Central Bank Credibility and Reputation: An Historical Analysis Presentation at the 7 th SEEMHN Conference Tales of Two Crises: The Great Recession and.

Central Bank Credibility and Reputation: An Historical Analysis

Presentation at the 7th SEEMHN Conference “Tales of Two Crises: The Great Recession and the

Great Recession in South-East Europe”Pierre L. Siklos, Wilfrid Laurier University

& Balsillie School of International Affairs, Waterloo, Canada

Back Story

• Prior to the ‘crises’, CB & even academics – at least in my part of the world – struggled to accept the relevance of historical analyses of MP

• Fortunately, some CB ignored the critics and pushed to create historical databases– A shining example: Norges Bank’s 2016 project

• “The joint publication of a data volume of harmonized long-run time series on monetary, financial and other macroeconomic variables” for SEE is welcome and I hope to use it for the present project!

Premise

• There are ‘swings’ in the credibility & reputation of central banks

• These swings may be tied to the reliance on the commitment to maintain ‘policy rules’ which contained ‘escape clauses’– An example: the Gold Standard• Some CB were successful: Britain, France, Germany• Others were not:

WARNING!

• This is very much work in progress (with Michael Bordo)– Comments are welcome– Our minds can be changed and our views will no

doubt evolve!– Only descriptive evidence today; formal results

will have to wait until April 2013

Objective

• Can we quantify the evolution of CB credibility over time and changes in its reputation?– Domestic component: legislation, monetary policy

strategy, overall economic & political environment– External component: exchange rate regime, CB

cooperation, financial crises• Emphasis in the literature so far has been on

theoretical predictions

CB Credibility: Definition

• The “…extent to which the public believes that a shift in policy has taken place when, indeed, such a shift has actually occurred.“(Cukierman 1986, p.6)

• It is best thought of as a flow like variable, rising and falling over time as monetary policy strategies change, subject to the evolving nature of central bank-government relations, and is even a function of political factors and the personalities who implement policies.

CB Reputation: Definition

• Can be likened to a stock variable. • The reputation of a central bank can be built

up over time as its performance improves or it can be diminished if a financial crisis is mishandled or there is a serious crisis of confidence in the governor or a debilitating conflict between the monetary authority and the government.

Elements of Credibility & Reputation

• ‘Identification’ problem has always been a challenge: two concepts are clearly intertwined

• Elements of credibility & reputation building– A ‘track record’...but this requires a benchmark: inflation, relative

to a forecast, a target, a definition of ‘price stability’• MEASURABLE IN THE SHORT-RUN AND OVER A LONG TIME SPAN

– A clear division of responsibilities with the fiscal authority • MEASURABLE OVER THE LONGER-RUN, A TIMES OF CRISES, IN THE

LEGISLATIVE RECORD, AND OVER A LONG TIME SPAN

– Clear and effective communication• PUBLIC AND PRIVATE COMMUNICATION BOTH MATTER, MEASURABLE

OVER A LONG TIME SPAN BUT...A SEA CHANGE TOOK PLACE BEGINNING IN THE 1990s especially

Theories of Credibility & Reputation

THE NARRATIVECB Credibility & Reputation: An Historical Analysis

Then and Now: Governance of Monetary Policy

• THEN– Little de jure

independence, but some de facto independence (Friedman’s personalities hypothesis): Strong, Norman, Moreau, Schacht• Not viewed as essential• CB were a ‘tool’ of the state

– Decisions followed a ‘top-down’ view (Genoa 1922, Bretton Woods 1944)

• NOW– CB autonomy begets

accountability which fosters transparency

– Greater international cooperation among CB

– Still a ‘top-down’ approach (e.g., G20, Basel)...is it still due to complexity?

Financial Crises

• SOURCES– Currency– Banking – Sovereign debt

• TYPES– Systemic– ‘borderline’

Then and Now: Financial Crises and CB

• THEN– Gold standard (later the GES) was

the coordinating/cooperating device. The concern was over adherence the ‘rules of the game’: ‘CB the only game in town’

– Ideology promoted a ‘steady as she goes’ approach, a reluctance to consider ‘unorthodox’ policies when required

– With few exceptions modest changes in the legislative/regulatory environment

– Initially, little sober second thinking about the role of MP

• NOW– Multiple agencies play a

role but are CB still ‘the only game in town’?• Governance implications

– International cooperation is fluid, there is no effective coordination mechanism• Concern over spillovers &

contagion but less so over the rules of the game

– “Do whatever it takes” attitude

Then and Now: Recoveries and CB

• THEN– Fiscal policy a secondary

or effectively non-existent tool

– Little thought given to the coherence between fiscal & monetary policies

– Blurring of fiscal and MP not a concern

• NOW– Fiscal policy has swung

from passive to active and seeks to be passive again

– Persistent doubts about the impact of fiscal policy

– Blurring of fiscal and MP is a recurring concern

The Narrative: A Summary

What Next?

• PREMISE– Central to the consequences of a financial crisis, recoveries,

and stability more generality are CB credibility & reputation– Credibility is reflected in deviations from trend in an

observable macroeconomic variable associated with MP, such as inflation/output• MACROECONOMIC PERFORMANCE DOMINATES

– Reputation is reflected in the duration of a MP strategy, low turnover in CB governors (i.e., personalities), ability to respond to a crisis• INSITITUTIONAL CHARACTERISTICS DOMINATE

Data

• Annual since the formation of 20+ central banks (around the world)

• Econometric & narrative approach co-mingled• Sources of data from datasets authors have

constructed over the years, as well as other available historical datasets

Methodology I

• Part I:– Central bank performance is typically assessed according to the

behavior of inflation or the price level• In an historical study it is unclear how we should define the

benchmark against which inflation deviates from some expected value• We consider a number of approaches:

– We apply statistical break tests to determine breaks in the price level. This permits us to evaluate one indicator of deviations of realized inflation from some expectation, namely deviations from a statistically estimated trend inflation rate;

– We can specify a time series model to estimate expected inflation. Models based on a New Keynesian type Phillips curve work well over long periods of time and, subject to ‘breaks’, can also serve as a useful benchmark;

– For the more recent history, that is, the last three decades of data, we can substitute private sector forecasts of inflation, or inflation targets set by central banks as a robustness test

Methodology II

• We then ask how the constructed credibility indicator reacts to the past history of inflation, various proxies for economic growth performance, or the output gap, the stage and shape of particular business cycle events

• We also aim to empirically establish whether credibility behaves asymmetrically over time

• A neglected aspect of the historical analysis of central bank credibility and reputation is how these can be transmitted across countries. Certain policy regimes (e.g., a pegged exchange rate) have long been known to permit countries to ‘import’ inflation performance from abroad. Whether central bank credibility and reputation can be similarly imported is another matter

Methodology III

• We estimate a model of credibility using a factor-augmented GVAR model– Designed to capture the transmission of shocks across

countries and across time– The GVAR model begins by estimating N-country specific VAR

models that focus on the role of largely domestic sources of shocks or global exogenous shocks (e.g., oil prices, wars)

– A second step then is to incorporate foreign shocks. This is accomplished by adding a cross-country weighted average of foreign shocks where the latter are assumed to be (weakly) exogenous and are constructed on the basis of VARs estimated in the first step

Methodology IV• Part II

– We relate reputation to the personality at the head of central bank and the length of time the governor is in place• We normalize this variable according to the length of time a governor is statutorily

entitled to be in office

– The reputation of a central bank will also be influenced by the quality of governance in the public sector• Indicators of governance (e.g., type of political system, political and economic

freedom, protection of property rights, quality of the justice system, etc…) are likely to be associated with central bank reputation

– Central bank governors and governance indicators might interact with each other. In this case a combined variable can be considered as a proxy for the reputation of the monetary authority• Reputation might also be determined by governance structures that define the

relationship between the central bank and government, including the degree of central bank independence

Methodology V

• Estimate a quantile regression model which allows us to estimate whether central banks that have particularly high reputations are associated with a type of policy regimes– Designed to estimate relationships where the investigator is not

interested in only mean responses but rather wishes to know whether the response of the variable of interest, here central bank reputation, differs across the distribution of the indicator of reputation

– Investigate whether central banks with high reputations are less prone to seeing a reduction when they are less credible, or in response to some other economic shock, as well as relative to central banks who have a lower reputation

Source: T. Lane (2012), “Weathering theHeadwinds to Canada’s Economic Growth”,Bank of Canada, 21 November.

Source: T. Lane (2012), “Weathering theHeadwinds to Canada’s Economic Growth”,Bank of Canada, 21 November.

Source: T. Lane (2012), “Weathering theHeadwinds to Canada’s Economic Growth”,Bank of Canada, 21 November.

1,240,000

1,280,000

1,320,000

1,360,000

1,400,000

1,440,000

1,480,000

I II III IV I II III IV I II III IV I II III IV I II

2008 2009 2010 2011 2012

real GDP 2012Q2 Potential real GDP 2008Q1Potential real GDP 2009Q2 Potential real GDP 2012Q2

Mill

ions

of 2

002

Chai

ned

$

5.6%

3.2%

0.5%

Latest release of real GDP against select vintages of potential real GDP. I have extrapolated potential real GDP, using 2008Q1, using a simple linear trend as the actual series is virtually linear towards the end of the sample.

The simple point is that, unless one is clear about whether a ‘structural’ shift has taken place and the source of this shift – this is entirely unclear to me I must confess – the cumulative output loss since the recession of 2008-9 could be rather large.

-4

-3

-2

-1

0

1

2

3

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

January 2008 December 2008December 2009

Perc

ent

-12

-10

-8

-6

-4

-2

0

2

4

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

July 2009 August 2009

-10

-8

-6

-4

-2

0

2

4

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

June 2010 December 2010March 2011 August 2011

Perc

ent

Output gaps - US

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2000 2002 2004 2006 2008 2010 2012 2014

CBO 2000 prices - 2009 vintage CBO 2005 prices - August 2011 vintage

US Potential Output- CBO EstimateN

orm

aliz

ed

Source: Almunia et. al. (2010)

Source: Reinhart & Rogoff (2012),“This Time is Different Gain?...”