“Evolution in Central Bank Governance Around the World”

Summary of Remarks by
Ellen Meade
Associate Professor, American University
March 29, 2007

The NEC was pleased to welcome Ellen Meade, Associate Professor at American University, to discuss evolution and implications of central bank independence and transparency over the past two decades.  Prof. Meade based her presentation on her working paper, “Evolution in Central Bank Governance Around the World” (coauthored with Christopher Crowe (IMF), Journal of Economic Perspectives, forthcoming in Fall 2008) to show that substantial increases in central bank independence and less dramatic, but still important, rises in transparency have occurred.
                               

To track central bank independence over the last two decades, Prof. Meade uses 2003 central banking laws for 96 countries, to replicate the Cukierman, Webb, and Neyapti (1992) study that measured legal characteristics of central bank statutes from 1980 to 1989 for 72 countries.  Cukierman developed a system that mapped legal independence into 4 categories resulting in a score from 0 (not independent) to 1 (fully independent) (see table).

Comparing these two studies, legal central bank independence has risen substantially from an average score of 0.3 to above 0.6 with 15 central banks scoring as highly independent (above 0.8).  The bulk of the increase is attributable to emerging market/developing countries that have been reforming laws and creating new central banks supported by modern independent statutes.  These increases mean nothing though if governments fail to support independence itself.  Leszeck Balcerowicz, former president of Poland’s central bank, can hardly appreciate Poland’s scoring the largest increase (0.1 to 0.9) after the government attacked his monetary policy as too restrictive.  A high score may indicate that independent laws are needed to combat potential political influences or a history of high inflation. Bosnia and Herzegovina recently scored the highest among all countries due to the constitutional provisions written during the Dayton Peace Accords, including requiring a foreigner be the initial bank governor.  This is in contrast to the Fed whose score remained unchanged, but was just below the mean relative to other central banks in the 2003 sample, partially due to its multiple mandates, which supports the argument that high credibility diminishes the need for “perfect” central banking laws.

High central bank independence scores are impressive, but only as they translate into real world results, such as lower inflation.  A previous association between higher central bank independence and lower inflation, found by Alesina and Summers (1993), disappears with the recent study.  As the mean and variance of inflation declined substantially over the last two decades, perhaps central bank independence characteristics have become less important.  However, legal and actual central bank independence might not be the same thing though, especially for developing countries. Cukierman hypothesized that higher turnover rates of the central bank governor can indicate lower actual central bank independence.  Turnover among advanced economies has increased (implied length of term in years went from 7.7 to 5.2 years) and turnover among emerging market/developing economies has decreased (3.4 to 4.8 term years).  As suggested by Cukierman, inflation does have a positive association with turnover in developing countries, although no association exists for the entire sample. 

Regression yields only a significant relationship between inflation and advanced economies, but not for the entire sample.  Problems with causality and omitted variables abound though in using regressions.  Turnover can purposely increase if inflation remains stubbornly high.  The attention brought to central bank independence may have encouraged nations to improve their own central banking laws.    Suggestively, regressing the changes in various central bank independence characteristics yields a significant negative association between inflation and central banking objectives.

Downsides can exist with too much central bank independence in some instances.  The “Fazio” problem appropriately refers to Antonio Fazio, former Italian central bank governor, whose life tenure became controversial when scandals led to a cry for his resignation by government officials.  Concerns over the ECB having a “democratic deficit” in terms of accountability exist as major steps, including ratifications of treaties, must be taken to change central banking laws. 

Greater central bank transparency through good communications, although difficult to measure, helps reduce uncertainty over what monetary policy is trying to achieve and allows the central bank to influence longer term interest rates.  Prof. Meade mapped two surveys, one from Fry, Julius, Mahadeva, Roger, and Sterne (2000) and another from Eijffinger and Geraats (2006) (increased in sample size by Meade), into five categories in order to compare central bank transparency (see table).

The 2006 scores ranged from lows of 0.15 (China, India, Singapore) to 1.0 (UK).  Only advanced economies showed statistically significant increases from 0.56 to 0.63.  Other statistically significant results include increases among countries that began targeting inflation, 0.56 to 0.71, increases among EMU members, 0.45 to 0.60, and a large decrease in countries that neither targeted inflation or were EMU members, including China, Russia, India, Singapore, and the Fed.  Explaining this drop can owe to central banks trying to conceal information from market participants, as might be the case for China, or that problems exist in the using surveys, such as interpretations, wording, and not capturing all of the actual improvement in transparency.  Over the last decade, the Fed made subtle, yet important, transparency improvements that affected the markets but might not have affected the survey results, such as releasing roll call votes on the Federal funds rate target.  

An inverse association exists between transparency and the level of inflation, but it is not statistically significant.  Christopher Crowe, Meade’s partner, previously suggested that inflation targeting can lead to improved accuracy and lower variances in private sector inflation forecasts.  To test this hypothesis, the standard deviations of one year forecasts from Consensus Economics for 28 countries were used.  Regardless of their level of transparency, trusted and credible monetary regimes of advanced economies showed lower uncertainty.  However, a negative association exists between uncertainty and transparency for developing countries, especially those not targeting inflation.  While most transparency literature suggests more is better, recent literature investigates transparency drawbacks, including the effect of exposing central bank committee deliberations to the public. Alan Blinder, former Vice-Chair of the Fed, suggests that “One size fits all” may not apply for transparency due to difference in central banks. 

Prof. Meade ended by discussingon a discussion about central banking committees and committee decision making.  Most central banks now have a committee.  Questions over optimal committee size exist as the ECB looks to expand even further with future EU member growth.  Using monetary policy experiments, Blinder and Morgan (2005) found that committees make “better” decisions in groups of five versus an individual and do not take longer to make them than individuals.  Sibert (2006) also has found through social science literature that the optimal size appears to be about five members. These results indicate that many central banks have committees that may be too large, such as the ECB and the Fed.  Research also points out that regional economic developments can affect regional member’s voting. Central banks range among their suppression of member’s internal dissent, with the ECB being highly restrictive and the Bank of England encouraging individualistic views.  Research also suggests that different monetary policy objectives can produce different decision making structures, but there is no evidence that this has any effect on monetary policy or interest rates.

Dr. Meade is an Associate Professor of Economics at American University with recent research focusing on central bank decision making, monetary integration, and exchange rate regimes. She previously served as a Senior Research Fellow at the London School of Economics, Senior Economist for the Federal Reserve Board of Governors and for the Council of Economic Advisers. She welcomes any comments to meade@american.edu. Her lunch slideshow and our podcast (NEC #11) are available online.

Kaylon McInelly, Rapporteur

Home About Calendar Speakers Links Contact Join