Governance Matters III: Governance Indicators
for 1996, 1998, 2000, and 2002
Daniel Kaufmann, Aart Kraay, and Massimo Mastruzzi
Six dimensions of governance are estimated covering 199 countries and territories for
four periods: 1996, 1998, 2000, and 2002. The indicators are based on several hundred
individual variables measuring perceptions of governance drawn from 25 data sources
constructed by 18 organizations. These individual measures are assigned to categories
capturing key dimensions of governance. An unobserved-components model is used to
construct six aggregate governance indicators in each of the four periods. Point esti-
mates of the dimensions of governance are provided as well as the margins of errors for
each country for the four periods. Methodological issues are also addressed, including
tests for potential biases, and the interpretation and use of the data, given the estimated
margins of errors for the indicators. The data and a Web-based graphical interface are
available online at www.worldbank.org/wbi/governance/govdata2002/index.html.
This article presents estimates of six dimensions of governance for 199 countries
and territories for 1996, 1998, 2000, and 2002 developed in the context of an
ongoing project to measure governance across countries. SectionI describesthe data
used in developing this round of the governance indicators, which include several
new sources. Data sources used in the earlier studies were updated forward to 2002
and backward to 1996, and previously estimated indicators for 1998 and 2000 were
revised to reflect the new data. The aggregation procedure, described in section II,
provides not only estimates of governance for each country but also measures of the
precision or reliability of these estimates. Although the new data have improved the
precision of the governance indicators, the margins of error remain large relative to
the units in which governance is measured, so that comparisons across countries
and especially over time should be made with caution. Measurement error is not
unique to these indicators but is pervasive among all measures of governance and
Daniel Kaufmann is director at the World Bank Institute; his e-mail address is dkaufmann@worldbank.
org. Aart Kraay is senior economist in the Development Research Group at the World Bank; his e-mail
address is akraay@worldbank.org. Massimo Mastruzzi is research analyst at the World Bank Institute; his
e-mail address is mmastruzzi@worldbank.org. The authors thank Merli Baroudi, David Cieslikowsky, Peter
Cornelius, Rui Coutinho, Guy Dunn, Rich Fullenbaum, Joel Hellman, Phil Keefer, Marta Lagos, Marc Levy,
Satish Mannan, Fiona Paua, Adrian Shute, Craig Webster, and Jennifer Windsor for providing data and
answering numerous questions; Steve Knack, Clay Lowery, and Steve Radelet for their comments; and Erin
Hoffmann for assistance. The support and collaboration of the World Economic Forum, the U.S. State
Department, the Netherlands government, and Italian Trust Funds is appreciated.
THE WORLD BANK ECONOMIC REVIEW, VOL. 18, NO. 2,
Ó The International Bank for Reconstruction and Development / THE WORLD BANK 2004; all rights reserved.
doi:10.1093/wber/lhh041 18:253–287
253
Public Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure Authorized
77472
institutional quality. An advantage of the measures used here is that explicit margins
of error reflecting this measurement error can be computed.
Section III examines issues related to the construction and use of the governance
indicators, such as the usefulness of subjective measures of governance relative to
alternatives. It also empirically investigates the importance of ideological biases in
expert assessments of corruption, finding little evidence that they exist. To illustrate
the consequences of the substantial margins of error associated with the governance
indicators, the aid allocation rules proposed for the U.S. government’s Millennium
Challenge Account, which rely on these measures, are examined in section IV.
Section V explores the limited evidence available on global trends in governance,
and section VI compares the Control of Corruption indicator estimated here with the
widely used Corruption Perceptions Index produced by Transparency International.
I. M
EASURING GOVERNANCE
In this study governance is defined broadly as the traditions and institutions by
which authority in a country is exercised. This includes the process by which
governments are selected and replaced, the capacity of the government to formu-
late and implement sound policies, and the respect of citizens and the state for the
institutions that govern economic and social interactions among them. This defini-
tion guides the construction of the governance indicators for this study.
Governance Clusters
Data on perceptions of governance from a large number of sources are organized
into six clusters corresponding to six dimensions of governance. The first two
clusters are intended to capture the process by which those in authority are selected
and replaced. One cluster, referred to as ‘‘voice and accountability,’’ includes
indicators of the political process, civil liberties, and political rights. These indica-
tors measure the extent to which citizens are able to participate in the selection of
governments. This category also includes indicators measuring the independence
of the media, which serves the important monitoring role of holding those in
authority accountable for their actions. The second cluster, ‘‘political stability
and absence of violence,’’ combines several indicators measuring perceptions of
the likelihood that the government in power will be destabilized or overthrown by
unconstitutional or violent means. This cluster captures the idea that the quality of
governance is compromised by the likelihood of a wrenching change in govern-
ment that directly affects the continuity of policies and undermines the ability of
citizens to peacefully select and replace those in power.
1
1. There is some ambiguity about the normative direction of a few of the subcomponents this
indicator. For example, a few sources rank countries such as Cuba and the Democratic Republic of
Korea highly in terms of their political stability, which simply reflects the longevity of the governments
in power.
254 THE WORLD BANK ECONOMIC REVIEW, VOL. 18, NO. 2
The next two clusters summarize indicators of the ability of the government
to formulate and implement sound policies. The ‘‘government effectiveness’’
cluster combines the quality of public service provision, the quality of the
bureaucracy, the competence and independence of the civil service, and the
credibility of the government’s commitment to policies. The focus is on inputs
the government needs to produce and implement good policies and deliver
public goods. The second cluster, ‘‘regulatory quality,’’ focuses on the policies
themselves. It includes measures of the incidence of market-unfriendly policies,
such as price controls or inadequate bank supervision, and perceptions of the
burdens imposed by excessive regulation in areas such as foreign trade and
business development.
The last two clusters summarize the respect of citizens and the state for the
institutions that govern their interactions. ‘‘Rule of law’’ includes indicators
that measure h ow we ll age nt s abi de by the ru les of society. T hese include
perceptions of the inci den ce of crime, th e e ffe ctiven ess and predictabi lit y of
the judiciary, and the enforceability of contracts. Together, these indicators
measure a society’s success in developing an environment in which fair and
predictable rules form the basis for ec onomic and social interactions and pro-
perty rights are protected. The final cluster, ‘‘control of corruption,’’ measures
perceptions of corru pt ion, con venti onal ly defi ne d as the exercise of pub lic
power for private gain. The focus of the various sources differs somewhat,
ranging f rom the frequency of additional payments needed to get things done
to the effects of corruption on the business environment and ‘‘grand corrup-
tion’’ in the political arena. Corruption is often a manifestation of a lack of
respect of b oth the corrupter (typically a private citiz en or firm) an d the
corrupted (typically a public official or politician) for the rules that govern
their interactions.
Sources of Governance Data
Some 250 individual measures from 25 sources produced by 18 different organ-
izations are used in constructing the 2002 indicators (table 1; further details on
each source and on how questions from each source were assigned to the six
governance clusters are available online at www.worldbank.org/wbi/governance/
govdata2002/index.html). These organizations include international organiza-
tions, political and business risk-rating agencies, think tanks, and nongovernmen-
tal organizations. Six new data sources are included in 2002: Afrobarometer, a
survey of individuals in 12 African countries; Reporters without Borders, an
assessment of press freedoms compiled by an international journalist organization;
Human Rights, a numerical coding of assessments of certain dimensions of
human rights as reported by the U.S. State Department and Amnesty International
(first reported in Cingranelli and Richards 2001 and subsequently updated and
expanded by Craig Webster); World Markets Online, a commercial risk-rating
agency; Voice of the People, a citizen survey sponsored by Gallup International;
and the World Bank’s Country Policy and Institutional Assessment (
CPIA),
Kaufmann, Kraay, and Mastruzzi 255
TABLE 1 . Sources of Governance Data
Country
Availability
Source Publication Code Type coverage
a
Representative 1996 1998 2000 2002
Afrobarometer Afrobarometer Survey AFR Survey 12 x
Business Environment
Risk Intelligence
Business Risk Service BRI Poll 50 x x x x
Business Environment
Risk Intelligence
Qualitative Risk Measure QLM Poll 115 x x x x x
Columbia University State Capacity Project CDU Poll 98 x x x
Economist Intelligence Unit Country Risk Service EIU Poll 115 x x x x x
European Bank for Reconstruction
and Redevelopment
Transition Report EBR Poll 26 x x x x
Freedom House Nations in Transition FHT Poll 27 x x x x
Freedom House Freedom in the World FRH Poll 192 x x x x x
Gallup International Gallup Millennium Survey GMS S 60 x
Gallup International 50th Anniversary Survey GALLUP Survey 44 x
Gallup International Voice of the People Survey GAL Survey 46 x
Heritage Foundation/
Wallstreet Journal
Economic Freedom Index HER Poll 161 x x x x x
Institute for Management
and Development
World Competitiveness
Yearbook
WCY Survey 49 x x x x
Latinobarometro Latinobarometro Surveys LBO Survey 17 x x x
Political Risk Services International Country Risk Guide PRS Poll 140 x x x x x
PriceWaterhouseCoopers Opacity Index PWC Survey 35 x
Reporters Without Borders Reporters Without Borders RSF Poll 138 x x
Global Insight’s DRI
McGraw-Hill
Country Risk Review DRI Poll 111 x x x x x
State Department/
Amensty International
Human Rights Report HUM Poll 159 x x x x x
World Bank Business Environment and
Enterprise Performance Survey
BPS Survey 18 x x
World Bank World Business Environment Survey WBS Survey 81 x x x
World Bank Country Policy and Institutional
Assessments
CPIA Poll 136 x x x x
World Economic Forum Global Competitiveness Report GCS Survey 75 x x x x
World Economic Forum Africa Competitiveness Report GCSA Survey 23 x
World Markets Research Center World Markets Online WMO Poll 186 x x
a
Countries included in the most recently available version of source.
256
an assessment of country performance constructed by World Bank country
economists.
2
Two of the new sources are also available for earlier years (Human Rights
and
CPIA). To improve the comparability of the governance indicators over time,
indicators for 1998 and 2000 were revised to incorporate these sources. Two
minor sources used in the past and no longer available were discarded.
3
For
convenience the revised indicators are referred to as indicators for 1998
and 2000, even though the measures are based on data for a two-year period
(1997–98 and 2000–01). A subset of indicators is also available for 1996. These
were used to construct new aggregate governance indicators for 1996.
Two categories of sources are used: polls of experts and surveys of business-
people or citizens. The choice between polls and surveys involves tradeoffs
between cross-country comparability and firsthand knowledge of local con-
ditions.
4
Polls of experts are designed to provide comparable results across
countries through elaborate benchmarking. However, their reliability depends
on the ability of a small group of experts to provide accurate assessments of
the governance dimensions being measured.
5
Surveys typically draw on the
responses of large numbers of respondents with direct knowledge of local
conditions. However, to the extent that ostensibly identical survey questions
are interpreted differently by respondents with different cultural or socioeco-
nomic backgrounds, it can be difficult to make cross-country comparisons.
6
How representative the sources are of the world as a whole is also important.
A number of sources cover a large sample of developed and developing eco-
nomies, whereas others cover very narrowly focused samples. Many of the
poorest and smallest countries tend not to be covered by commercially oriented
2. Transparency International’s Corruption Perceptions Index (CPI) is not used as a component of our
aggregate corruption indicator because the
CPI is itself an aggregate of sources that are already included in
the corruption indicator constructed here.
3. These are the Central European Economic Review, which rated a sample of transition economies
and ceased publication after our only use of this source in the 1997/98 indicators, and the Political and
Economic Risk Consultancy, which has also discontinued its rating of a small number of Asian eco-
nomies. Dropping these sources does not affect country coverage, and it makes the aggregate indicators
more comparable over time.
4. For a more detailed discussion of the advantages and disadvantages of polls of experts relative to
surveys of market participants, see Kaufmann and others 1999b, 2002, and Kaufmann and Kraay 2002.
5. Most of the polls of experts cover large groups of raters. For example, the Economist Intelligence
Unit, based in London, draws on the views of a worldwide network of correspondents for its assessments,
as does Freedom House, which is based in New York, and Reporters without Borders, based in France.
Other polls of experts have a narrower institutional affiliation for their respondents. For example the
European Bank for Reconstruction and Development Transition Report ratings are based primarily on
the assessments of its staff in London, and the State Department component of our Human Rights
measure reflects the views of U.S. State Department employees.
6. The three main sources of firm-level survey data (the Geneva-based World Economic Forum’s
Global Competitiveness Report, the Lausanne-based Institute for Management Development’s World
Competitiveness Yearbook, and the Washington-based World Bank’s Business Environment Surveys)
interview primarily domestic rather than foreign-owned firms in the countries they cover.
Kaufmann, Kraay, and Mastruzzi 257
