players on financialmarkets when it leads to less wage discipline and higher average inflation. 2002 Elsevier Science B.V. All rights reserved.

Keywords: Central bank transparency; Communication; Inflation uncertainty

JEL classification: E58

1 . Introduction

1The openness of central bank decision making has recently received new attention in the literature .It has been argued that more openness reduces uncertainty for players on financial markets and makesfuture decisions more transparent (cf. Blinder et al., 2001). In this paper I argue that the opposite maybe the case. The argument is based on a model that studies the interaction of major macroeconomicplayers with the central bank.

In the paper I make a distinction between uncertainty about the central banks objectives andinflation uncertainty. This distinction turns out to be crucial. In the spirit of Cukierman and Meltzer(1986), I assume that the public is uninformed about the weight which the central bank puts on itsvarious objectives. The disclosure of information affects the degree of uncertainty about central bankobjectives. However, actual inflation uncertainty is affected by these objectives and by the actions ofall macroeconomic players. In so far as uncertainty about central bank objectives affect trade unions

*Tel.: 149-521-181-1886; fax: 149-621-181-1884. E-mail addresses: hgruener@rumms.uni-mannheim.de (H.P. Gruner), http: / /www.vwl.uni-mannheim.de (H.P. Gruner).

1Recent contributions include Blinder et al. (2001), Cukierman (2001) and Gersbach and Hahn (2001a,b).

0165-1765/02/$ see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0165-1765( 02 )00125-8

196 H.P. Gruner / Economics Letters 77 (2002) 195198

behavior, it may also have a positive impact both on price stability and on inflation uncertainty. Moreuncertainty about future monetary policy objectives leads to more wage discipline, which in turnlowers average inflation (Sorensen, 1991, derives the same result from a very similar model).Moreover, I show that the variance of inflation may be reduced as well.

1 .1. The model

The model is a two stage game in which a monopoly union interacts with a central bank. Previousmodels of this sort can be found in Cukierman and Lippi (1999), Gruener and Hefeker (1999), andothers.

Two variables are determined in this game: The (log of the) nominal wage w and inflation p. Thenominal wage is fixed by the monopoly union before the central bank fixes p. Both variablesdetermine the (log) real wage W5w2p and unemployment u5 a ?W. Without loss of generality Ifix a5 1.

The objectives of both players are the following. The central bank cares about inflation andemployment. Its utility function is

2 2C(p, u)5 2 Ip 2 u , (1)where I is a measure of the central banks inflation aversion. The union cares about the real wage andemployment. Its objective is to maximize

A 2]U(w, p)5W2 u . (2)2The parameter A measures the unions aversion to unemployment. The reaction function of the centralbank can be derived as

p 5 b ?w, (3)where b5 1/11 I. The trade union is not perfectly informed about the central banks preferences

when it fixes w. It knows the mean of the inflation reaction E(b)5 b and the variance2 2

s 5E[(b2 b ) ]. (4)b2Measures that make the decision process more transparent are assumed to result in a reduction of s .b

1 .2. Equilibrium

Taking into account the central banks expected reaction the union maximizes the followingfunction at stage 1:

A 2F ] GE w2p 2 u (5)2A 2F ] G5E (12 b)w2 ((12 b)w) (6)2

H.P. Gruner / Economics Letters 77 (2002) 195198 197

A 2 2F ] G5E (12 b)w2 (12 2b1 b )w . (7)2The solution of this problem is

E 12 bs d]]]]]]w5 (8)2f s dgE A 12 2b1 b

or

12 b]]]]]]]w5 (9)2 2

A 12 2b 1 b 1ss db1 .3. Results

An immediate consequence is Sorensens (1991) result:

Proposition 1. Uncertainty about central bank preferences reduce wages, average inflation andunemployment.

2Proof. Wages decline with s as can be seen from (9). Moreover, according to (3) average inflationband average unemployment decline as well. Q.E.D.

Next we analyze whether more uncertainty about the central banks preferences necessarily createsmore inflation uncertainty. It is useful to define

2 2 2 2 2

s [E[(p 2p ) ]5E[(bw2 bw) ]5w ? s (10)p b

We have2ds dwp 2 2]] ]]5w 1 2w s (11)2 2 bds dsb b

12 b2 2]]]]]]]5w 2 2ws (12)b 2 2 2 A(12 2b 1 b 1s )b

12 2 2]]]]]]5w 2 2w s . (13)b 2 2 (12 2b 1 b 1s )b

This derivative is positive iff:2 2

12 2b 1 b .s (14)b2 2

(12 b ) .s . (15)b

Hence we have:

198 H.P. Gruner / Economics Letters 77 (2002) 195198

Proposition 2. Inflation uncertainty increases with uncertainty about central bank preferences whenthe latter is low. It decreases when uncertainty about b is high.

Proof. see above.

2 . Discussion

The present paper adds another theoretical argument in favor of limited central bank transparency tothe contributions by Sorensen (1991), Cukierman (2001), Cukierman and Meltzer (1986) andGersbach and Hahn (2001a). According to Sorenson, uncertainty about central bankers preferencesmay lead to more wage discipline. Wage setters act more carefully when they know less about the wayin which the central bank might react to their wage claims. This reduces equilibrium wage claims,inflation and unemployment on average. Governments that are concerned about inflation andunemployment should therefore be careful about introducing too much transparency.

I have shown that, even sticking to the narrow objective of low inflation uncertainty, it may notalways be optimal to improve the publics information about the central banks future objectives. Thisis the case when there is an upper bound on the degree of informedness of the public. Consider e.g.

2 2the case where new rules for information disclosure can reduce uncertainty to a level s $s . 0.b b,min2

If this minimum level exceeds s12 bd then any reduction of uncertainty about future moves leads toan increase of inflation uncertainty.

A cknowledgements

The paper was written while I was visiting the economics department at London Business School. Ithank this institution for its hospitality.

R eferences

Blinder, A.S, Goodhart, C.A., Hildebrand, P.M., Lipton, D., Wyplosz, C., 2001. How Do Central Banks Talk? GenevaReports on the World Economy 3.

Cukierman, A., 2001. Are Contemporary Central Banks Transparent about Economic Models and Objectives and WhatDifference Does it Make?, mimeo.

Cukierman, A., Lippi, F., 1999. Central Bank Independence, Centralization of Wage Bargaining, Inflation and Unemploy-mentTheory and Evidence European Economic Review.

Cukierman, A., Meltzer, A.H., 1986. A theory of ambiguity. Credibility and inflation under discretion and asymmetricinformation. Econometrica 54 (5), 10991128.

Gersbach, H., Hahn, V., 2001a. Should the individual voting records of central bankers be published?, Deutsche Bundesbank,discussion paper 02.01.

Gersbach, H., Hahn, V., 2001b. Voting transparency and conflicting interests in central bank councils Deutsche Bundesbank,discussion paper 03.01.

Gruener, H.P., Hefeker, C., 1999. How will EMU affect inflation and unemployment in Europe? Scandinavian Journal ofEconomics 101, 3347.

Sorensen, J.R., 1991. Political uncertainty and macroeconomic performance. Economics Letters, 37.

How much should central banks talk?A new argumentIntroductionThe modelEquilibriumResults

DiscussionAcknowledgementsReferences