What is the Corruption Perceptions Index?

The TI Corruption Perceptions Index (CPI) this year ranks 102 countries in terms of the degree to which corruption is perceived to exist among public officials and politicians. It is a composite index, drawing on 15 different polls and surveys from nine independent institutions carried out among business people and country analysts, including surveys of residents, both local and expatriate.

For the purpose of the TI indices, how is corruption defined?

The CPI focuses on corruption in the public sector and defines corruption as the abuse of public office for private gain. The surveys used in compiling the CPI tend to ask questions in line with the misuse of public power for private benefits, with a focus, for example, on bribe-taking by public officials in public procurement. The sources do not distinguish between administrative and political corruption.

Why is the Corruption Perceptions Index a valuable tool?

Because the CPI is derived from 15 different surveys that garner the perceptions of both residents and expatriates, both business people and risk analysts, the index provides a snapshot of the views of decision-makers, who take key decisions on investment and trade. The CPI builds public awareness of the corruption issue, and it adds to pressure on governments to directly address the issue and the damaged image of their nation that low rankings in the CPI reflect.

What role is played by exporters in international criminal transactions?

On 14 May 2002, TI published the second Bribe Payers Index (BPI), which ranked exporting countries according to their propensity to offer bribes. (The first BPI was published in 1999.) This BPI is accessible on the internet at http://www.transparency.org/surveys/index.html#bpi. The BPI complements the CPI and underlines the point that corruption in international business transactions involves both those who take and those who give. Looking only at those who take, the CPI provides an incomplete picture.

Is it right to conclude that the country with the lowest score is the world's most corrupt country?

No. Firstly, the country with the lowest score is the one perceived to be the most corrupt of those included in the index. The CPI is based on polls, which are snapshots in time and reflect both opinions and experience. Furthermore, there are more than 200 sovereign nations in the world and the CPI 2002 ranks only 102. Although this is the highest number of countries in the history of the CPI, TI does not have sufficient reliable data for all countries.

Why is the CPI based on perceptions only?

It is difficult to base comparative statements on the levels of corruption in different countries on hard empirical data, e.g. by comparing the number of prosecutions or court cases. Such cross-country data does not reflect actual levels of corruption; rather it highlights the quality of prosecutors, courts and/or the media in exposing corruption. The only method of compiling comparative data is therefore to build on the experience and perceptions of those who are most directly confronted with the realities of corruption.

Was there any change in the target groups polled for the CPI this year?

There are changes this year, which should serve as a warning not to overstate year-on-year comparisons. This year we used 15 polls from nine institutions compared with 14 polls from seven institutions in 2001. The robustness of the CPI findings is enhanced by the fact that residents' viewpoints were found to correlate well with those of expatriates. In the past, expatriates surveyed were often western businesspeople. The viewpoint of less developed countries seemed underrepresented. This has changed. On behalf of Transparency International, Gallup International surveyed respondents from emerging market economies, asking them to assess the performance of public servants in industrial countries. The results from this survey correlate well with other sources, indicating that the CPI gathers perceptions that are invariant to cultural preconditions and represent a global perspective.

What are the criteria in determining which surveys are used?

TI seeks excellent data for the CPI and, to qualify, the data has to be well documented, and it has to be sufficient to permit a judgement on its reliability. TI strives to ensure that the sources used are of the highest quality, that the survey work is performed with complete integrity and that the methodologies used to analyse findings are first class. TI is confident that these criteria apply to the CPI. A more detailed description of the underlying methodology has been written for the 2002 index and is available at www.gwdg.de/~uwvw/2002_CPI_FD.pdf. The methodology used is reviewed by a Steering Committee consisting of leading international experts in the fields of corruption, econometrics and statistics. Members of the Steering Committee make suggestions for improving the CPI, but the management of TI takes the final decisions on the methodology used. Consequently, neither the CPI methodology nor its presentation modality ought to be interpreted as necessarily constituting endorsement by the Steering Committee or its individual members.

Which countries are included in the CPI 2002?

TI requires at least three sources to be available for a country before considering the database sufficiently robust for that country to be ranked in the CPI. Countries for which there might be only one or two data sources available are not included in the CPI. Albania, Angola, Belarus, Ethiopia, Georgia, Haiti, Jamaica, Madagascar, Morocco, Paraguay and Sri Lanka have been included because three sources are now available, but were not in 2001.

Were there any countries that were included in the CPI 2001, but not in the CPI 2002?

No. All countries included in the CPI 2001 are also included in this year's CPI.

Is the country score a reliable measure of a country's perceived level of corruption?

In terms of perceptions of corruption, the CPI is a solid measurement tool. The reliability differs, however, between countries. Countries with a low number of sources and large differences in the values provided by the sources (indicated by a large Standard Deviation) convey less reliability as to their score and ranking.

Are old surveys used in the CPI?

The CPI 2002 is based on 2000-2002 data. Since fundamental changes in the levels of corruption in a country evolve only slowly, while public perceptions may change more swiftly and be influenced to some extent by short-term events, TI determined to base the CPI on a three-year rolling average. Hence, this year's CPI is based on survey data collected exclusively between 2000 and 2002.

Which sources have contributed to the assessment of each individual country?

A list of the sources that contributed to the assessment of each country is available on the Internet as an Excel sheet www.gwdg.de/~uwvw/2002CPI.xls. This list also provides further information on confidence intervals for each country. It also lists all the countries (including their scores) that did not qualify for inclusion in the CPI due to lack of data.

Can data from one year be compared with that from a previous year?

Comparisons to the results from previous years should be based on a country's score, not its rank. A country's rank can change simply because new countries enter the index and others drop out. A higher score is an indicator that respondents provided better ratings, while a lower score suggests that respondents revised their perception downwards. However, year-to-year shifts in a country's score can result not only from a changing perception of a country's performance but also from a changing sample and methodology. With differing respondents and slightly differing methodologies, a change in a country's score may also relate to the fact that different viewpoints have been collected and different questions been asked.

As compared with the CPI 2001, in 2002 Bangladesh's score has improved by 0.8 points. However, this change was due solely to methodological changes: the new standardisation technique avoids negative numbers. For example, last year the worst individual score provided to Bangladesh was -1.7. This year the worst standardised score is 0.3 due to the methodological changes. Bangladesh's original values had remained largely constant. As a consequence, the higher score in 2002 in no way reflects actual improvements. Quite the contrary: the results obtained in 2001 were corroborated by this year's CPI. There is now even much more agreement (that is, a low standard deviation of the data) that Bangladesh is perceived to have exceptionally serious problems with corruption. For the expert, maybe the following additional piece of data might be relevant. Had we used our former methodology, in 2002 Bangladesh would have obtained a score of 0.2, with a standard deviation of 1.4. The High-Low range would have been from -1.4 to 1.8.

Which countries' scores deteriorated most between 2001 and 2002?

Making comparisons from one year to another is problematic. However, to the extent that changes can be traced back to a change in the results from individual sources, trends can be cautiously identified. Noteworthy examples of a downward trend are Argentina, Ireland and Moldova. The considerable decline in their scores does not result from technical factors - actual changes in perceptions are therefore likely.

Which countries improved most compared with last year?

With the same caveats applied, on the basis of data from sources that have been consistently used for the index, improvements can be observed for Hong Kong, Slovenia, South Korea, Dominican Republic and Russia.