070131 CPI analysis and comments (final)

© Michel van Hulten, January 2007

www.corruptie.org

Ten years of Corruption (Perceptions) Indices

as established by Prof. Dr. Johann Graf Lambsdorff[2] (University of Passau)

and yearly published by Transparency International in the years 1995-2005

METHODS – RESULTS – WHAT NEXT? An analysis

What went right, what wrong?

What can be done to improve this tool?

by

Dr. Michel van Hulten[3]

Individual Member of Transparency International

31 January 2007

ISBN 978-90-811048-2-1

THESIS

The annual publication of the Corruption Perceptions Index (CPI) by Transparency International, which gives the scores and ranks with regard to the prevalence of corruption in most countries of the world, ranking those from ‘most clean’ to ‘worst’,  legitimizes the widespread but erroneous popular belief that corruption is a problem of the poor and of the Third World.

PURPOSES OF THIS STUDY

1. To invite TI to stop the annual publication of the Corruption Perceptions Index by the Transparency International Board and the TI Secretariat in Berlin.
2. To begin a process of social and economic research and of global scientific discourse, aiming for a better qualification and classification of countries and/or business sectors and/or particular groups or individuals in societies, with regard to the extent of corruption in their socio-economic environment. The results should help the fight against corruption and the struggle to arrive at more integrity in local, national and international governmental institutions and actions, and in all business dealings at home and abroad. All research already done or at present underway for these same purposes should be accessed and used as building blocks for next phases of studies and actions.

Acknowledgements

I sent my first draft for this document late in 2006 to some of my colleagues fighting corruption and safeguarding integrity, asking comments, criticism, additional information and arguments to strengthen my findings and conclusions.

I harvested support, agreement, inspiring thoughts, and leads for rethinking what I had done. All that input helped me to improve considerably the original text.

Of course, I did receive a lot of criticism on my imperfect use of the English language as well as on the substance of the report. Sometimes very conflicting remarks, ‘have a native speaker correct your English’, and ‘your English is perfect for all those “English as second language readers”’. On the substance I met similar contradictions: some advised me to ‘go deeper’, while others wrote me: ‘please, be more popular, otherwise no one will read this’. Some applauded the structure of this document, others advised me to shorten it, or to give it another structure. No one disagreed on the main lines.

In many respects I followed what was advised, but not all and not always. I studied all comments that I received very carefully, happy that there was such an interest, but also because I felt obliged to do this for those who were willing to spend some of their time on this important subject and on helping me to improve in the first place.

I did not listen to everybody and did not honor all suggestions. I think I can justify this also because this document is only a starting point. Much more is probably needed to get many to think about the problems of ‘measuring the immeasurable’ as Fredrik Galtung called this in the title of a chapter he published in 2006. I hope that soon more researchers and activists in this fight against corruption will share my view – and I know now that this is already the view of at least some of them. ‘The CPI is a tool that we should not use anymore because of its deficiencies’, the more that conclusion spreads, the more we will begin or will strengthen our efforts to get better tools. I hope that TI will take the lead in doing so. This means that much of what has been written in my recent correspondence with my colleagues will come back anyhow again as this will help to make better proposals. And I hope that many more voices will join this choir.

This text is fully my responsibility.

Peer reviewers:

Paul Arlman, The Hague, the Netherlands, paul@arlman.com.

Julie Bajolle, MSc, Corruption Research Network, researcher, France, juba1500@aol.com
Fredrik Galtung, Chief Executive TIRI, London UK, www.tiri.org

Arjan Geveke, Regulatory Reform Advisor Better Regulation Executive Cabinet Office, London, UK

Pawel P. Mlicki, Ph.D., Senior policy advisor VNG International, Warsaw, e-mail: pmlicki@hotmail.com

Claudio Weber Abramo, Executive Director Transparência Brasil, São Paulo, www.transparencia.org.br

CONTENTS

Thesis. Purposes of this study                                                                       page    2

Acknowledgements                                                                          page            3

Contents                                                                                            pages            4-5

Summary and conclusions                                                                pages            6-7

METHODS                                                                                       pages            8-38

Paragraph

1-14 Introduction

15-21 Apples, melons, potatoes and strawberries

22-25 Early problems

26-28 Number of sources (see attachments III and IV)

29-31 Precision of perceptions

32-36 Evolution

37-43 Reactions

44-49 How do you measure the degree of corruption?

50-54 Double standards

55-78 Sources of information

Francophone and lusophone (59)

Available? (60-61)

Continuity and consistency? (62-68)

Why discontinuity in the sources? (69-71)

Africa (72)

Where is the Arab world report? (73-74)

Table 1 Arab-world rank (73, page 22)

Should source-reports be recent? (75-78)

79-83 Should CPI-reports be transparent?

84-86 Index Advisory Committee

87-91 Other problems underlying the CPI

92-105 Respondents

Table 2, Who was asked in 1996? (95, page 27)

The respondents in the CPI-2005 Methodology report (98)
Are the samples representative? (99-105)

Table 3, Who was asked in 2005? (104, page 30)

106-123 Definition of corruption?

Table 4, What was asked in 1996? (111, page 32)

Table 5, What was asked in 2005? (123, page 35/36)

124-125 Sources for CPI-2005 compared to 1999.

RESULTS                                                                                                     pages   39-46

126-131 More South – more corrupt?

132-137 To begin with some examples

138-145 Worldwide corruption increased?
146         More than US$ 1 trillion?

147-166 CPI high ranking dominated by rich countries

Top (147-155)

Bottom (156-162)

Culture (163-166)

WHAT NEXT?                                                                                             Pages   47-52

167-193 Proposals

—–

pages

Attachment I,      Ranks, and Ia, Yugoslavia                                                            53

Attachment II,     Scores, and IIa, Yugoslavia                                                          58

Attachment III,   CPI sources, names and numbers per year                                               63

Attachment IV,   Number of scores required for inclusion in the yearly list                   65
Number of countries included                                                      65

Attachment V,     Singapore                                                                                      66

Attachment VI,   How national governments reacted to the CPI, three

reactions Pakistan (Bhutto), Malaysia, Argentina                                   69

Attachment VII, Feedback to the CPI-1998 from a journalist in Tanzania             71

Attachment VIII, CPI 2006 rank and score list                                                        73

Attachment IX,   International Corruption: Strains of sleaze. How trustworthy

   is the experts’ verdict on governments’ honesty?                        80
(Nov 9th 2006, From The Economist, print edition, Volume 381 Number 8503, page 69)

Attachment X,    Britain is ‘as corrupt as worst African states’, by Duncan
Campbell on Monday September 4, 2006, Business Guardian         82

Attachment XI,  World Economic Forum, Global competitiveness
index 2005 and 2006                                                                      84

Attachment XII, Stephen Knack, World Bank Measuring Corruption: A Critique

   of Cross-Country Indicators                                                          85

—–

Summary and conclusions

Methodologically, the Corruption Perceptions Index (CPI) is not rigorous enough in its scientific research, which is a pity, as it was over the past ten years an excellent public relations tool that wakened the world over this scourge of our society.

One of the peer-reviewers, asked to comment on my earlier draft of this document wrote:

‘My short statement is: as a marketing tool is the index in some societies very effective but this is the only excuse of its existence. It cannot be considered by no means as a serious statistical tool for measuring whatsoever.’

An outsider was even more rude and honest: ‘TI reinforced stereotypes by depicting African nations as the most corrupt’, (see attachment X).

I agree that the scores calculated per country, and the resulting rank-list of countries, year after year strengthened the general public, governments, international organizations, journalists and donors, mostly in the ‘rich world’, in their views that it is the underdeveloped nations where corruption is most rampant.

As I will prove it is the faulty methodology used over so many years that created these scores and rankings in which the richer countries are the ‘cleaner’ ones. The poor not only remain poor but it is ‘their own fault as they are not well governed and not honest’.

The scores, and therefore the resulting rank-order of countries, are built on samples of respondents, particularly working in the Western business-world: country-analysts/experts and journalists, no blue collar-worker is in sight.

Why are other categories of respondents missing? Perceptions of corruption held by others do not play a role? Why is it that perceptions of corruption held by women, consumers, the old, the young, the sick, the poor, the powerless, the ones that suffer under corruption, do not play a role? Why are no perceptions portrayed that are held in the trade-unions? Why are there no contributions from investigative journalists?
But even the businessmen alone, willing to answer questions about corruption, constitute already a biased choice. In most countries of the world you better shut up.

Remarkably enough, also missing are the elected politicians and the public servants.
Is not bribe-giving and -receiving, two-way traffic?

Most of the respondents (maybe the analysts/experts qualify better?), work as individuals in private companies and judge on the basis of their personal experiences (and on hearsay?) in daily life, and in their business relations with politicians and public servants.

If we accept that the perceptions of corruption as held by them, and as they believe do exist among their business-colleagues, why is it that their perceptions are correct and valid for the total of ‘countries’ and populations?

The CPI is a ‘composite index’ based on the work by other researchers. TI cannot possibly monitor the quality of the data-sources and surveys on which the CPI is based. No one contributes directly his or her perceptions to the final TI-results. All perceptions have already been processed and reported by others before they enter the statistical analysis and processing by TI.

Methodologically this is unacceptable as in the supporting surveys from which the composite index is constructed, questions asked differ widely. They cover perceptions of corruption, as well as perceptions of fraud, or of a general feeling of uneasiness in the business-world, or of lack of integration in a foreign land or culture. Questions do not only differ between surveys, they also differ within the same survey over the years. They do reflect a variety of definitions for corruption.

‘Corruption’, the most essential concept used in the CPI-reports is defined as ‘misuse of public power for private gain’. In later years the word ‘entrusted’ replaces the word ‘public’, an important change, as the focus shifts from solely public corruption to also including private corruption.

Moreover, over the years there is no rigidity in the choice of sources that is used.

Should it be only the most recent ones? Or, should all sources be used that originate from the same year as the CPI of that year? Or, should all reports be used in three successive years according to availability (‘roll over’ as is said in some of the CPI-reports?)

TI chooses for a variety of ways in various years.

Francophone and lusophone sources are altogether missing.

The use made of sources is sometimes discontinued, although they were available. Why? Only two sources have contributed in all the years. Why is in particular years abstained from the use of an earlier used source, also available in following years, is not motivated.

A most remarkable statement is that by making use of ever more sources, the quality of the one composite index will become better. TI/Lambsdorff genuinely believes that the good qualities of one or some of the sources, compensate for eventually lacking qualities of the others. Why not use only those sources that have good qualities? Because nobody knows which ones are good?

Undisclosed remains, which sources have been considered but have not been used for inclusion in the composite index, no reasons are given.

The World Bank estimate of the total of bribes paid in the world in 1995 was 50 billion dollars. In 2005, the World Bank estimates (calculates?) a total of 1.000 billion dollars, twenty times more.

TI/Lambsdorff agrees with this higher figure, but do not explain how this multiplication could not affect the scores and rank-order countries get assigned in the CPI.

Is bribe-paying rising in all countries proportionally with the same rate? This is highly unlikely to be true.

All over the years we see that basically the same countries are on top, the same remain in the tail of the rank-order. The CPI has no educational value?

In the ‘What next?’ chapter suggestions are given for improvements.

*****

Introduction

1. TI’s CPI-website says for 2005:

‘The annual Corruption Perceptions Index (CPI), first released in 1995, is the best known of TI’s tools. It has been widely credited with putting TI and the issue of corruption on the international policy agenda. The CPI ranks more than 150 countries by their perceived levels of corruption, as determined by expert assessments and opinion surveys.’

2. This is perfectly true.
   Indeed, it is a perfect public relations instrument. ‘The CPI has been cited in thousands of newspaper articles’[4]. In the words of Galtung:

‘The CPI was a formidable instrument in raising awareness about the international scope and shared burden of corruption and driving corruption onto the front pages of newspapers throughout the developing world’.

3. Ask Google and type ‘corruption perception index’ and you get 994.000 hits. The plural ‘corruption perceptions index’ make you score 1.160.000 hits [21 Jan. 2007].
   The CPI has put the issue of corruption on the international agenda. It came at the right moment when the world started to realize that corruption has to disappear from our societies and that integrity of governance should be promoted[5].
4. Peter Eigen, from 1993 to 2005 the president of Transparency International, wrote:

‘Many of the world’s poorest nations are perceived to be among the most corrupt. The CPI helps to draw attention to this link and it represents a challenge to leading foreign aid granting agencies to make fighting corruption a key priority. We are delighted that an increasing number of these agencies now see the CPI as a valuable tool and are evolving constructive anti-corruption strategies for developing nations.
In the last three years many leading international organizations, such as the United Nations, the World Bank, the International Monetary Fund, the Council of Europe, the European Union, the Organization of American States, the OECD and the Global Coalition for Africa have articulated anti-corruption policies, often with TI involvement.
The annual CPI sensitizes public opinion world-wide to the corruption issue, influences the policies of major aid agencies and is a factor in the foreign investment decisions of multinational corporations.’

5. He is very much convinced, as are many others, that decision-making by aid-agencies and by investors is improved because of the insights offered by the CPI.
6. Indeed, more and more decision-making authorities are guided by CPI-scores and ranks. However, do we have to be happy with this result if scores and ranks as established are of a poor quality and have to be doubted?
   If, because of a wrongly established low score and/or rank on the level of corruption, a country failed or fails to get aid and investments, does this satisfy us?
7. What the CPI-website and the eleven CPI-reports do not say is that doubts have been voiced repeatedly and continuously about the results and the methods used to get these results.
   What is ‘the message’ we receive through the yearly publication of the CPI?
8. After more than ten years of existence of the CPI, it is useful to look at the results over these years, to evaluate the methods, and to ask how reliable the results are.
   Therefore, I want to discuss the validity of the results, the quality of the methods and of the data-sources used to construct the composite index which is the CPI, and the [in my eyes] biased sampling of the respondents chosen to give their perceptions.
9. This is a difficult task, also because Prof. Lambsdorff is convinced that:

‘The strength of the CPI lies in the combination of multiple data sources in a single index, which increases the reliability of each individual score’,
[and]
‘The benefit of combining data in this manner is that erratic findings from one source can be balanced by the inclusion of at least two other sources, lowering the probability of misrepresenting a country’s level of corruption.’

10. Of course, we can always try whether this ‘can be balanced’. But does it work that way? How is it, if all three sources used for a particular country happen to deliver ‘erratic findings’? How could we possibly know that ‘findings from one source’ are ‘erratic’? Why should we use data knowing that they are ‘erratic’? How can we be sure that the ‘two other sources’ are not ‘erratic’, and that they will lower the chances to misrepresent the country’s level of corruption? And what does ‘erratic’ mean if results are based on ‘perceptions’? How do we measure and score ‘misrepresentation’ of ‘perceptions’? These questions alone already show how vulnerable the CPI is.
11. I discussed this issue of ‘erratic findings’ earlier with prof. Lambsdorff.[6]
    He disagreed with me and responded [20 June 2005]:

‘Concerning the term “erratic”: Whenever you add data that randomly distribute around a common mean, by aggregation the variance of the mean is lower than that of the individual observations as soon as the error terms are not perfectly correlated. My descriptions thus relate to a standard statistical feature.

12. As I am not at all a specialist in statistics but kept my doubts, I consulted Austin Murphy, professor of Finance at Oakland University, Rochester, USA[7]. He
    reacted to the quoted above Lambsdorff ‘erratic’ remark and my criticism and wrote to me:

’He is indeed correct if each of the measures is unbiased. However, as you point out, there probably are substantial biases given the ratings are largely by Westerners (I assume they are dominated by surveys of Western firms?). Adding biased ratings to biased ratings can actually increase the bias if the added ratings are more biased.
In fact, the entire concept of measuring only bribes of government officials is incredibly biased. Countries with bigger government sectors (and regulation) will of course show up as more corrupt using that measure just because relatively less of the trade (the business-business sector) is excluded from the measure. China itself may score low on the CPI because its firms are perceived as government firms (because of majority government control), and so more of the bribery there is included in the CPI surveys, whereas the substantial bribery of retailers in the USA by manufacturers is excluded from the CPI.
Also surveying businesses allows a bias to begin with. Businesses that are happy with the bribery system are not likely to complain. However, companies that are kept from doing things in countries because of strict regulators who are not bribable are likely to rate such countries as corrupt (and possibly even claim that the amount of the bribery needed to get anything done was just plain excessive).’

13. In the following, I will return to questions of sampling and reliability of the sources used for the CPI, and to the results as shown in the yearly CPI-reports.

    I will not discuss the statistical methodology neither the statistical calculations nor the standardization techniques used to arrive at the CPI as a composite index. I see no need to do so, after I have shown in the following that the sampling of the data-sources and of the respondents is biased, and that sources and respondents deal with different definitions of the concept ‘corruption’. They do not deal with verifiable facts but only with perceptions. Moreover these personal thoughts, called ‘perceptions’, are not about the same phenomena. The data-input is so deficient that it becomes useless to have a closer look at the statistical work done in processing these data. Clever statistical operations can never improve improperly collected data and cannot upgrade the quality of the whole process and improve the final outcome.
    Nevertheless, those who also want to read a statistical critique about the CPI and the way this composite index is constructed, should read the chapter written by Galtung mentioned in footnote 4 above.

14. As I see it, the CPI is biased as it reflects the views (‘their perceptions’) of observers all of whom belong to the culture of the Western financial and economic spheres. Their yardsticks are culturally determined.
    This does not mean that I disregard Lambdorff’s constant endeavors to include and give more voice to people from less-developed countries.
    My point is that the respondents from the less-developed economies, like their peers from OECD-countries, in sharing their corruption perceptions with us, do study, work and live in a culture dominated so to say by the Harvard Business School and the Davos Economic Summit.

    It is not a matter of skin-color but of culture.

Apples, melons, potatoes and strawberries

15. Imagine a basket with apples, some of them perfect, some of them nearly perfect, some of them overripe, some of them in various states of rot.

    We begin and put all apples on a row, beginning with the most perfect ones and at the far end the absolutely uneatable ones. Some among us may differ in opinion about what is perfect and what is impossible to consume anymore. They will define a perfect apple and the state of rotting of other apples in a different way, and therefore their scores for some of the apples will be different from the scores others give. Their rows will place the apples in another order. Nevertheless, if we ask enough people to classify our apples, somehow an average will be established which is closest to the observations as defined by each one of them.

16. However, we do not only have apples but we have also melons, potatoes and strawberries, and also among those fruits we have perfect and rotten ones, all of them in the same basket with the apples. We put them equally in rows. This would mean that we get four distinct rows in which the quality of the fruit is going down following the line.
    Yet, we are not satisfied as we want one row of all fruits, so we mix the apples, melons, potatoes and strawberries and line them up in one row according to their rot. Now we are in real trouble. Will we still get agreement of everybody on how to mix the four lines into one? Will an apple with a score of 40 percent rot be placed before or after a melon with the same percentage rot (but much more rotten body-mass), or do we put them alphabetically: first the apple, second the melon? It follows than that the potato will be on third place and the strawberry will be always last.
17. Is the classification of the apples, the melons, the potatoes and the strawberries influenced by the ones who do the classification work? You, he, I will classify? Researchers, my friends, your friends will do so? Will producers order differently from consumers? Will rows be different if classified by professional fruit-farmers, or by traders who have to make profits? What about the shopkeepers, the shoppers, the final consumers, those who paint a still-life? Those who are specialists on apples? What about the specialists on melons? Those who want to promote growing and consuming fruits, or those that want to make money the quickest way possible?
18. There is the more reason to ask these questions since in 2005 Michael Johnston[8] so convincingly explained that corruption takes quite different forms in different parts of the world and in different business cultures. Why is he quite cynical about all efforts to tell corrupted regimes that they should try to improve to get the same quality of ‘good governance’ as the countries in the top of the TI-list!
19. In summary Johnston qualifies corruption in four different shapes as ‘Influence markets’ (examples USA and North-Western Europe), ‘Elite cartel’ (Italy and South Korea), ‘Oligarch and clan’ (Malaysia and Russia), and ‘Official Moguls’ (China and Nigeria). It is too much to explain here in full length his thoughts, but it will be clear to everybody that as a consequence of this differentiation, he emphasizes that defining and fighting corruption and safeguarding integrity vary quite a bit around the world. Observers’ perceptions can be misleading given these differences in corruption.
20. In other words: if we follow the theoretical concepts of Johnston, it is not possible to make one ranking of countries according to the prevalence of corruption.
21. Scoring countries, and subsequently ranking them, according to a criterion which is more or less ‘the use made per country of the possibility to buy power and favorable decisions’ [quote from TI, the definition of corruption is discussed later], is comparable to scoring the rotting state of fruits and ranking them accordingly.

    However, it is much more difficult, as at least the apples, melons, potatoes and strawberries can be physically observed which makes that we judge on the basis of observable facts, call qualifications, and rank these. While corruption takes place in hiding, facts are difficult to observe as payer and payee are inclined to deny corrupt behavior and certainly their involvement. Are respondents, as used in the surveys aggregated in the CPI-reports, fully aware of this problem?

Early problems

22. TI recognizes in the yearly CPI-reports since 1995 that the qualifications given by businessmen and financial journalists (in later years it is not any more ‘journalists’ but ‘experts’ or ‘analysts’) are not more than perceptions of what they think they observe.

‘It is an assessment of the level at which corruption is perceived by businessmen as impacting on commercial life.’

23. Nevertheless, the press release issued by TI in 1995 already postulated:

‘New Zealand best, Indonesia worst in world poll of international corruption’

This sounds more as a judgment based on facts than on presenting a perception of reality. The title of the 1995-CPI-report ‘Corruption Index’ radiated the same view. From 1996 onwards TI adopted quickly the concept of ‘perception(s)’ of corruption.

24. The 1995-CPI was the first one ever on corruption that had a global ambition. We should consider it to be a try-out.
    It surveyed only 41 countries ‘a list that does not include some countries, because the polls surveyed do not include them’. ‘The hope’ was already expressed ‘to broaden the scope of the index in future years’.
    Later, two scores available for a country, considered enough in 1995, would be considered not enough. For instance, already in the next year, 1996, ten surveys were used and TI reported that at least four scores were required for a country to be included in the list. In later years TI reports that three scores is enough for the CPI. Why not any more four, is not explained.
25. A larger number of scores to be included was emphasized again in 2001, through a footnote attached to the CPI-ranking of that year (see attachment I, p.59, footnote 23), which expressly mentioned that data for Bangladesh (the country ended that year as the last one on the list) ‘need to be viewed with caution’ because ‘data was available from only three independent survey sources, and each of these yielded very different results’.

Number of sources (see attachments III and IV)

26. Not only the number of countries included in 1995 was rather limited, also the number of sources. Only seven surveys, three each from two institutions: the Institute for Management Development in Lausanne (1992-1993-1994), and from the Political and Economic Risk Consultancy Ltd, Hong Kong (1992-1993-1994), and a 1980 (!) survey from Business International, New York, contributed to the final result.
27. The explanatory note published with the list of countries and their scores and ranking mentioned: ‘the index only includes countries for which a minimum of two scores (and in some cases as many as 7) exist’. The methodology includes that disagreement between the various polls was taken into account by averaging the scores and calculating the variance and publishing these results with the score-list and the rank-order. It is not explained why a 1980-survey was also included in this ‘poll of polls’ established in 1995.
28. Numbers went up in 1996. Ten surveys were used and at least four surveys were required for a country to be included in the list.

‘The index is a “poll of polls”. It has been prepared using ten sources, including three from the World Competitiveness Report, Institute for Management Development, Lausanne, three from the Political & Economic Risk Consultancy Ltd, Hong Kong, one small survey by Peter Neumann, published in the monthly German magazine, Impulse, No. 4/1994, two assessments by DRI/McGraw-Hill Global Risk Service and by the Political Risk Services, East Syracuse, NY, and the first incoming replies to the internet service of Göttingen University (http://www.uni-goettingen.de/~uwvw/) which gives contributors the possibility for anonymous contributions and also directly approaches employees of multinational firms and institutions.’
 

Precision of perceptions

29. The expression of precision was strengthened in the early years, 1995-1997, by presenting the scores per country with two digits after the comma. This was too much precision as the produced information was based on perceptions and not on well-established facts that could be proven.
30. Obviously this could not last. From the score-list in 1998 onwards this was reduced to scores with one digit. A great advantage of presenting two digits is that all countries could easily be ranked. With one digit, countries may more easily end up with an equal score, in which case the alphabetical order of their country-names decides about their rank in the CPI. TI solves this problem every year by listing those countries ex aequo[9].
31. With a growing number of countries included in the list, chances increase for countries to reach the same score, as is shown. Most of it is in the lower scores.

Evolution

32. More important is, that after more than ten years of experience, and indeed a growing interest in the whole world in corruption and integrity, nobody denies anymore the existence of corruption everywhere.
33. Shifts in the name (1995: ‘Corruption Index’, 1996: ‘Corruption Perception Index’, 1998 until today: ‘Corruption Perceptions Index’), confirm that the making of the TI-index is a learning process.
34. Shifts can also be observed in the sources used to establish the scores. Some sources of earlier years have disappeared, new ones added
    [see Attachment III, CPI sources, names and numbers per year].
35. The number of countries included in the yearly list grows from 41 in 1995 to 159 in 2005, (163 in 2006).
36. Also growing is the number of institutions and surveys available from which to draw the data required for the TI-scores and ranks. The number of surveys grows from 7 in 1995 to 16 in 2005 [see Attachment IV[10]].

Reactions

The Impact of the CPI

37. TI noted that since its inception in 1995, the CPI has served the constructive purpose of stimulating public debate about corruption. In some countries it has also led to substantive anti-corruption reform. It needs to be emphasized, though, that it can take some time for these actions to influence international perceptions, and be consequently reflected in the CPI.

38. Reactions by Governments, in particular those with bad scores in various years, placing a particular country at the bottom-end of the rank-list, were often rather strong[11].
    Peter Eigen, the TI Chairman noted already in 1997 that ‘the CPI has political impact’.

‘While some governments rejected the implicit criticism out of hand, others have acted on it, initiating reforms to strengthen their integrity systems’.

39. Some donor-governments [themselves always ranked high, see below 50-54, Double standards, why this is a dubious result] use a low ranking of an aid-recipient country on the CPI-list, as sufficient argument to stop a donor relationship with that developing economy and to withhold aid-money. For instance, the Dutch withheld aid to Cameroon in 1998 when the country ended as number 85, and last one, on the list.

40. TI distances itself from such use of the CPI in the context of aid-conditionality, an application to which the CPI is wholly unsuited. TI could louder reject this misuse of its findings. A strong [but also remarkable] reason given for this caution was that

‘a high degree of observed corruption may [….] reflect a high standard of ethics and a rigid application of rules rather than a high degree of real misbehavior.’

41. In other words, the stronger a government fights corruption, the clearer becomes the presence of corrupt behavior to everybody, and this may mislead observers. Should this warning not be taken more seriously by TI itself?

42. I assume that also TI realizes that this reason for a low/bad score is quite exceptional and that giving this reason as a possibility, is disastrous for the TI/Lambsdorff research approach to use ‘perceptions of corruption’.
    The corruption-perception may reflect a high standard of ethics!
    Moreover, is not repeatedly said that we do not have the facts available to establish real scores about the prevalence of corruption? Why than write about ‘observed corruption’?

43. In the following I will try to list and to review the available information, try to find reasons for changes in the methods and the sources, ask questions useful for improvement and take an overall critical stand, as the main questions should be:
    – Does the yearly list of TI reflect the reality in the world?
    – Does the CPI educate, in other words: are countries becoming less corrupt because of blame when found on the lower ranks of the ladder and can they climb the ladder by fighting corruption?

How do you measure the degree of corruption?

44. The CPI-1996 report stated already very briefly the methodology.
    The TI-Corruption Ranking is a joint initiative of Transparency International and (then) Göttingen University, and is updated once a year. The major operational work is conducted by Prof. Dr. Johann Graf Lambsdorff [address see footnote 2 on the front page] and a research team at the University of Passau, since he moved there from the Göttingen University.

‘The index is not an assessment of the corruption level in any country as made by TI or Dr. Johann Graf Lambsdorff. Rather it is an attempt to assess the level at which corruption is perceived by people working for multinational firms and institutions as impacting on commercial and social life. To the extent that any country has a problem with its ranking, this lies not with the index but rather with the perception that businessmen polled apparently have of that country. Their perceptions may not always be a fair reflection on the state of affairs [italics by MvH], but they are reality.’ ‘Differing perceptions may be due to a change in awareness rather than real corruption’.

45. The scores obtained from the participants in various studies are being used by Lambsdorff/TI to produce a composite index.
    The CPI reports claim that

‘in combining these different sources into a single index further serves to strengthen the index with respect to validity and reliability’.

This claim appears in all reports (with some variation in the wording) accompanying the yearly CPI-publication.

‘The greater the number of surveys in which a particular country has been included, the more reliable the assessment’, [italics by MvH]
is already written in the ‘Explanatory note’ attached to the CPI-1995 and also repeated often.

46. The foregoing paragraph signals already in two statements the major shortcomings of the CPI over the years:

• ‘Perceptions may not always be a fair reflection on the state of affairs’
  [in other words the qualifications given by the businessmen and analysts in the various surveys used for the TI-CPI may not reflect the reality in a particular country].
• Multiplication of surveys would enhance reliability of the composite index.

47. They correctly draw attention to the fact that the ‘perception of corruption’ does not necessarily reflect the ‘real level of corruption’. Furthermore, that ‘a low performance in the index does not make corruption a justifiable business. Also in badly perceived countries is corruption sensed to be an illegitimate conduct leading to public dismay.’
48. One more caution was built into the reporting of the CPI which could be repeated in any one of the following years. It deals with the qualification as ‘most corrupt country in the world’. For instance in 1997, this happened to be Nigeria that qualified as nr. 52, and last one, of that year’s ranking list.
49. The pertinent and logical answer that year and all following years was that the number last (in this case Nigeria) is only, as perceived by business people, the most corrupt country which has been included in the TI-list. Some other countries, not included because of lack of information, could have scored worse than Nigeria, had they been included.

Double standards

50. A caution missing here, which I found in another Report by TI, the Bribe Payers Index 2006 (BPI, p.10) is the variation in behavior of the same people in different life circumstances. What does the BPI 2006 say?

‘Perhaps the most important finding regarding the comparison of assessments by respondents in LICs [Low Income Countries] and OECD countries is the apparent double standard employed by foreign companies in the two groups. While the scores for companies from the majority of countries tend to be considerably higher in the OECD than in the full sample, their performance falls when looking at scores in LICs. Thus it would seem that many foreign companies do not resort to bribery while operating in the ‘developed’ world, where institutions are strong and there is a significant threat of legal retribution for illegal activities. However, in LICs, many of which are characterized by poor governance and ineffective legal systems for dealing with corruption, it appears that many companies resort to corrupt practices. The result is that the countries least equipped to deal with corruption are hardest hit, with their anti-corruption initiatives undermined. This helps trap many of the world’s most disadvantaged people in chronic poverty.’

51. After this statement should not the first question for LICs be: ‘who is corrupt?’, the country or the companies?
    And the second question: are not these respondents the same kind of respondents that answer questions about the perception they have of corruption?

    Is the perception of corruption domestically also influenced by differences between LICs and OECD-countries? What is it respondents observe and on what do they reflect? Is this different if they are asked about bribe paying in stead of about bribe receiving?
    Differs their perception if they are asked about other fraud or other shortcomings in any society? Is it not remarkable that so many of these lists of countries have all the same countries in the top quartile and the same in the bottom quartile of the list?

52. Do they observe and experience corruption, and do they reflect and report as their perception of corruption what they observe in their own environment? Are they more sophisticated looking at corruption elsewhere? In other words, do they perceive that it is their President who is corrupt accepting money from an oil-company, or is their observation that it is the Director of the foreign oil-company who pays, that is the corrupt man, company and country? The more so this becomes a problem if companies seem to behave differently in countries where the law reigns than where the law is absent as is observed in the Bribe Payers Index?

    Why do we find this observation only in the BPI-reports, and not in the CPI-reports?

53. An effort to explain different behavior by respondents comes from a paper by Jakob Svensson, Bribes, Lobbying and Development [12], in abstract:

‘Why are firms more likely to pay bribes to bureaucrats to bend the rules in developing countries while they instead lobby the government to change the rules in more developed ones?

Should we expect an evolution from bribing to lobbying, or can countries get trapped in a bribing equilibrium forever?

Corruption and lobbying are to some extent substitutes. By bribing, a firm may persuade a bureaucrat to “bend the rules” and thus avoid the cost of compliance. Alternatively, firms may lobby the government to “change the rules”. But there are important differences. While a change in the rules is more permanent, the bureaucrat can hardly commit not to ask for bribes also in the future. Based on this assumption, we show that (i) an equilibrium with corruption discourages firms to invest, (ii) firms bribe if the level of development is low, but (iii) they switch to lobbying if the level of development is sufficiently high.

Combined, the economy might evolve from a bribing to a lobbying equilibrium, but too large bribes may discourage the necessary investments for lobbying eventually to become an equilibrium. The outcome is a poverty trap with pervasive corruption. This poverty trap is more likely if penalties on corruption are large and the regulatory costs are high.’

54. Does this explain why North-Western Europe and like-minded countries elsewhere in the world always come high (corruption-free) in the CPI-reports (see my later discussion in this paper of the results (see para. 147-162) in the form of scores and rank-numbers presented by the CPI-reports) and lesser economically developed countries with weaker governmental institutions, come low(er) on those lists? We bend or change the rules, and they pay bribes?

Sources of information

55. Over the years, data for the successive CPIs came from 24 independent institutions and the composite indices made use of their 95 polls or surveys as they are mentioned in the yearly editions of the CPI-reports that can be found on the TI-website.
56. Some of these were used in a roll-over fashion, i.e. they were used for several consecutive years. In Attachment III with the names of the institutions and the titles of the surveys, this is made visible by giving the years in which particular surveys have been used.
57. For instance, for the CPI-1995 use was made of the World Competitive Reports of 1992, 1993 and 1994 issued by the Institute for Management Development as mentioned under no.1 WCR on the first row of the table in Attachment III. For the CPI-1996 from the same Institute its Reports 1993, 1994 and 1995 contributed data.
58. All counted together 95 surveys have been used. The reported available number of surveys per year was rather low in the years 1995-1997 (7-10) and higher in 1998-2005 (12-18).

Francophone and lusophone sources are missing

59. A very general remark regarding the full list of sources is that the francophone and the lusophone world of analyses are practically speaking missing. Of course, French and Spanish/Portuguese speaking countries are included in the various surveys, but the perception of everybody reading the CPI-reports cannot be another one than that the approach chosen appeals very much to those living, working and observing from an Anglophone background.

Available?

60. With regard to the availability, one question comes up immediately: what is the exact meaning of ‘available’? Looking at the table in Attachment III again, we see for instance that in 1997 for source no.1, WCR, not only the Report 1997 was available, as was reported in the CPI-1997 Report, and is indicated in the table, but also the Reports for 1995 and 1996, mentioned respectively under 1996 and 1998. Why have these data not been used in 1997?

    This has to be a rather pertinent question in the light of the remark already made in the CPI-1995 Explanatory note: ‘the greater the number of surveys included, the more reliable the assessment’. Or is this not true for all data available in 1997?

    When are a larger number of available surveys of importance for more reliability?

61. The same question can be asked e.g. for the reports mentioned under no. 2, PERC, for the years 1997, 1998, 2002 and 2003. Obviously, in each of those years more reports from this source were available, but they have not been used. Why?

Continuity and consistency?

62. Only two sources have been used over all the years 1995-2005, the WCR-World Competitiveness Reports of the Institute for Management Development in Lausanne, and the Asian Intelligence Issues of PERC-Political & Economic Risk Consultancy of Hong Kong.
63. If we exclude more early years as ‘try-outs’ and we look only at the years since 1999 than we see that for all following CPI-reports the continuity of sources improves as five sources are continuously used. Next to the already mentioned reports from Lausanne (WCR) and from Hong Kong (PERC), we see incoming the Global Competitiveness Survey/Report of the World Economic Forum (Switzerland), reports from the Economist Intelligence Unit (EIU, London) and of Freedom House (New York).
64. All other 19 institutions figure only in a limited number of years. Some (4) began to be used from CPI-1998 onwards, some (5) from CPI-1999, some more provided data for later CPI-reports. However, several sources/institutions are not delivering any more data for later issues of the CPI.
65. This is equally important because ‘the perception of a country’s integrity may also vary with the availability and choice of sources’, quote from the CPI-1997 report, paragraph ‘The General Approach’.
    Galtung (see footnote 4) remarks that TI in the 2003 CPI press release, states that

‘Norway, France and Germany improved their scores in recent years’
and he continues:

‘a significant percentage of the upward trend of the scores for these and other countries would be accounted for by the exclusion of PRS [Political Risk Services, used from 1996 till 2000, see attachment III] from the data sources’.

66. Lambsdorff himself is aware of the consequences of discontinuity in the sources.
    In the CPI-2005 Methodology (p. 3) he remarks:

‘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 comparisons of a country’s score do not only result from a changing perception of a country’s performance but also from a changing sample and methodology. Old scores drop out of the index and new sources enter, disturbing the consistency of the assessment. The index primarily provides a snapshot of the views of businesspeople and country analysts, with less of a focus on year-to-year trends.’
[italics added by MvH to draw more attention].

67. Is not this a self-killing argument by Lambsdorff? Is not the purpose of yearly assessments to have a yardstick that can measure whether the perception of the corruptness of a country over the years is improving or deteriorating? Not for the researcher who claims that a higher score can result from change(s) in the methodology. If you cannot trust anymore that a score for a country that goes up over the years, means that the perception of that country’s state of corruptness is improving, what is it you are assessing?
68. But the outside world does! When the new CPI is published, what is it all journalists are looking for? What happened to country X and Y? Who is at the top of the list? Who is at the tail end? Which countries move up the list and which ones go down?

    Lambsdorff himself lists countries that ‘improve’! To confirm that this may be caused because of change in the sources that have been used for the amalgamation of data disqualifies the CPI. The more so, if the discontinuity of data sources could be less than is the case.

Why discontinuity in the sources?

69. The reports do not give an answer to that question. Why was the EIU not yet included before 1998? How is it justified that in the CPI-1996 use is made of ‘one small survey by Peter Neumann’ published in 1994 in the monthly German magazine Impulse. Did it just happen to be around in 1996? Why have Neumann’s 1994-data not been used in the CPI-1995? Why did they not roll-over in the CPI-1997?

70. The same question can be asked for the World Business Environment Survey published yearly by the World Bank and included as a source dating from 2002 only up to the CPI-2003. The World Bank BEEPS-Business Environment and Enterprise Performance Survey of 2002 are used for the CPI-2003 and CPI-2004. The same surveys for 2004 (issued 11^th February 2005, in any case in time for CPI-2005) and 2005 (issued July 2005) are not used for CPI-2005. Why?

71. Some may argue that the correlations calculated by Lambsdorff prove that his scores and therefore also his yearly ranking of countries is more valid and reliable than I observe. With regard to his statistical work, my point is that if your basic data are doubtful, whatever statistical work you do, this does not improve their quality. Mud remains mud! Some do call this ‘statistical manipulation’.

Africa

72. Another remarkable report is missing. For the CPI-reports 1999-2002 use is made of the World Economic Forum’s Africa Competitiveness Reports 1998 and 2000. According to the methodology chosen, a survey is used in not more than three successive years. So it is correct that the Africa Competitiveness Report 2000 is used in 2000, 2001 and 2002, and is not used any more from 2003 onwards. Its value had expired. But why is the continuity not restored since the (third) Africa Competitiveness Report 2004 was published? Its data could have been used for the CPI-2004 and 2005?

Where is the Arab world report?

73. Next to the World Competitiveness Reports and the Africa Competitiveness Reports was also twice issued the Arab World Competitiveness Report, in 2002-2003 (first edition) and in 2005 (second edition) with in that year the following ranking for 12 countries:

Table 1, Arab-world ranks

Three Arab countries present in the CPI-2005 list: Kuwait with the CPI-rank 45, Syria 76, and Libya 122, are not included in the 2005 Arab World Competitiveness Report.

74. For ease of comparison, the CPI-2005 rank-numbers have been added. Between the rank orders from the two reports there is not much difference. The first six from the Arab list appear more or less in the same order in the CPI-2005 (and vice versa). The same is true for the second group of six countries. It appears that not much is lost by omitting the use of the Arab World Report, but does this answer sufficiently the question why this report has been neglected?

Should source reports be recent?

75. No explanation is given about the use in the CPI-1995 report of a 1980 survey from Business International. [It is unclear what this name stands for. Is it the Business International subsidiary of the Herald Tribune?]. Why was this outdated source used? The very early date for this study on comparisons of corruption and integrity in the world, [as most interest in this subject dates back no longer than to 1990], would have justified some more explanation why this survey has been used? Was it because of lack of other sources? Here certainly we cannot speak of ‘up-to-date’ information.
76. Several of the CPI-reports proudly mention that their data are very fresh. The CPI-1997 report is very explicit in stressing that all data come from reports of 1997, but one that covers data from 1996/97:

‘The 1997 Corruption Perception Index CPI is the most up-to-date and reliable index on corruption so far. Using only data from 1996/7 its figures are more precise than in previous years when the CPI covered a period of several years.’

77. However, the CPI-2005 report explains the contrary view as the best one! The data should not be only the most recent ones:

‘the 2005 CPI combines assessments from the past three years to reduce abrupt variations in scoring that might arise due to random effects. Some sources, such as II [Information International, Beirut] provided only one recent survey. Others such as WEF, IMD and PERC conducted annual surveys between 2003 and 2005, which are all included.’

78. Over the years there is no rigidity what sources to use. All sources that is available? Only the most recent ones? Or, all from the same year as the CPI of that year? Or roll over three years with reports available, or make a choice of reports available?

Should CPI-reports be transparant?

79. For the CPI-1996 and CPI-1997 use has been made of

‘incoming replies to the internet service of Göttingen University which gave contributors the possibility for anonymous contributions and also directly approaches employees of multinational firms and institutions.’

80. The initiative to answer came sometimes from the respondents (‘anonymous’), and sometimes was asked (‘direct approach’).
    How many in each category? Who made the selection of those to be ‘directly approached’? Did researchers see differences in results from these two categories of respondents? Did everybody who was approached, cooperate? How many, what percentage did not? Was factually established that the internet-respondents were international businesspeople or journalists or experts? Was the internet service closed down? Why was it not used in later years?

81. For the CPI-2003 and CPI-2004 an undisclosed multilateral development bank (row 22 in Annex III) enters as a source of data.

82. Luckily, it was concluded to discontinue the internet service of Göttingen University (79) and the use of data from the undisclosed multilateral bank (81). In these cases that was the best thing to do.

83. In the Framework Document by Lambsdorff published with the CPI-1999 [on page 6] is remarked that

‘a number of other possible sources has been considered for inclusion and rejected because they mixed corruption with other issues such as xenophobia, nationalism or related to changing corruption rather than levels of corruption.’
An attached list of those rejected sources, with the rejection motivation, would have been useful, also in other years.

Index Advisory Committee

84. In several CPI-reports the following lead can be read. This one is taken from the CPI-2006:

‘The CPI methodology is reviewed by an Index Advisory Committee consisting of leading international experts in the fields of corruption, econometrics and statistics. Members of the committee make suggestions for improving the CPI, but the management of TI takes the final decisions on the methodology used.’

85. The names and functions/qualifications of the Members of this Index Advisory Committee are not given in the CPI-reports. In earlier years, the name of this Committee is given as ‘Index Steering Committee’. Neither is indicated what the criteria are for the management of TI either to follow or not to follow the ‘suggestions’ of the Committee. Has ever TI-management not followed a ‘suggestion for improving the CPI’? About what subject, source, data, or whatever else?
    I ask these questions as I could not understand why persons, with qualifications as mentioned by TI, would accept to sit on such a committee, and to make suggestions which subsequently might not be followed by management. I also wonder whether TI should not be more transparent about this advisory committee that deals with the most important tool in use by TI worldwide.
86. As I found this unsatisfactory I asked the TI-Secretariat additional information. Dr. Robin Hodess, TI Director of Policy and Research answered:

‘The Index Advisory Committee (it used to be called a Steering Committee but the name was changed in the past year or so) is managed here at TI S.
The idea behind this is very simple: we have a group of expert volunteers (from academia above all, but also from relevant private sector and international organisations) who contribute to our discussion of TI indices. This goes beyond the CPI, and extends to the BPI, Barometer and other current and planned TI measurement tools. The group meets by telephone or in person, as is possible/necessary. Whether what the group does is called steering or advising, TI always takes final decisions about its tools and how they will be carried out, with careful consideration of the points raised by the committee.
The members of the committee have changed over time (it was set up in the late 1990s, long before I became responsible for it), as some have ceased active participation. There may have once been a list of members on the website, but it probably dates to before the website was redesigned in 2005. Right now, we do not have a current list of members posted online, since those who participate have not necessarily chosen to participate on the basis that we make them open to public queries. If you have anything you would like to share with the committee, I would be happy to do so on your behalf. No meeting of the committee is planned at present.’

Other problem(s) underlying the CPI

87. I quote from the CPI-1995 reporting that the focus of TI is on ‘international business transactions’. But the Introduction of the Methodology of CPI-2005 begins with: ‘The goal of the CPI is to provide data on extensive perceptions of corruption within countries’. Why is it then mostly businessmen that are asked their perceptions?
88. The TI ranking is

‘an assessment […] in which existing polls of international business interests and financial journalists have been analyzed and collated. It is thus a picture of how international business sees the levels of corruption in the 41 countries ranked in the survey.’

89. An obvious question is: do businessmen have the ability to judge a country in expressing their perception of the state of corruption in such a country? Is their personal experience not rather limited to only a few countries and within those countries to particular trades, companies and individuals? How can we trust that their observations and/or experiences, by definition limited in time and space, from which they draw their perceptions of the level of corruption in a country, are valid ones for a whole country?

90. The CPI-1997 mentions: ‘the perception by contributors may refer back to experiences made long before the assessment is made’.
    Yes, of course, and nevertheless all experiences from long ago and from more recent days are amalgamated into one result for a particular year’s Index? What is than the importance of working with most recent surveys?
91. And if respondents are capable to score particular countries in which they live and/or work, why would that capacity qualify them to judge countries of which they may have heard only?

Respondents

92. Behind the institutional reports from which the data are collected to establish the composite index which is the CPI of a particular year, hide the people that give their perceptions and answer the questions of the researchers. Who are they? Do they represent ‘the people’? Do they know what happens in a particular country? Do they have sufficient knowledge of, and insight in what happens in society, a broader experience than what they observe as businessmen?
    Obviously, their judgments are mostly about countries in which they do not live and work (see number of respondents and countries in the tables 2 and 3 below).

93. The institutions involved in this ‘poll of polls’ of 1995 are located in Lausanne, Hong Kong and New York. This alone should worry bystanders about the picture they deliver ‘on average’. Obviously, the index deals with domestic corruption. The intention is expressed ‘to broaden the scope’ and ‘to include assessments of the levels of corruption practiced by businesses from various major trading countries in their dealings abroad’. This indeed comes some years later in the form of the ‘Bribe Payers Index’ [published by TI in 1998, 2002 and 2006, not included in this paper].

94. In the CPI-1996 report the respondents are classified as

[…] ‘people working for multinational firms and institutions as impacting on commercial and social life’ … [and a few lines later] ‘… the perception of businessmen’.

95. As the CPI is a composite index based on other research results in the field of corruption, TI claims in the CPI-1996 report that

‘the sources we included put a high effort on comparative judgments’,

and TI continues with the statement that:

‘People working for internationally operating firms and institutions are able to provide a comparative judgment and apply the same internalized definition to different countries. In this perspective a cross-country approach can contribute to a valid comparison of real degrees of corruption.’

Although this is a quote from the report, it cannot be serious that is seriously proposed that by assembling ‘perceptions’ [here is used the word ‘judgment’], the ‘real degree of corruption’ [in other words ‘facts’] can be found.

Table 2, Who was asked in 1996?

96. I question also the rather narrow definition of ‘people working’ as executives, bankers, managers, staff, embassies, Chambers of Commerce [see table 2]. What are the perceptions of people working at the production lines and in the administration, bookkeeping and accountancy of the firms? Are they not ‘able to provide a comparative judgment’? Over the years, whistleblowers have proven that they are quite capable to do so. Do we have to understand that the men and women on the production lines and those in administrative functions are represented by the views of ‘staff’ of risk assessment companies and by the ‘anonymous callers’ on the hotline of Lambsdorff himself in his university? The CPI-1997 says: ‘ordinary citizens’ are not included ‘as they do not fit our definition of corruption’ (p. 11), and ‘their results are of lower quality’!
    These shortcomings are repeated in later CPI-reports.

97. It goes too far in detail, to go over all CPI-reports of all the years since 1996, and to find out who the respondents are that ultimately delivered the information that now circulates as Corruption Perceptions Indices.
    I will restrict myself now to the CPI-2005 Methodology Report, reproduce what is said in that report and try to conclude what is different from the CPI-1996 and what is acceptable for good research.

The respondents in the CPI-2005 Methodology report

98. In the Methodology written by Lambsdorff for the CPI-2005, a bit more is explained than in 1996 about the sources used for constructing the CPI-2005 and the people that was involved in the collection of the data given in the 16 surveys supporting the CPI-2005. [The text is quoted from the source].

‘While the sources all aim at measuring the extent of corruption, the sample design differs considerably. Basically, three different types of samples are used.

A first group of sources, namely CU [no. 7 in the following table 3], EIU [8], FH [9], MIG [11] and WMRC [16], assemble the perceptions of non-residents, turning in their experienced perception with regard to foreign countries. These assessments are carried out by respondents from developed countries of the western hemisphere such as North America and Western Europe, often supported by networks of local correspondents.

A second group of sources assembles also the perceptions of non-residents, but these respondents are largely from neighboring countries, which are often less developed countries. This year, only the data by II [10] were of this type – in previous years further sources were available that adhered to this approach.

There is an advantage to perceptions vis-à-vis foreign countries because they are not vulnerable to a “home-country bias”. Such a type of bias would be relevant if respondents assess their home country purely according to local standards. Such a standard would be problematic because it can differ from one country to another, impairing the validity of cross-country comparisons.

A third group of sources, namely IMD [1-2-3], PERC [4-5-6], UNECA [12] and WEF [13-14-15], gather assessments made by residents with respect to the performance of their home country.

These respondents are partly nationals but sometimes also expatriates from multinational firms. While such data might be susceptible to the aforementioned “home-country bias”, they are not susceptible to introducing an undue dominance of “western business people’s” viewpoint. Such a viewpoint would be inadequate if foreigners lack a proper understanding of a country’s culture.’

Are the samples representative?

99. In all surveys it is people that are willing to express their perceptions, otherwise there could be no survey. In many countries this can be a problem as the powerful that enter into corrupt dealings have reasons to keep corrupt activities secret. Generally they are not pleased by disclosures of unwanted corrupt behavior. They cannot be pleased either by insinuations or reflections on their honesty. The respondents willing to cooperate in the various surveys may constitute biased samples as they are persons not afraid of the danger for them in disclosing one’s perceptions.

100. There is no indication in the report that there are systematic differences between perceptions as collected from in-country local residents and from in-country (quasi-) permanently living and working expatriates. It is most likely that there are. Maybe it is impossible to discover this difference as being an expatriate does not explain everything. For instance part of the expatriates might be fluent in the local language and another part might not. It is unlikely that such a difference would have no consequences for the development of one’s perceptions about corrupt behavior, for information that can be obtained, and for the perceptions being formed. It might be also different for expatriates working in a local company or in an international company?

101. Lambsdorff indicates already that ‘local standards’ might bear an influence on the genesis of ‘perceptions’. Indeed, the question is what is considered to be corrupt behavior and what is not? How does this influence the genesis of perceptions of corruption?
     This is the more so a problem because the various CPI-reports are not clear about their definition of corruption and in any case the subject of study is – between them – often quite different..

102. Finally, let us have a look at the details that are given in the CPI-2005 report as description of the persons involved in the various surveys.
     In my view those surveyed share some characteristics that are most important for the development of their perceptions of corruption. It is not that they are all white, or all from Northwestern Europe, or all (maybe some women?) male.
     What we may assume is that they are all working at higher and managerial levels in their companies, well-educated and fairly well enumerated. Probably they are (nearly?) all male and between 25 and 50 years of age. Let me make the educated guess that most of them have an MBA.

103. Missing are the women, the consumers, the old, the young, the sick, the poor, the powerless, the ones that suffer under corruption. Why no trade-unionists? Why no investigative journalists?

104. The issue is not as is stated in the CPI-1997 that this index is ‘less biased against developing countries’. It is biased in more than one direction. And on top of this it is biased in as far as western-world (Europe, USA/Canada, Australia/New Zealand) representatives in the samples are dominant over third world representatives.
     In this respect the real bias is that those among the samples coming from the countries in the so-called rich world resemble very much the other ones coming from countries in Africa, Asia and South-America. (Nearly?) all of them share predominantly the same business education and business-world background. At best, they are the researchers of corruption, at worst they are the ones that corrupt. Should they decide what the perceptions of corruption are?

Table 3, Who was asked in 2005?

105. Lambsdorff makes very clear in his reporting on corruption as published in the yearly Corruption Perceptions Indices that the samples chosen to present their perceptions of corruption, cannot do this with any legitimacy as they do not represent ‘the people’.

Definition of ‘corruption’?

106. It is odd, that nowhere in the documents surrounding the 1995 Corruption Index publication (Press Release, Explanatory Note, Rank-list of the 41 countries, Note for Editors, see the TI-website[14]), you can read what definition is used for the concept ‘corruption’.

107. Neither is the concept ‘corruption’ described, as it was understood by the businessmen and other respondents of the various sources, in making their judgments. What is said is that they were ‘polled in a variety of contexts’. In these documents ‘kickbacks’ and ‘extortion’ are used as synonyms for corruption. The word ‘integrity’ is not really discussed in these early documents. This word is used to explain the index as an

‘overall integrity ranking (out of 10). A ten equals an entirely clean country while zero equals a country where business transactions are entirely dominated by kickbacks, extortion etc. No country scores either ten or zero.’

108. In later issues of the CPI the definition-question is tackled. In 1996 [and again in 1997 in exactly the same wording] the question is specifically asked:

‘How do you define corruption?
Corruption is the misuse of public power for private benefits, e.g. the bribing of public officials, taking kickbacks in public procurement or embezzling public funds. The external surveys we included were mostly very close [emphasis by MvH] to this definition. The ranking tries to assess the degree, to which public officials and politicians in particular countries are involved into corrupt practices.’

109. In later years the first line in this definition is changed into ‘Corruption is the misuse of entrusted power for private benefits’, in stead of ‘public’ power. This is an important change as the scope of the definition includes now not only the public sphere of politicians and administrators, but also the private sphere of businessmen and their interactions.
110. The 1996-CPI publication discusses the definition a bit in a paragraph called

‘How can you compare between countries?’

‘Corruption is defined by some researchers as a particular public reaction to political/administrative behavior rather than as an illegitimate act as such. Looking for appropriate definitions, this approach assigns a much more active role to the public perception and reactions towards corruption. A high degree of observed corruption may in such an approach reflect a high standard of ethics and a rigid application of rules rather than a high degree of real misbehavior. A cross-country comparison of levels of corruption would hence not be applicable since the underlying standards of ethics may not correspond between countries. However, the sources we included put a high effort on comparative judgments.
People working for internationally operating firms and institutions are able to provide a comparative judgment and apply the same internalized definition to different countries. In this perspective a cross-country approach can contribute to a valid comparison of real degrees of corruption.’

111. The definition of ‘corruption’ as used can be deducted from the questions asked in the various polls. In other words, what subject was considered by Lambsdorff sufficiently covering the ‘corruption’ issue?
     Most of these questions refer specifically to the misuse of public power. One at least deals expressly with public and private business (no.7), while PERC (no. 4, 5, 6) and DRI (8) give no indication.

     Table 4, What was asked in 1996?

112. Not only is the definition as used in the CPI of importance. In order to collect comparable data, also all sources should use the same definition. This appears to be difficult, although all sources generally apply a definition of corruption such as the misuse of public power for private benefits, e.g. bribing of public officials, kickbacks in public procurement, or embezzling of public funds. Each of the sources also assesses the “extent” of corruption among public officials and politicians in the countries in question.
113. The later shift from the word ‘public’ to the word ‘entrusted’ in the definition of corruption as used in the various CPI-reports is remarkable, as some of the sources are continuous, see for instance in Attachment III the numbers 1 and 2. They clearly do not change their definitions [they changed questions going through the years].
114. With the focus on problem, prevalence, pervasiveness, commonality, extent and the number of cases of corruption, all these sources refer to some kind of ‘degree’ of corruption, which is also the aim of the CPI.
115. The CPI-1998 report continues with the observation that this common feature of the various sources is particularly important when one considers that corruption comes in different forms.

‘It has been suggested in a multitude of publications that one should endeavor to differentiate between petty and grand corruption, frequency and price of corruption, etc. Yet, none of these special forms of corruption have been found to dominate in one source and be less important in another. The sources can be said to aim at measuring the same phenomenon.’ [This cannot be said based on the information as reprinted from the reports in the following table 5].

116. The discussion is continued in the CPI-1999 report:

‘All sources generally apply a definition of corruption such as the misuse of public power for private benefits, e.g., bribing of public officials, kickbacks in public procurement, or embezzlement of public funds. Each of the sources also assesses the “extent” of corruption among public officials and politicians in the countries in question.’

117. However, the same CPI-1999 (pages 2 and 3 in the Framework Document by Lambsdorff, October 1999), says

‘the precise legal definition of bribery and corruption can be different in each national context, the differences drawn between bribery, embezzlement and fraud may be troublesome and the statistical methodology of counting and aggregating used in each national agency can differ considerably from that used elsewhere.’

118. How to assess and to aggregate what is defined in different ways? Is everybody talking about the same subject? Is everybody interpreting the same subject? What is in the minds of the respondents, their experience, or what they think is the socially acceptable definition of corruption among their peers?
119. That same report continues with:

‘Apart from this, countries as Singapore and Hong Kong have extremely high per capita conviction rates for bribery. This lends itself to the conclusion that the data are to a large extent determined by the effectiveness and capacity of a country’s judiciary in prosecuting corruption. High levels in this case indicate the success of anti-corruption initiatives rather than high levels of actual corruption. As such problems commonly arise with objective data, international surveys on perceptions serve as the most credible means of compiling a ranking of nations.’

120. I do not understand this exclamation. I could have understood this conclusion of the researcher, if Singapore and Hong Kong – because of the reason he gives – had landed at a low score and at the bottom of the CPI-1999. But that is not the case.
121. If it is true that both countries have high conviction rates (I do not have these data at hand) and if because of that there is less corruption in those countries, is not that the explanation why they score extremely well in the CPI-1999, respectively with scores of 9.1 and 7.7, resulting in rank-numbers 7 and 15?
     I don’t see the problems mentioned by the researcher that arise with the objective data.
122. Is it not that all researchers of the corruption issue in the world would love to have facts in hand, and that all do their utmost best to get facts so that corruption is no longer the unknown factor influencing governmental and business decisions? And here comes a researcher with the proposal to forget about facts as ‘perceptions serve as the most credible means of compiling a ranking of nations’? I cannot grasp this.
123. Facts are rejected because they conflict with perceptions. Is seriously meant to say that what people imagine as existing in their socio-economic environment, is more credible than facts that have been researched and judged? Why would an, in my view biased, choice of respondents, deliver better results? Is not the absence of known facts about corruption the major complaint of all researchers of this issue? And here facts are explicitly rejected? Is this science?
     Compare what was the subject of questions in 1996 (table 4) with what was asked and researched in 2005 (table 5).

Table 5, What was asked in 2005?

Sources for CPI-2005 compared to CPI-1999

124. In five cases the same institution was also used as a source for the CPI-2005 report.
     All questions of 2005 as reprinted below come from the document ‘The methodology of the TI Corruption Perceptions Index’ by Professor Dr. Johann Graf Lambsdorff, September 2005, issued with the scores and rank-list of CPI-2005.
     It should be noted that questions asked are not the same as in 1999, and may convey another message. Nevertheless they are used in establishing the composite index and claim to be comparable.

• The IMD (Institute for Management Development, Lausanne) asks respondents in 1999 to assess whether ‘Improper practices (such as bribing and corruption) prevail or do not prevail in the public sphere.’

2005: IMD asks respondents to assess whether ‘bribing and corruption prevail or do not prevail in the economy’. [Not explained are the purpose and the consequences of the shift from ‘public sphere’ to ‘the economy’.]

• The PERC (Political and Economic Risk Consultancy, Hong Kong) asks in 1999: ‘To what extent does corruption exist in the country in which you are posted in a way that detracts from the business environment for foreign companies?

2005: PERC asks ‘to rate on a scale of zero to 10 how bad they considered the problem of corruption to be in the country in which they are working as well as in their home country.’ [The difference with the 1999 question is quite large.]

• The PRS (Political Risk Services) determines a variable ‘Corruption in Government’ and assesses the overall spread of corruption. [not included in 2005.]
• GI (Gallup International) asks ‘From the following groups of people, can you tell me for each of them, if there are a lot of cases of corruption given, many cases of corruption, few cases or no cases of corruption at all. The following groups were considered for the CPI: politicians, public officials, policemen and judges. [not included in 2005.]

  · The WEF (World Economic Forum) asks in 1999: ‘Irregular, additional payments connected with import and export permits, business licenses, exchange controls, tax assessments, police protection or loan application are common/ not common.’

2005: WEF asks ‘in your industry, how commonly would you estimate that firms make undocumented extra payments or bribes connected with 1. Exports and imports, 2. Public utilities (e.g. telephone or electricity), 3. Annual tax payments, 4. Public contracts, 5. Loan applications, 6. Influencing laws and policies, regulations, or decrees to favor selected business interests, 7. Getting favorable judicial decisions.’
Each of the seven categories could be answered with ‘common’ or ‘never occur’. TI took from these questions the simple average as determined.

• The WB/UB (World Bank and University of Basel) asks two related questions with respect to corruption: first, ‘Please judge on a six point scale how problematic [corruption is] for doing business’; second, ‘It is common for firms in my line of business to have to pay some irregular ‘additional payments’ to get things done. This is true always, mostly, frequently, sometimes, seldom or never.’ [not included in 2005.]
• The EIU (Economist Intelligence Unit) defines corruption as the misuse of public office for personal (or party political) financial gain and aims at measuring the pervasiveness of corruption in 1999. Corruption is one of over 60 indicators used to measure ‘country risk’ and ‘forecasting.’

2005: EIU asks ‘to assess the incidence of corruption and defines corruption as the misuse of public office for personal (or party political) financial gain.’
[Assessing the ‘incidence of corruption’ is quite something else than ‘measuring the pervasiveness of corruption.’]

• FH (Freedom House, Nations in Transit) and CEER (Central European Economic Review of the Wall Street Journal) in 1999: determine the ‘level of Corruption’ without providing further defining statements.

2005: FH asks ‘to assess the implementation of anticorruption initiatives’.

• The ICVS (International Crime Victim Survey) asks in 1999: ‘In some areas there is a problem of corruption among government or public officials. During 1995, has any government official, for instance a customs officer, police officer or inspector in your own country, asked you or expected you to pay a bribe for his service?’ [not included in 2005.]

125. The question remains open what kind of perceived corruption is meant by the respondents? Is it corruption at home (domestic corruption)? Corruption in a country abroad where they happen to live as expatriates but have a good local knowledge? Is it corruption abroad of which they have knowledge, e.g. as analysts for a surveying institute like the EIU? Or is the subject ‘international corruption’, buying and selling of entrusted power for private gain in another country?

 
RESULTS[16]

More South – more corrupt?

126. It is common knowledge – erroneously, should be added immediately – that the more South one goes, the more one will find corruption in business and daily life.
     In the North of the Netherlands (dominated by Calvinism from the 16^th till the 20^th century), the belief is generally held that the South (dominated by Catholicism in the same years) is more corruption-stricken than the North. In Dutch public opinion is likewise held the belief that officials and politicians in Belgium and France are more corrupt than in the Netherlands. In all of Northern Europe is held a similar belief with regard to Southern Europe. Do indeed people in Italy consider Milan as less corrupt than Naples?
127. Of course, along the same lines of thinking, many in the North agree that Africa is more corrupt than Europe. Widely held is the opinion that it is quite natural that Africans are more corrupt than Europeans, as this is part of their culture, while corruption is supposed not to be part of the European culture!
128. If we look at the American continent we see a similar geography. Canada scores high (i.e. corruption-free) on the CPI of TI: the country never ranked lower than place number 14 in 2005 (with a score of 8.4), all previous years it ranked between number 5 and 12 (scores of 8.87-8.5). Geographically next in southern direction comes the USA, in scores and ranks every year lower than Canada: with ranks 15-19 and never a score above 7.8.
129. If we go again more south and look at the scores and ranks of Mexico, we see the same phenomenon: worsening figures. Mexico’s rank-number is 66 in 2005 with a score of 3.5. Mexico never had a higher score than 3.7 in 2001 and 3.6 in the years 2002-2004. Ranks reached were numbered 32 to 66. (The places 32 in 1995 and 38 in 1996, Mexico thanked to a lower total number of competing national scores).
130. Transparency International seems to deliver proof that the thinking is correct that the further south a country is geographically located on the globe, the more it is corrupt. The TI yearly publication of rank-orders and scores of countries with regard to the perception of prevalence of corruption proves this hypothesis. Does it?
131. Before we go any further it is good to have another look at scores and ranks and how they are influenced for any country included as well by the results obtained for other countries as by the entry of newcomers on the list that may push earlier rankings downwards.

To begin with some examples

132. Indonesia ranked as number 41, and last one, in 1995. The country was ranked as number 140 in 2005. Did it worsen?
     If we go over the rank-lists of those two years than we see that the country’s rank sunk to number 140 in 2005, because 100 new countries entered the list in the course of the years and all of them got ranked ahead of Indonesia. The resulting place in the rank-order would have become number 141, except for Pakistan, that ranked in 1995 as number 39, and finished in 2005 below Indonesia as number 146.
133. Certainly, it must have been disappointing for the Government and the people of Indonesia that despite an improvement in the score-list showing that in fact Indonesia improved from a score of 1.94 in 1995 to a score of 2.2 in 2005, this was not reflected in a better position in the ranking of nations.
134. If we do the same exercise for China, than we see a slightly better result. Five of the countries that had been better classified than China (nr. 40) in 1995 (Argentina 24, India 35, Philippines 36, Venezuela 38, Pakistan 39), sank below China in 2005 (respectively 98, 92, 124, 136, 146) and only 43 of the newcomers got in the rank-list ahead of China, lowering China not more than from rank 40 in 1995 (40 plus 43 newcomers minus 5 ranking lower) to rank 78 in 2005. All countries that ranked better than China in 1995, ranked also better in 2005.
     But also China improved on her score in between those years from 2.16 to 3.2.
135. Below, in para.163-174, CPI high ranking dominated by rich countries, we will see that also in the higher echelon of the nearly corruption-free countries (if we follow the ranking for the perceptions of this phenomenon) interesting changes occur in the rank-order. The Netherlands climbed from being number 9 with a score of 9.0 in 2002, to become number 7 a year later, although her score lowered to 8.9. Nevertheless the higher place in the rank-order was hailed in the Dutch media as a great achievement. The United Kingdom scored 8.7 in both years 2002 and 2003, but was number 10 in 2002 and number 13 in 2003, a defeat in the fight against corruption?
136. All data taken together we see already that we have to be careful in drawing conclusions from the rank reached by a country on the yearly list of Transparency International made by Professor Lambsdorff. The scores need to be considered simultaneously, as they are the basis for the ranking.
137. ‘As pointed out repeatedly’, I quote Prof. Lambsdorff, ‘the time series value of the TI data is distorted because of the annual changes in the composition of sources.’ First question, why then these changes? Second question, if sources used to establish a score on prevalence of corruption, why should these scores not be comparable if the sources are reliable?

Worldwide corruption increased?

138. The total value of bribes paid each year as published by the World Bank rose between 1994 and 2004 from 50 billion US dollars to 1.000 billion (one trillion) US dollars. One would expect that this has an impact on the scores and rank-order of countries in the yearly CPI which is not the case. This can only be correct if the 20-fold increase of the total world-figure translates itself in also a 20-fold increase per country, in other words the spread of corruption over the world remains unchanged. This comes close to a miracle! This does not seem to be very likely. The estimate of the total amount of corrupt money paid/received per country, as well as the perceptions per country, were equally wrong for all countries?
139. The ‘50 billion’ was repeatedly used by the World Bank and in her footsteps by TI in the late ‘90s. Later, early in the first decade of this century ‘400 billion’ was used as a reliable estimate or approximation. The ‘1.000 billion US dollars’ was published for the first time on the internet website of the Bank in the monthly Newsletter dated 8/3/2004, under the title ‘the costs of corruption’. The journalistic form chosen for the publication of this latter figure was an interview with Dani Kaufmann, the World Bank Institute’s director for governance. That Newsletter also mentioned an article by Peter Eigen, Chairman of TI, titled ‘Grease Trade’, in which the same figure of one trillion US dollars was used.
140. Understandably, I was intrigued by the quick rise of the estimated or calculated total value of bribes paid worldwide. I asked by e-mail both authors what study was made and published, justifying this higher figure. TI/Peter Eigen pointed to the Bank as the source of this information and Kaufmann informed me by e-mail of 1^st of December 2004, that
     a full fledged report/study is still underway as background to the estimates given in these two articles, which we should expect to have at least in draft soon.

     Further e-mail correspondence with Kaufmann (September 6 and 7, 2005) helped me in receiving from him the full text of the article from which the Newsletter had quoted and he wrote:
     The estimation approach is simply based on extrapolating from thousands of survey responses from firms and households on bribes paid, which implies that we get an estimate for procurement bribes from firms and also other types of bribes, from firms and households. The trillion dollar estimate is obviously subject to a margin of error, it could be somewhat less or more. […] We are obtaining new enterprise data to try and have an estimate subject to less margins of error (yet they will remain) – – the sense is that an estimate ranging between 500bn -1.2 trillion may be appropriate given data so far, but as we obtain more it may be possible to narrow the range further and thus have a better estimate, which is why we will finalize a draft only in the (near) future’.
     
     He continued, answering my queries about the figures as used in the various years that
     ‘indeed in the past there was no survey data like the one we have had available more recently (which is the data I base the current estimate), and therefore the 2 estimates you cite cannot be compared
     — the old one is likely to be a vast under-estimate.   In fact, there is no comparable over time data at our disposal enabling one to make an overtime comparison, which makes it impossible to infer whether the total volume of bribes has increased or not.   As per other work, we know from the qualitative indicators that corruption has not improved much on average, yet it has not deteriorated markedly either, so it is possible that at least in real terms, the volumes have not changed dramatically (and certainly they couldn’t have changed by 20-fold, among others for reasons such as the one you mention).’

141. Remark:
     It is interesting to see that the World Bank collects her own primary data and is therefore working independently from intermediary reporting. TI/Lambsdorff makes the CPI-reports based on information obtained from using reports by others as input.
     Interesting is also that the World Bank expressly mentions the use of information from ‘households’, indicating herewith another source than the CPI that uses information from ‘businessmen, experts and journalists’.
142. I am not much satisfied with a ‘margin of error’ as indicated by Kaufmann. In my view, you cannot justifiably qualify his margins as ‘somewhat less or more’, when you publish a figure of 1 (one) trillion US dollars as your estimate for world bribes paid. However, I will not concentrate now on this aspect of the new figures of the World Bank. In later publications the WB stuck to these figures.
143. An interesting background figure was given on the internet WB-website: the $1 trillion figure, calculated using 2001-02 economic data, compares with an estimated size of the world economy at that time of just over US$30 trillion, and does not include embezzlement of public funds or theft of public assets.[17] Only 3 percent of the world economy paid in the form of bribes, it sounds rather likely.
144. For those who doubt the reliability of the one trillion figure, I add the additional information by Aleksander Bouksman, Deputy Prosecutor-General of Russia, who stated in November 2006, at the time of publication of the CPI-2006 report by Transparency International which classified Russia as number 121 (ex aequo with eight other countries), that the yearly sum of bribes paid in his country is some 240 billion dollars, about as much as the total budget of the state and also as one quarter of the World Bank global estimate. Undoubtedly, this makes it more likely that the 1 trillion for the whole world is a correct figure and maybe is still an underestimation.
145. In earlier years the World Bank published already as a reliable figure that business costs in Uganda go up on average with 8 percent because of bribes paid. For Mexico was calculated that – also on average – each household pays 6.9 percent of income for the costs of bribes.

More than US$ 1 trillion?

146. This amount of 240 billion dollars for Russia alone seems to be already quite high, however, Kaufmann (quoted above) warns in the Appendix to his chapter ‘Myths and Realities of Governance and corruption‘ in the Global Competitiveness Report of the World Economic Forum 2005, page 98, that

‘a new study estimating bribery in Russia (Satarov and Levin, 2005), if validated, would hint at a vastly larger estimate of worldwide corruption. The report estimated an annual bribe amount exceed US$316 billion, or 73 percent of Russian GDP. Even if figures such as these are, in fact, substantial overestimates, and the actual figure is much smaller for Russia, the implications for worldwide bribery would suggest a global estimate that may vastly exceed an annual figure of US$1 trillion.’

CPI high ranking dominated by rich countries
Top

147. If we look at the top ten over the years (see attachment I), than we see that there are 5 countries always among them: Finland, New Zealand, Denmark, Singapore and Sweden. Iceland is also always in the top ten since it was included for the first time in 1998. These 6 are also the top 6 in 2006.
148. Another 8 countries made it also to the top ten, but not in all the years: The Netherlands and Norway 9 times, Canada 7, Switzerland 6, Australia 5, Luxembourg 3, the UK 2 and Austria 1 times.
     None of the other countries ever included in the CPI-reports made it into the top ten.
149. It is interesting to see Singapore figure consistently in the top-class. Some more light is shed on this qualification by the message received from Michael Backman, see attachment V. Given his data, anyone will wonder how Singapore in the corruption perceptions of businessmen ever got the reputation to be nearly corruption-free. Or is Mr. Backman an unreliable source?
150. If we do the same exercise for the top-25 countries as presented in attachment I, a table comprising all CPI-reports rank-results in the eleven years from 1995 to 2005, and all countries considered by TI in those years, we see that in 2002 among the first 25 of the 102 classified (i.e. ‘most free from corruption’) there are 15 European countries. Among the 10 remaining countries in the top quartile of that year, although geographically located in other parts of the world, 8 could be qualified as belonging to the North-American and West-European economic/cultural sphere: New Zealand, Singapore, Australia, Canada, Hong Kong, USA, Chile, Israel. These eight are financially and economically strongly linked with the economy of the USA and the top-European countries. The only exceptions are Botswana and Japan. But also for these two we may point out that they are strongly linked to the financial-cultural sphere of the 15, Botswana through her diamonds and Japan through her industry.
151. Three members of the European Union did not figure among the first 25 countries on the CPI-list of TI in 2002. However, they ranked reasonably well: Portugal at 26, Italy at 31, and Greece at 44. We note also that in 2002 none of the West- and North-European countries appears in the bottom two-thirds of the list.
152. In 2005 the same picture arises from the results as presented in table I.
     Among the top-25 of the 159 countries listed, there is one more European country (newcomer Malta, all others are the same as in 2002), the total is now 16 of which 13 EU-members. All other 9 ‘new EU-members’ rank between number 27 Estonia and 53 Latvia, plus at a distance 74 Poland. From the ‘old members’ of the EU, Portugal ranks again as number 26, Italy as 41 and Greece as 48.
153. Botswana fell out of the top-25 and became 32. Israel also lowered its rank from 19 to 28. In the top-25 are 9 non-European countries geographically located in other parts of the world, 7 could be qualified as belonging to the North-American and West-European economic/cultural sphere: New Zealand, Singapore, Australia, Canada, Hong Kong, USA, and Chile. The same seven as in 2002 (plus then Israel) are financially and economically strongly linked with the economy of the USA and these European countries. Additionally, Japan stayed among the top-25 and Barbados became number 24.
154. If we broaden our view to the first quartile of countries on the CPI-2005, we find among the first 40 countries 5 more European: Portugal (26), Estonia (27), Slovenia (31), Cyprus (37) and Hungary (40).
155. From the EU-countries only Italy (41), Lithuania (44), Czech Republic (47), Greece (48), Slovakia (50), and Latvia (53), could not make it to be in the first quartile. As we can see from their rank-numbers, all of them came close and are among the first one third of the rank-list. Poland (74) obviously is a black sheep among the Europeans, the only one EU-country that ended lower on the list.

Bottom

156. If we look at the bottom end of the CPI list (‘most corrupt’) we see a similar phenomenon. In 2002 it is a stable list of poor countries most of them in the so-called Third World. These are joined by European countries that did not enter the free-market-economy realm until recently: the Ukraine at rank 86, Georgia at 85, Albania at 81, Rumania at 79 and Russia at 74.
157. Above the bottom-quartile, between ranks 46 and 55 we find Bulgaria, Poland, Croatia, Czech Republic and Slovakia. All of these are only recently beginning to get involved more and more with the European financial and economic sectors.
158. In 2005 in the bottom quartile, rank numbers 121-159, we still find as European countries: Albania 126, Russia 128 (is it ‘European’ as it is partly Asian?), and Georgia Also here Azerbaijan 137, former Soviet Union, now to be considered Asian? All others are in Africa: 18 countries, Asia 13, and Latin America 4.
159. The Ukraine went from 86 (score 2.4) in rank down to 113 (score 2.6) and Romania down from 79 (score 2.8) to rank 87 (score 3.0), but the total number of countries went up from 102 to 159, showing again that we have to be very careful in using the rank numbers.
160. Why do I go so far in detail into this ranking and the scores of the CPI?
     The reason is that it seems to be possible to draw two conclusions:
161. The first conclusion is that from the rank-positions in the CPI-listing we can conclude that the level of corruptness in particular economies and societies is rather stable: once you are a rich country, you rank high (corruption-free), and once you are poor, you are low on the list (very corrupt). Not a single country moved over the twelve CPI-years from the first quartile to the last one, or vice-versa. The most of changing places you can see is that some switching takes place across the borderlines between neighboring quartiles: between the first and second quartile, the second and third, and between the fourth and third. In all cases upwards as well as downwards.
162. Also remarkable is the second conclusion that seemingly the rank-position of Germany and France did not change over the years, despite strong involvement of their top politicians: Chancellor Helmuth Kohl and President Jacques Chirac, in quite some corruption cases in both countries. Obviously the perceptions did not change, although the reality is quite corrupt and draws abundant media-attention.
     The rank and score of these countries did not change considerably. Does this confirm that the CPI-methodology is not good enough to depict the reality? Have a look also at the USA, with a score remaining between 7.5 and 7.8. No effects from ENRON, Olympics, Anderson?

Culture?

163. Is the CPI too culturally determined? Does the chosen methodology of research logically result in the presented scores and ranks? All sources on which the ranking is based belong to the Western financial and economic sphere, all respondents represent the official (visible?) economy. The black economy, the ‘shadow economy’ of Professor Friedrich Schneider in Linz (Austria), and ‘the second economy’ of Professor T. Luca Maliyamkono in Dar-es-Salaam (Tanzania) is missing, so are their views and the perceptions of those who are immersed in these economies. And even in the rich countries, the second economy is about one-sixth of the official economy, in poorer nations the second economy is probably a much greater share of the total economy. An illustrative example from the Netherlands is that yearly Dutch money-laundering is estimated to be in the order of €18 to 25 billion, most of which, some €14 to 21 billion, comes from the USA (according to a study of the Utrecht School of Economics in 2005, financed by the Ministry of Finance). The generated additional annual tax-income for the Dutch government is estimated to be some €500 million. Nevertheless, the perception-score of the Netherlands is nearly perfect.
164. The first paragraph of this document stated already that the CPI-list is a wonderful PR-tool that draws attention every year again to the phenomenon of corruption, and generates positive feelings about an organization like TI. But should we be satisfied with this result? Maybe this was good enough for the initial years, but now that we have had the same outcome for ten consecutive years, should we (in TI) not start to be a bit more demanding and less satisfied with the world’s attention and appreciation for the CPI-list?
165. Lambsdorff is quite pertinent in his statements that inclusion of information from sources that mix the level of corruption with other variables, such as xenophobia, nationalism, political instability or expected risks due to changes in corruption, is not a valid one as this would distort measurement of perceived levels of corruption.
166. Interestingly enough, he himself takes also into account cultural differences where at the end of his paper in the Global Corruption Report 2003 he states that correlation has been observed between viewpoints of residents of countries and those of expatriates in those countries. Adding:

‘in the past, the expatriates whose view were included in the CPI were often Western businesspeople, and the viewpoint of less developed countries seemed under-represented.’

 
What next?
167-192 Proposals

167. The purpose of this paper was to show that it is unwise to continue to go on routinely with composing the CPI in the way this was done since 1995. I conclude that I gave sufficient reasons not to continue.

     We need to know what is the level of corruption in all countries? We need to know this also in a comparable way. Which countries are the role-models for others? We need to know whether in some countries the fight against corruption is being won, or whether corruption grows? A subsidiary question is whether this knowledge helps to fight corruption? What helps countries to find the best methods to fight corruption?

168. Our next task is, logically as the CPI is discarded, to develop a better tool to register the level of corruption in the countries of the world. This requires a lot of thinking and working for which many capable heads are needed, as well as time and money.

     Nobody can do that alone. In my view, TI should take the lead, answer this challenge and invite its members and experts to join hands in a project which could be most useful for the global battle against corruption.
     Some indications of what should be done I gave already in the previous pages. Some more I will give below.

169. The best advice for TI is to change course and not to look for minor improvements in the CPI as it has been developed over the past ten years.

     Now in 2007, it is better possible than in earlier days to collect facts and abstain from perceptions. This is possible as we know much more than we did twelve years ago. In the meantime a lot of research has been done. Therefore also it is not advisable, neither useful, to marginally improve the CPI methodology. We should now go for the best results, not only try to get better ones.

170. As also part of the World Bank research on measuring the level of corruption and the quality of governance deals with perceptions of corruption, we should – after this analysis of the TI/Lambsdorff methodology and research – have a similar study on what the World Bank is doing. Luckily and understandably, the World Bank relies also on other methods, and is more critical of its own behavior than TI.
     We/TI in our efforts to find better approaches to collect facts and data helping us to measure the level of corruption and the quality of governance, should have a close look at this most recent message from the World Bank:

171. January 2007 Newsletter – Governance and Anti-Corruption

******************************************************************

On Measuring Governance A Roundtable Discussion
On January 11th, 2007, a special Roundtable on Measuring Governance was held at the World Bank to discuss the opportunities and pitfalls in measuring governance across countries and over time. A discussion paper was prepared for the roundtable, as well as a presentation and the discussion.
On Measuring Governance: Framing Issues for Debate by Daniel Kaufmann and Aart Kraay, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, dkaufmann@worldbank.org , akraay@worldbank.org.
[Draft Issues paper prepared for the Roundtable on Measuring Governance, hosted by the World Bank Institute and Development Economics, The World Bank, on January 11th, 2007].
Both authors are involved in the Worldwide Governance Indicators project that began in the late 1990s (WGI), for details visit www.govindicators.org. From that website it is also possible to access a brief discussion paper as well as the power-point presentation given by the authors of this draft issues paper during the roundtable.

***********************************************************************

172. The theoretical understanding of all kinds of corruption has also grown immensely. This would allow TI to set up its own practical research, to find its own respondents, to collect its own data and to analyze these according to its own criteria, in order to raise its standing as a qualified advisor in the fight against corruption and for safeguarding integrity.
173. TI has one great advantage over the World Bank. It has chapters with members of very diverse qualities, background, training and expertise all around the world. These members are rooted in their own societies and have combined insight knowledge of their socio-economic environment which is unrivalled, and therefore of great value in collecting information on corruption and integrity and understanding what is going on.
174. TI should commission regular independent surveys.
     Distinctions could then be made between corruption and extortion, between large and petty bribes, maybe using a wider catalogue of corrupt acts [borrowed from Galtung]: including nepotism, extortion, patronage, facilitation payments, collusive networks, administrative and political corruption.
175. Fredrik Galtung identified (in his study mentioned earlier, see footnote 4 above) seven failures of the CPI. We could use these seven and convert them into recommendations for a new scoring system, allowing us to do exactly what the CPI wanted to do, but could not do:
176. scores should punish the takers as well as the givers and abettors,
     2. country coverage has to be regular and controlled,
     3. samples of respondents have to be representative,
     4. sources should secure a precise definition of corruption,
     5. sources should be precise and correctly measure corruption,
     6. scores should measure trends,
     7. scores should point reformers in meaningful direction(s).

The most important characteristic of a new tool to describe, score and rank the level of corruption should be that it allows us to reward genuine reformers.

176. The corruption syndromes of Michael Johnston could lead us to four lists with rankings under the names he gave them: ‘Influence markets’, ‘Elite cartels’, ‘Oligarch and clan’, and ‘Official moguls’ [see para. 19 above], in stead of one list.
     This could help us to justify the development of various ways of fighting corruption: everybody will understand that you have to fight corruption in a society dominated by ‘influence marketing’ with other means and motivations, than you would chose to use for fighting corruption in a ‘Mogul state’.
177. Up till now the TI-chapters have not played a role in the collection and processing of data for any global corruption index. The TI-Secretariat could play a leading role inviting the chapters to collaborate in collecting data, whereas TI-S should set-up the mainframe and ask the questions. In other words restructure TI-S in such a way that it becomes capable to run a centralized activity that has the ambition to reach for global conclusions on the state of corruption, while collecting the required information on the basis of nationally decentralized data collections.
     TI could invite all its chapters to collect data, each one to begin with their own country, local/national markets, financial institutions, politicians and businessmen, consumers and taxpayers, men and women, etc., making sure that this database represents the reality in the country, and to provide the TI-Secretariat with these data for processing and analysis in an organized scientific way.
     This could also work the other way around: TI-Secretariat indicates what research is needed and asks chapters to collaborate and use the formats as prepared by TI-S.
178. Here it should be mentioned also in passing that the TI national chapters do not play any role of this kind at the moment, neither in financing the global (net)work of TI. At best they finance their own national needs and not always only from personal individual membership contributions. Seemingly, the chapters are only a justification for the claim that TI is a global organization.
179. TI should learn from Amnesty International that such a global organization can be enriched with a global personal individual membership that supports its thinking and activities and assures the financial base for its existence, actions and continuity.
180. TI as a corruption-fighting organization should become independent from the resources the governments of the world want to spend on TI, and from business’ financial contributions. Only with an independent and strong financial base of its own through an international personal membership, it is possible for TI to be independent in its own search for the facts, establishing its own scores and rankings. I know that this will be a difficult shift in the policies of TI. We will find out quickly after having become financially independent from the powers in this world, how welcome our fight against corruption is.
181. TI will have to become serious in the fight against bank secrecy (see attachment X). It is relatively easy to prove that those in power in poor countries profit privately from illegal payments. Often cash, gold, diamonds and other presents change hands visibly. TI should not shun away from proving that those in power in rich countries do the same but in a more sophisticated way, using hidden forms, and protected by bank secrecy. Seemingly ‘corruption-free’ countries like the Scandinavian ones, Canada, Switzerland, Singapore, provide banking and investment safe havens for looted public funds and corruptly obtained fortunes.
182. Earlier, maintaining bank secrecy was considered necessary because of privacy protection. However, in the summer of 2006 it became known that SWIFT, the Society for Worldwide Interbank Financial Telecommunication, allowed the CIA already for years – illegally – to look into her databank for information, in this way enabling the American Information Agency to look for data about possible terrorists and their financial links and networks. This undermines, of course, the theory and practice that bank-secrecy has to be maintained at all times. Likewise, bank accounts should be open for the collection of information needed in the fight against corruption. TI should pronounce clearly that what is acceptable behavior in order to fight terrorism, should also be accepted as acceptable behavior in fighting corruption.
183. Professor Johann Graf Lambsdorff is probably the first one to agree that we should begin to use facts. Proof is that one of the Passauer Diskussionspapiere for which he signed, Consequences and Causes of Corruption – What do we know from a cross-section of countries?[18], hardly mentions his research that deals with corruption perceptions. The opening sentence of this paper is:

‘There is currently a wave of empirical investigations on the causes and consequences of corruption.’

Let us make use of the results of these empirical investigations.

184. He summarizes in the abstract that his paper

‘includes research on the impact of corruption on investment, GDP, institutional quality, government expenditure, poverty, international flows of capital, goods and aid. Causes of corruption focus on absence of competition, policy distortions, political systems, public salaries as well as an examination of colonialism, gender and other cultural dimensions.’

185. We can begin with all data that are already published and publicly available. To begin with what we can find in the country-reports made in the last decade by GRECO[19] (Council of Europe, States against Corruption) and OECD. The last ones[20] have also been analyzed, compared and evaluated by TI[21].
186. From all countries in the world, facts can be collected and aggregated with regard to investigations, public prosecution and judicial processes, to begin with those countries that together deal with 80 percent of all world trade. The World Bank sits on an enormous mount of information.
187. Because of the strong link between ‘human rights’ and ‘corruption’, Transparency International should collaborate much more with Amnesty International (AI).
     Was not TI’s name inspired by the name of AI?
     Where human rights are endangered, corruption is around the corner. This collaboration could be established at the level of the International Secretariats of both associations, but should also be developed at national, regional and local levels.
188. More of this kind of coalitions could offer useful additional weapons in the fight against corruption, think for instance of an NGO as Publish what you pay pushing oil-companies to make public what they pay to governments as royalties for the oil they extract from their territory, and export from the oil-producing countries) and Reporters without borders, as also the freedom of press and information is an important element in the fight against corruption. Without those freedoms and without investigative journalists, the fight against corruption has no eyes, ears and voices.
189. Gallup International should be asked to continue its useful work, and could be asked to follow wishes of corruption researchers to carry additional questions and answers in order to get more valuable data available from the polls done anyhow.
190. Quite a number of sources of data and information on corruption and integrity are listed at http://www.irisprojects.umd.edu/anticorruption/sources.htm. This is a list at the USAid website that gives the major indices commonly used by the donor-community to assess overall corruption levels in countries. It also gives the World Bank Governance Indices and Other indicators from the World Bank, Civicus, European Commission (Eurobarometer), OECD, Open Society Institute, ODI and Dag Hammarskjöld Foundation and the center for Public Integrity.

     These could be usefully consulted, analyzed and – if based on facts – could be used for much stronger surveys by TI, than at present is possible with the collection of perceptions.

191. Of course, all this work requires much more effort, probably also more staff, and therefore more financial means so that it can be executed.
     As I indicated already, preferably the financing should come from individual members in order to make the organization really independent (this is also the way AI collects its budgeted income). As long as this is not within reach and as the work done by TI is of great importance to all governments, they should contribute to covering the costs. Financing could be allocated by the UN and/or through UNODC on a project-basis, no country could object as long as all agree that fighting corruption is a common good valuable for the whole world community, as less corruption is better for the economy of all nations, and better for all consumers and taxpayers.
     Maybe those governments responsible with their people and their companies for 80 percent of all global production should pay in the first place? This means that tax-money would pay for the costs.
192. The business community could also show that she understands that all business sectors and all companies profit, if all deals are made on the well known ‘level playing field’ where corruption (bribes paid) do not give undeserved advantages to some. Business should pay fees for the services rendered by TI as the fight against corruption and the safeguarding of integrity is also a business interest.

*****

Attachment 1

CPI Rank Table 1996-2005

Transparency International is the global civil society organization leading the fight against corruption.

Transparency International commissioned Prof. Dr J. Graf Lambsdorff of the University of Passau to produce the CPI table. For information on data and methodology, please consult the frequently asked questions and the CPI methodology or www.icgg.org.

Attachment I

Corruption Perception Indices 1995-2005, ranks

Attachment Ia

Former Yugoslavia and FY-states in a nutshell

CPI Score Table 1995-2005

Attachment II

CPI Score relates to perceptions of the degree of corruption as seen by business people and country analysts and ranges between 10 (highly clean) and 0 (highly corrupt).

Transparency International is the global civil society organization leading the fight against corruption.

Transparency International commissioned Prof. Dr J. Graf Lambsdorff of the University of Passau to produce the CPI table. For information on data and methodology, please consult the frequently asked questions and the CPI methodology or www.icgg.org.

Corruption Perception Indices 1995-2005, scores

Attachment IIa

Former Yugoslavia and FY-states in a nutshell

Attachment III

CPI Sources, names and numbers per year

Attachment IV

Number of scores required for inclusion in the yearly list

Number of countries included

Attachment V

Singapore

 
Subject: FW: Singapore none too fussy about the source of wealth
Date: Thu, 27 Jul 2006 11:42:38 +0800

By Michael Backman who writes a regular column on Asian business, politics and culture for The Age newspaper, the principle daily broadsheet newspaper in Melbourne, Australia’s second largest city (see www.michaelbackman.com).

Singapore none too fussy about the source of wealth in its financial sector

YOU are an Indonesian businessman. You’ve bribed a state bank official to
give you a $US200 million ($A265 million) loan without sufficient
collateral, or a risk assessment, for a business venture you know won’t get
off the ground.

The authorities have found out and you’re facing arrest. You need somewhere
to go where authorities can’t touch you. So where do you go? The answer is
Singapore. Why? Because it is a half-hour flight from Jakarta, or 45 minutes
by ferry from the Indonesian island of Batam, and, most importantly, it does
not have an extradition treaty with Indonesia.

It is largely ethnically Chinese, just like many of Indonesia’s white-collar
criminals, if only because Indonesians of Chinese ancestry dominate that
country’s business sector.

Singapore finally agreed to negotiate an extradition treaty last year after
years of Indonesia begging for one. The process has been ridiculously drawn
out. At least six rounds of talks have been held. Indonesia is angry and
feels that Singapore is being obstructionist. But why should Singapore be
slow? Probably because it is a haven for Indonesian crooks on the run, and
they bring their money with them. Billions of dollars in corruptly obtained
funds have flowed into Singapore’s property market and its banks.

It’s a sensitive matter because financial services account for 22 per cent
of Singapore’s economy. You can imagine the situation from Jakarta’s point
of view. Singapore lectures Indonesia about the importance of the rule of
law while giving its criminals a haven.

Despite the billions it gets from Indonesia, it gives back only a fraction
in foreign assistance but then decries Indonesia for being insufficiently
grateful.

Among the Indonesian crooks and suspects believed to be on the run in
Singapore are Bambang Sutrisno and Adrian Kiki Ariawan, who were found
guilty of embezzling the equivalent of $US162 million from Bank Surya;
Sudjiono Timan, who was convicted of improperly diverting $US120 million
from a state-owned investment company; Lidia Mochtar, who is wanted over the
embezzlement of $US20 million from Bank Tamara; Agus Anwar, a suspect over
$US214 million that’s unaccounted for from Bank Pelita; and Pauline Maria
Lumowa, who is wanted over $US184 million that’s missing from Bank BNI.
Others whose whereabouts are unknown are able to safely visit Singapore.

The US doesn’t have an extradition treaty with Indonesia but co-operation by
US officials saw the fugitive Indonesian David Nusa Wijaya, wanted in
connection with embezzlement of about $US140 million, return to Indonesia
from San Francisco earlier this year.

The US embassy in Jakarta said at the time: “The US Government understands
that returning fugitives and stolen assets from abroad in corruption cases
is a top law-enforcement priority in Indonesia.”

Singapore argues that because its laws are based on English common law and
Indonesian law is based on Dutch codes, the two systems are incompatible,
making an extradition treaty difficult.

But that didn’t stop India from signing such a treaty with the Philippines
in 2004, or Australia from signing one with Indonesia. Fugitive Indonesian
banker Hendra Rahardja, who embezzled almost $US300 million, was on the
verge of being extradited from Australia in early 2003 when he died of
cancer in Sydney. His funds in Australia were frozen and returned to
Indonesia.

A corollary of Singapore’s reluctance to sign an extradition treaty with
Indonesia is its apparent lack of fussiness about the sources of the funds
attracted to its banking sector.

Singaporean officials make all the right noises when it comes to monitoring
illicit funds. But there is a perception that in practice Singapore is not
fully meeting international expectations and obligations. One person
involved in monitoring international money flows for a Western government
told me last week that the results of Singapore’s efforts to date were
disappointing.

And a senior fund manager in the region had this to say: “Singapore has
truly become the global centre for parking ill-gotten gains. The private
banking teams are huge and in practice ask almost no questions (compared
with the branches elsewhere, including Switzerland).

“An acquaintance of mine who made $US13 million through a corrupt deal (in
Indonesia) was not asked about how he got the money despite obviously having
a job that would not have allowed such amounts to have been accumulated.
Russians, mainland Chinese and Indonesians are pouring money into Singapore.
High-end property has risen 30-50 per cent in the last 18 months or so.”

Singapore, he argues, is out of step internationally. He cites a recent case
in which even a Swiss bank co-operated with the Indonesian Government in
tracking down $US5.2 million in allegedly improper funds deposited by the
former head of Bank Mandiri, Indonesia’s largest state-owned bank.

Attention is now being turned to China. Singapore is working hard at making
itself more attractive to Chinese mainlanders, be they tourists or
individuals, with funds to park. Singaporean Government representatives are
trawling through China, promoting Singapore over Hong Kong as a safe
destination for funds and property investment. Direct flights are being
established with regional centres across China. Casinos are being set up.
There has even been an influx of mainland Chinese prostitutes into
Singapore’s quasi-legal sex industry. And there’s no extradition treaty, or
little chance of one.

Of course, Singapore will argue that it takes money laundering seriously and
has all types of detection methods in place. But that is not the point. It’s
what happens in practice that counts. After all, even Chinese laundries can
have window dressing.

michaelbackman@yahoo.com
www.michaelbackman.com

Attachment VI

How national governments reacted to the Corruption Perception Index

[text from the Press Release of CPI-1997]

Bhutto: “The most honest administration in Pakistan’s history …”

The impact of the index was perhaps greatest in Pakistan. The anger of people in Pakistan over their government’s participation in rampant corruption was catalysed by Pakistan’s position as second-worst in the world table. Suddenly, this anger became focussed, accompanied by the bitter feeling that Pakistan had “deserved better” from their political elite. The reaction to the index in Pakistan was remarkable: Embassy and opposition party representatives visited TI in Berlin to ask for clarification. Many Pakistanis contacted TI which promoted the creation of a network in Pakistan and made TI a household name (as the extensive media coverage and the 300 leading citizens who crowded out a TI-Pakistan seminar in Karachi showed).

Many speakers at this conference stated that the index had contributed to the downfall of the notoriously corrupt Bhutto administration. It was the former Prime Minister after all, who had erupted angrily when the index was referred to in parliament claiming that her’s was “the most honest administration in Pakistan’s history”. Only days later she was dismissed from office by the President who was reportedly influenced in his decision to act by Ms Bhutto’s wholly irrational response to the index. Ms Bhutto lost the ensuing elections in a landslide.

The new National Chapter is targetting public procurement and working independently with the new government to reduce levels of corruption which have plagued Pakistan’s development for two generations.

Malaysia: Index as focal point for national campaign

In Malaysia, the government initially reacted strongly. Prime Minister Mahatir called the index another example of Western “cultural imperialism”. He added it was now time to set up watchdog agencies to monitor the West and their export of corruption. However, a serious effort to understand the methodology of the index was made.

A delegation of the Malaysian Anti-Corruption Agency (ACA) was sent to Berlin where the mechanics and methodology of the index were explained to them by TI.

The government then started an anti-corruption campaign – continually pointing to the TI index in its public statements and parliamentary debates as the reason why all Malaysians needed to be mobilised to counter corruption. Prime Minister Mahathir saw corruption as threatening Malaysia joining the ranks of the most industrialised nations by the year 2020. The government bolstered both the powers and the budget of the Anti-Corruption Agency.

The net outcome: The index is forming the focal point for an official national awareness-raising programme, and is often referred to in the speeches of the Deputy Prime Minister. And TI is a “name” now in Malaysia and present in the media – an excellent precondition for the future work of the nascent National Chapter of TI in the country.

Initially viewed with suspicion, TI-Malaysia is now seen as an independent partner in the push to enhance the country’s integrity.

Argentina: New anti-corruption push in the provinces

In Argentina, the index was top news for weeks. The public debate even led to a dispute between the government and Poder Ciudadano, the National Chapter of TI in Argentina.

According to Argentine press reports, Minister of the Interior Carlos Corach said the TI

Corruption Perception Index “conveys a lie, is unjust and absurd”. He added that the information was all the more irresponsible as President Carlos Menem had mounted “the most formidable campaign to eradicate structural corruption”. Hence, it was unjust and arbitrary to speak of Argentina “in such terms”.

The President himself insinuated that TI and its members were unqualified.

Poder Ciudadano presented a host of sources supporting the findings of the CPI when Luis Moreno Ocampo, its chairman, was called by the Chief Minister of Cabinet, Mr. Jorge Rodríguez, for explanations. Later on, Mr. Rodriguez and other members of government met with Luis Moreno Ocampo, Roberto de Michele and other members of TI Argentina to express the feeling of the government and the President regarding the publication of the index. One of their comments was that the index did not reflect the efforts of the government to control corruption.

While on the federal level the problem of corruption goes unabated – despite the government’s claims to the contrary -, on the provincial level there is now much real dedication in the fight against corruption. The province of Mendoza is a case in point, where Poder Ciudadano is working to include “Integrity Pacts” in all government procurement.

Attachment VII

TI CPI 1998 feedback

page editor: Georg Neumann  last modified: 15/12/2005

Feedback to the 1998 Corruption Perceptions Index

East African
Tanzania
28 September, 1998

Is Tanzania Really So Corrupt?
from our special correspondent in Dar es Salaam

TANZANIA’S disastrous showing on Transparency International’s 1998 corruption perceptions index (CPI) – ranking 81st out of 85 countries polled – prompts a number of questions. Is Tanzania really on a par with Nigeria, for many years considered (along with the former Zaire) as one of the two most corrupt countries in Africa? Is Tanzania really more corrupt than Kenya – the pace-setter in grand corruption in East Africa – or Uganda for that matter? Is Tanzania really one of the five most corrupt countries in the world, or at least in TI’s sample of countries?

Those who would answer “no” to these questions point to the high profile that President Mkapa has given to the corruption issue since coming to power three years ago. The report of the Presidential Commission against Corruption was published as soon as it was completed – an unprecedented move by any standards. The report pulled no punches, documenting the extent of both grand and petty corruption throughout the state machinery, among politicians, police, government officials at all levels, and the law courts. Washing its dirty linen in public – so the argument goes – may have affected public perceptions, making corruption appear to be worse in Tanzania than in countries less prepared to be open about the issue. Tanzania may be being punished unfairly for having the courage to raise the corruption issue in the first place. Those who would answer “yes” to the above questions argue that, although President Mkapa has taken a principled stand on the corruption issue, there is a widespread feeling that he is unwilling to confront the issue head-on since very few of those in high office are clean, and even those in the system who are relatively clean are reluctant to distance themselves from the relatively (or absolutely) dirty. An opinion poll of 500 Dar es Salaam citizens undertaken by TADREG last December showed that around two out of three of those interviewed thought that the police, the ruling CCM party, ministers, and top government officials were not doing enough to fight corruption. The poll painted a picture of a President more or less totally isolated from his fellow party and government leaders in his anti-corruption efforts. Both Kenya and (particularly) Uganda have taken anti-corruption initiatives, and it is not at all self-evident that these have been less successful than those in Tanzania. It was the Auditor General who first revealed the Goldenburg scandal in Kenya. In Uganda the Minister of Education was recently censured for corruption by Parliament, an event virtually inconceivable in Tanzania, where the National Assembly is still widely seen as a rubber-stamp institution for the government of the day.

But the above argument still does not address the question of whether Tanzania deserves to be so far down the list of corrupt countries, and significantly below neighbouring Kenya and Uganda. The CPI measures perceptions, not actual levels, of corruption. These are both local perceptions, including those of local businessmen, and those of foreign investors and risk analysts who advise on investment opportunities and risks. There is a relatively high degree of agreement between both locals and outsiders on the degree of corruption in Tanzania. Part of the problem with comparing levels of corruption is that the type of corruption which dominates or has the greatest visibility in country A may be different in country B. Grand corruption and looting, for example, have long been associated with the higher echelons of the Kenyan state, whereas Tanzania has a President still widely considered to be “Mr Clean”. This does not mean that looting is not practiced in Tanzania, of course. It may simply be that the looters are more dispersed and have a lower profile in Tanzania than elsewhere. Added together, however, the total amount of looting of tax revenues and aid receipts may not be significantly lower in Tanzania than in other hugely corrupt countries, including Kenya. Also, and perhaps related to the above, many perceive petty corruption among government officials in Tanzania as being more widespread than in the other East African countries. A third factor may be the predictability of officials delivering on bribes which they receive. There’s a joke to the effect that you pay bribes in Tanzania at your own risk, since the system is so complex that nobody can guarantee that a bribe, even a big one, will be effective. In this respect, there is no reason why Tanzania should not be a greater risk from the point of view of investors than, say, Nigeria or Kenya. Lastly, foreign investors are clearly concerned with the bureaucratic obstacles encountered and legal protection which they enjoy for their investments, and compare countries on these criteria. Anecdotal evidence suggests that Tanzania loses serious investors in key sectors on these two scores, as well as on the amount of kickbacks expected by senior officials.

While the controversy surrounding the CPI’s verdict on Tanzania will no doubt continue, at least one thing is clear. The perception that Tanzania is riddled with high levels of corruption of all kinds, which gravely undermine the country’s development potential, threaten human rights, and deepening poverty, is essentially correct. This perception is unlikely to change until the Tanzanian government, with support from the general public, the business community and aid agencies, wake up to the fact and try to do something about it.

home – about us – contact us – news room – global priorities – regional pages – policy and research – tools – publications – support us – credits – copyright

Attachment VIII.

CPI 2006

The ranks and scores as published in the CPI-2006 Report.