Central bank transparency: Causes, consequences and updates

Theoretical Inquiries in Law

Volumn 11, Issue 1, 2010, Pages

  Dincer, Nergiz ^c   Eichengreen, Barry ^a,b  

^a State Planning Organization   (Turkey)

^b UNIVERSITY OF CALIFORNIA   (United States)

^c NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 77649169753     PISSN: None     EISSN: 15653404     Source Type: Journal    
DOI: 10.2202/1565-3404.1237     Document Type: Article

References (53)

1. DONALD E. MOGGRIDGE, BRITISH MONETARY POLICY, 1924-1931, at 160 (1972).
2. See WILLIAM GREIDER, SECRETS OF THE TEMPLE: HOW THE FEDERAL RESERVE RUNS THE COUNTRY (1987).
3. Robert Barro & Roger Gordon, A Positive Theory of Monetary Policy in a Natural Rate Model, 91 J. POL. ECON. 589 (1983).
4. See, e.g., Petra M. Geraats, Why Adopt Transparency? The Publication of Central Bank Forecasts (July 2002) (unpublished manuscript, on file with the University of Cambridge);
5. Petra M. Geraats, Central Bank Transparency, 112 ECON. J., at F532 (2002).
6. See Richard Clarida, Jordi Gali & Mark Gertler, Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory, 115 Q.J. ECON. 147 (2000);
7. Larence J. Christiano & Christopher Gust, The Expectations Trap, ECON. PERSP. (Fed. Reserve Bank of Chi.), Q. II, 2000, at 21, available at http://www.chicagofed.org/publications/ economicperspectives/2000/2qep2.pdf.
8. Faust and Svensson consider a model in which the public attempts to infer whether or not the central bank is serious about limiting inflation from information on policy outcomes. Its inferences are imperfect because of unanticipated policy-implementation errors that the public is incompletely able to observe. Greater transparency about economic conditions thus enables agents to infer the central bank's preferences more accurately. In turn, this gives the monetary authority an incentive to build a reputation for valuing price stability. The private sector becomes more sensitive to unanticipated policy responses and actions, attenuating the incentive for the central bank to engage in them. The result is thus greater sensitivity of inflation expectations to policy actions, less benefit to the central bank for fueling inflation, and less inflationary bias. In this way, increased transparency about control errors improves social welfare. Greater transparency about the central bank's objectives may be more of a mixed blessing in this framework. In addition to reducing uncertainty about future inflation and output, which would, all other things equal, be welfare-improving, transparency about objectives removes the central bank's incentive to curtail inflation in order to signal its type, which could be welfare-reducing. See Jon Faust & Lars Svensson, Transparency and Credibility: Monetary Policy with Unobservable Goals, 42 INT'L ECON. REV. 369 (2001).
9. Moreover, while observers may never have been exactly ignorant about the influence of monetary policy over economic and financial conditions, it was less fully appreciated than today, when traders hold their collective breath on the days when the central bank announces its interest-rate decision. And where the extent of the franchise and of political contestability was limited, there was no more need for the central bank to be democratically accountable than any other branch of government.
10. Morris and Shin illustrate the point in a model in which individual welfare depends not just on the state of the world, but on the actions of other individuals (there exists a coordination externality). Starting from a position where both private and public information are imperfect, they show that greater precision of public information can lead individuals to attach inadequate weight to private information. In the absence of coordination motives, the precision attached to the public and private signals will be commensurate with their relative precision. When coordination motives are present, however, agents attach greater weight to the public signal, since they know this to be common information. But since the public signal is noisy, this weight on the public signal may be suboptimal from a social-welfare point of view; agents may be led to coordinate on the basis of an inefficient equilibrium. The more precise the private information, the more likely this adverse outcome is to ensue. See Stephen Morris & Hyun Shin, The Social Value of Public Information, 92 AM. ECON. REV. 1521 (2002);
11. Stephen Morris & Hyun Shin, Central Bank Transparency and the Signal Value of Prices, [2005] 2 BROOKINGS PAPERS ECON. AACTIVITY 1.
12. Thus, some critics of the European Central Bank, whose transnational status, embedded in an international treaty, makes it difficult for national politicians to sanction those responsible for its policy, question whether policy transparency provides adequate political accountability.
13. See Sirkka Hamalainen, The ECB's Monetary Policy - Accountability, Transparency and Communication (Sept. 14, 2001) (speech delivered to the Conference on Old Age, New Economy, and Central Banking), available at http://www.ecb.int/press/key/date/2001/html/sp010914.en.html for an early statement to this effect.
14. Andrew Clare & Roger Courtenay, Assessing the Impact of Macroeconomic News Announcements on Securities Prices Under Different Monetary Policy Regimes (Bank of Eng., Working Paper No. 125, 2001).
15. Carin A.B. van der Cruijsen, Sylvester C.W. Eijffinger & Lex H. Hoogduin, Optimal Central Bank Transparency (June 2008) (unpublished manuscript, on file with the De Nederlandsche Bank, Tilberg University and University of Amsterdam).
16. Charles Goodhart, Monetary Transmission Lags and the Formulation of the Policy Decision on Interest Rates, FED. RES. BANK ST. LOUIS REV., July/Aug. 2001, at 165.
17. Some central banks go partway toward indicating the prospective future path of interest rates; the Federal Reserve, for example, regularly signals its "bias" toward future interest-rate increases or reductions by issuing a "balance of risks" statement.
18. Frederic Mishkin, Can Central Bank Transparency Go Too Far? (Nat'l Bureau of Econ. Research, Working Paper No.10829, 2004).
19. See also Nergiz Dincer & Barry Eichengreen, Central Bank Transparency: Where, Why and with What Effects? (Nat'l Bureau of Econ. Research, Working Paper No. 13003, 2007).
20. Mishkin, supra note 15.
21. Sylvester Eijffinger & Petra Geraats, How Transparent Are Central Banks?, 22 EUR. J. POL. ECON. 1 (2006).
22. The overall index thus runs from 0 to 15. Adopting the same criteria used by these previous investigators has the advantage of facilitating comparisons across studies and frees us of any suspicion that we have constructed our measures so as to maximize or minimize the impact of transparency.
23. An earlier study by Bini-Smaghi and Gros also considered 15 aspects of central bank transparency, although they considered only four central banks. See Lorenzo Bini-Smaghi & Daniel Gros, Is the ECB Sufficiently Accountable and Transparent? (European Network of Econ. Policy Research Inst., Working Paper No. 7, 2001). De Haan, Amtenbrink and Waller developed a similar index for six countries, while De Haan and Amtembrink apply a similar approach to 15 countries. See Jakob De Haan, Fabian Amtenbrink & Sandra Waller, The Transparency and Credibility of the European Central Bank, 42 J. COMMON MKT. STUD. 775 (2004);
24. Jakob De Haan & Fabian Amtenbrink, A Non-Transparent European Central Bank: Who Is to Blame? (May 2002) (unpublished manuscript, presented at the Bank of England to the conference on Monetary Policy Transparency). Siklos expands coverage to twenty central banks, all from advanced industrial countries.
25. See PIERRE SIKLOS, THE CHANGING FACE OF CENTRAL BANKING: EVOLUTIONARY TRENDS SINCE WORLD WAR II (2002). The most comprehensive previous study in terms of country coverage is Fry et al., although they construct indices only for 1998.
26. See Maxwell Fry et al., Key Issues in the Choice of a Monetary Policy Framework, in MONETARY POLICY FRAMEWORKS IN A GLOBAL CONTEXT 1 (Lavan Mahadeva & Gabriel Sterne eds., 2000).
27. Adding this time dimension was particularly challenging, since many central bank websites describe current practice, but not that of prior years. For that we had to rely mainly on published documents. We were able to access a relatively complete run of these on the basis of holdings in the University of California and Joint IMF-World Bank libraries. We are grateful to the staff of the Joint Bank-Fund library for granting us access to their collection.
28. This is the vast majority of central banks in the world (recall that there are more countries than there are central banks, given the existence of monetary unions, countries that have unilaterally adopted the currencies of other countries, etc.). A subsequent study by Crowe and Meade constructs measures of transparency very similar to our own, but only since 2000 and for a much smaller (37 country) sample (their purpose being to compare measures of central bank transparency and independence). See Christopher Crowe & Ellen Meade, Central Bank Independence and Transparency: Evolution and Effectiveness (Int'l Monetary Fund, Working Paper No. 08/119, 2008).
29. Among the omissions are Bolivia, Ecuador, Chad, Iran, and Afghanistan. We are aware that this creates a form of sampling bias: we tend to oversample more transparent central banks. Although econometric corrections for this bias (involving strong assumptions) do exist, we have not implemented these yet. Our defense is that the number of consequential omissions is relatively slight.
30. Up from 47 in 1998.
31. -2006 being the most recent year for which all the ancillary variables are currently available.
32. Carmen Reinhart, Kenneth Rogoff & Miguel Savastano, Debt Intolerance, [2003] 1 BROOKINGS PAPERS ECON. ACTIVITY 1;
33. Barry Eichengreen & Raul Razo-Garcia, The International Monetary System in the Last and Next 20 Years, 21 ECON. POL'Y 393 (2006).
34. Daniel Kaufmann, Aart Kraay & Massimo Mastruzzi, Governance Matters VI: Governance Indicators for 1996-2006 (World Bank Policy Research, Working Paper No. 4280, 2007).
35. Where we use variables from the Polity database, we are forced to end the analysis in 2004.
36. Readers may be concerned that the exchange rate regime is endogenous - that countries with experience with monetary policy transparency may be better able to operate regimes of greater flexibility. In the event, dropping the exchange-rate regime leaves the other results unchanged.
37. The other variables do not approach statistical significance at conventional confidence levels. For what they are worth, the point estimates suggest that central banks of countries with better developed financial markets tend to be more transparent. Similarly, central banks of countries with a history of inflation tend to be more transparent, presumably as part of a credibility-building strategy. This is not something that might have been anticipated from the contrast between transparency in advanced and developing countries.
38. When we include multiple political variables, it is voice and accountability and government efficiency that are most often significant at standard confidence levels.
39. Note that the "polity" variable in the final column is the difference between "democracy" and "autocracy."
40. Political transparency also appears to decline with financial depth, which is not a pattern for which we have an immediate explanation. Interestingly, other components such as economic and procedural transparency are positively associated with financial depth, as can be seen from the corresponding tables.
41. The standard Hausman and Breusch-Pagan tests reject random effects and simple pooling in favor of fixed effects (the Hausman test statistic is reported at the foot of the tables). See also the further discussion below.
42. Mishkin, supra note 15.
43. But not between the level of inflation and transparency. See Maria Demertzis & Andrew Hughes Hallett, Central Bank Transparency in Theory and Practice, (Jan. 2003) (unpublished manuscript, on file with the Netherlands Bank and Vanderbilt University).
44. Gregory Mankiw, Ricardo Reis & Justin Wolfers, Disagreement About Inflation Expectations, 2003 NAT 'L BUREAU ECON. RES. MACROECONOMICSANN. 209 (2004);
45. Andrew Levin, Fabio Natalucci & Jeremy Piger, Explicit Inflation Objectives and Macroeconomic Outcomes (European Central Bank, Working Paper No. 383, 2004).
46. Refet Gurkaynak, Andrew Levin & Eric Swanson, Inflation Targeting and the Anchoring of Long-Run Inflation Expectations: International Evidence from Daily Bond Yield Data (June 2005) (unpublished manuscript, prepared for the Riksbank Workshop on Inflation Targeting, Stockholm) (on file with authors).
47. Results using alternative instrument lists are discussed below and are available from the authors on request.
48. Though the confidence level in question depends on which specific controls are included or excluded.
49. The squared term, like the level, is constructed from the fitted value of transparency derived from the first-stage regression.
50. If the coefficients are taken literally, they suggest that the benefits in terms of reducing inflation variability dissipate and inflation variability begins to rise with transparency when the index for the latter exceeds five, a suspiciously low threshold. Thus we do not want to push these particular point estimates too far.
51. Van der Cruijsen et al., supra note 12.
52. The versions of the equation run with instruments are again estimated by GMM. The two-stage least squares version simply applying the instruments to transparency yields the same results.
53. The coefficient on the fifth component enters with a t-statistic of 4.2. None of the other coefficients have t's in excess of one.