Earlier this year, when I was in Liberia collecting data for a research project, a national radio station announced that Liberia had made some big steps towards becoming a less corrupt country. According to the Transparency International Transparency Perceptions Index (CPI) of 2012, Liberia had climbed from 22 to 41 points over the past seven years, which measures the perceptions of corruption in public sectors of 174 countries. With this score, Liberia outperformed countries such as Greece, Panama, and China. They were almost on equal footing with Italy, which scored 42 points. In Africa, Liberia now ranks 11th on the list of least corrupt countries, beating substantially more developed countries on the list such as Tanzania and Morocco. For those people that have never been to Liberia, these results might simply be surprising. Perhaps Liberia is one of those exceptional countries that do well out of war and that adopt sensible policies in order to further economic and social development. To me, it rather led me to question the index itself.

Liberia is not the only country that stood surprisingly high on the list (75th). Rwanda even made it to the 50th position, just before Georgia, which ranked 51st. In contrast, some countries that scored surprisingly badly considering their economic situation were Thailand (88th), Indonesia (118th), and Argentina (102nd).

Transparency International defines corruption as “the abuse of entrusted power for private gain”. Measuring corruption in a quantitative or absolute manner is not possible, warns Transparency International. “Corruption generally comprises illegal activities, which are deliberately hidden and only come to light through scandals, investigations or prosecutions. There is no meaningful way to assess absolute levels of corruption in countries or territories on the basis of hard empirical data.” The reports that do exist on court cases and uncovered scandals rather reflect the effectiveness of prosecutors and the media to expose corruption practices. The CPI therefore looks only at the perceptions of corruption in countries´ public sectors. The index measures administrative and political corruption.

Living and working in a country, or even just travelling through it, is a great way of perceiving administrative and political corruption. In Liberia, police officers at road-blocks asking for “cold water”, the local euphemism for a bribe, are a way of life. A 20 dollar handshake to get a visa renewed is simply seen as the way it works. These are just small forms administrative corruption, but they form the tentacles of a much bigger creature. The national ministry of finance only collects a fraction of the taxes they should. This is partly due to a lack of administrative capacity, resulting in a situation in which the government basically has no idea who pays their taxes and who doesn´t. However, when it comes to big multinational investors, politicians know precisely how to stash millions of dollars on private bank accounts. Embezzlement and patronage seem to be almost structural phenomena.

Admittedly, Liberia also has a lot going for it. President Ellen Johnson Sirleaf, the first female African head of state, has been tough on corruption since she was elected in 2006. Several officials have been fired after corruption scandals, and police officers are unarmed, which is primarily a security issue, but also makes it more difficult to push for bribes. The last two democratic elections were internationally regarded as generally free and fair. Liberia was also the first country to be compliant with the Extraction Industries Transparency Initiative (EITI), and is part of the Kimberly Process Scheme for the legal exportation of diamonds. These are noteworthy achievements, although they are also known for being more successful on paper than on the ground.

Leaving aside the case of Liberia, Transparency International’s claim that countries such as Rwanda, Georgia, and Liberia are perceived less corrupt or equally corrupt as Italy, Greece, and Argentina seems counter-intuitive to say the least. Perhaps this feeling is caused by the misleading sense that corruption is negatively correlated with wealth. As the Guardian warned with regard to the CPI: “Wealth seems no easy antidote to corruption: some relatively rich countries, including Russia, fall at the bottom of the global league table. Meanwhile, some of the world’s poorer states do comparatively well: Botswana, Bhutan, Cape Verde, and Rwanda all appear among the 50 ‘cleanest’ countries.” The newspaper also noted that “Arab Spring countries, and many Eurozone countries – particularly those affected by the financial crisis – are doing worse and worse”. There is anecdotal evidence of Greek politicians stashing away money on Swiss bank accounts in order to evade taxes, and of families that receive state-pensions for their long-dead parents and child support for children that were never born. Surely, the role of the Mafia also plays a role in ranking the Italy’s perception of corruption. But one wouldn’t dare to bribe a local customs-officer at a Milanese airport. Also, the fractions of public money that might be siphoned off by Greek officials might amount to the fractions of public money that actually make it to the treasury in a country such as Liberia.

Analyzing the Corruption Perceptions Index

So how do some countries end up so high on the Corruptions Perception Index, while others are placed confusingly low? Which indicators does the CPI use in order to give countries a place on the 100-point scale, on which Denmark, Finland and New Zealand scored 90 points, and on which Afghanistan, Korea, and Somalia scored no more than 8 points?

The index is based on thirteen indicators that intend or claim to measure corruption perception from different angles. All the indicators are other indexes constructed by expert organizations, consultancy bureaus, and international institutions, such as the African Development Bank, the Bertelsmann Foundation, the World Bank, the World Justice Project, Freedom House and the Economist Intelligence Unit, among others. The indicators normally measure corruption, but as part of wider indexes that focus on sustainable governance, political risk, and the rule of law, but also on economic competitiveness and economic risk for investors in countries. One indicator, Transparency International’s own “bribes payers survey”, solely focuses on the act of bribery.

There seem to be two basic problems with the sources used for creating the index. Firstly, the scores given for ‘corruption perception’ in some surveys about country risk or competitiveness of countries judge the problem of corruption in relation to other problems that affect that same country. Secondly, some of the indexes used in the CPI only use data on corruption from African countries, while others only use data from developed or industrialized countries. However, the scores that countries get in the CPI index count as if they came from the same league.

To sketch an example that illuminates the first problem, in the World Economic Forum Executive Opinion Index, one of the indicators, experts were asked about the “most problematic factors for doing business” . In Argentina, corruption is in the top three, with 11.5% considering corruption as a major problem. In Greece and Italy the sentiment against corruption is equally critical, with 11.6% and 7.1%. Contrarily, in Rwanda corruption was recognized as an obstacle by only 0.4% of those asked. The three top problems were Access to Financing (20.2%), inadequately educated workforce (19.6%), and Tax Rates (17.5%). But does this mean that corruption almost doesn’t exist in Rwanda, or rather that it doesn´t form a real obstacle for those doing business in Rwanda?

Interestingly, the CPI does not always lie in line with other research on corruption by Transparency International. For example, Transparency International also has facilitated research that found that corruption is at times positively correlated with economic growth. According to Transparency International´s Corruption Barometer, an insightful global poll that tries to measure public opinion on corruption, Rwandans are indeed surprisingly optimistic about corruption. Despite this optimism, still 11% of Rwandans admitted to having paid a bribe to the judiciary during the past twelve months, and 23% admitted to having paid a bribe to the police. In Argentina, Greece, and Italy these figures were substantially lower (6% and 16% for Argentina, 12% and 4% for Italy, and 6% and 4% in Greece ), but this apparently doesn´t withhold Argentineans, Italians, and Greeks from being extremely critical of their government and judiciary when it comes to corruption. It could very well be that complaints about corruption are also related to economical downturns, although I have not checked this. In Liberia a staggering 77% percent admitted to paying bribes to the judiciary and the police, but according to the World Economic Forum only 9.6% considered corruption to be among the most problematic factors for doing business. Lacking basic infrastructure and access to credit, many underdeveloped countries simply have bigger problems on their minds.

We could pull two careful conclusions from these observations. Firstly, some indicators in the index measure corruption relative to other problems in the same society. On the ranking, this ‘benefits’ underdeveloped countries that have much bigger problems to worry about than corruption, even if they suffer from systematic administrative and political corruption. Meanwhile, several countries that have their basic political needs covered perceive corruption as a terrible societal monster, even though they might be doing ‘better’ than several countries with a higher rank. The second conclusion is that there might be a lot of people that perceive their society, or the societies they do business in, as corrupt, while they don’t necessarily see corruption as an obstacle.

This last observation leads us to ask whether corruption should by definition be regarded as bad for social and economic stability and equality. As a matter of fact, researchers have written about the ambiguous relationship between corruption and various economic and political factors since the 1960′s. For example, in 1964 Nathanial Leff explained how corruption can “grease the wheels of an economy”, and that corruption can introduce an element of competition in what is otherwise a “comfortably monopolistic industry”. More contemporary authors have made arguments following the same logic, claiming that corruption can benefit economic growth and entrepreneurship in poor but highly bureaucratized states with little institutional capacity. Mironov (2005) and Heckelman & Powel (2008) argue that under conditions of institutional weakness, corruption is positively correlated with capital accumulation and economic growth in developing countries. In weakly insitutionalized states, corruption can also function as a social mechanism to decrease income-inequalities, as it benefits lower class workers in the informal sector getting jobs they wouldn’t get in a more formalized system (Dobson and Rodriguez Andés, 2010) . The other side of the equation is that a reduction of corruption is favorable for growth only when political and, more importantly, economic institutions are strong. These findings might help us in explaining why businessmen in Greece and Italy consider corruption a burden to bureaucracy, while those doing business in Rwanda and Liberia rather see it as an alternative to bureaucracy. Fighting corruption is thus only worth the trouble if there is sufficient institutional capacity to take away the incentives for corruption.

The second methodological problem with the Corruption Perception Index is that several of the thirteen indicators only measure corruption in developed countries, while others only measure corruption in Africa or Asia. For example, the African Development Bank Governance indicator only applies to African states. According to the CPI, the ADB rewarded Liberia with a generous 55 points on a scale of 100, which put them on par with South Africa and Ghana, admittedly worth a compliment. But this high score on governance was only possible because of the other players in the League. Liberia didn’t have to compete with European countries in order to get this score. On the other side of the equation, the IMD competitiveness index, also part of the CPI, only uses a populations of fifty-nine industrialized economies, ranging from Denmark and Finland at the top (95 points) to the Ukraine and Venezuela at the bottom (25 points). Italy and Greece respectively scored a mere 39 and 35 points on the indicators that measured corruption in the IMD´s competitiveness ranking, which leaves out developing countries. Similarly, the Bertelsmann Sustainable Governance index only compares 31 OECD countries. Not surprisingly, Greece and Italy again scored low in this ‘champions league’.

Luckily for many developing countries, they were left out of these competitions. For example, no African states except for South Africa were part of the Bribe Payers Survey. Where many developed countries were indexed through the Bertelsmann Sustainable Governance survey, many developing countries were indexed through the Bartelsmann Transformation Index. This division boosted the ranking for countries such as Bulgaria and Romania, thereby beating, again, Greece and Italy, who remained at the bottom of the other Bertelsmann index.

There are thus two reasons that explain why some countries rank suspiciously high on Transparency International’s Corruption Perceptions Index, while others rank deceptively low. The first reason is that several of the most corrupt countries in the Index simply have bigger problems to worry about, making the issue of corruption relatively less important to people doing business. In countries such as Liberia, corruption is perceived as a secondary problem after more fundamental problems that keep a state from functioning. At the same time, countries that have the basics covered have more time and money to dedicate to ‘secondary problems’. I do not say this to downplay the importance of fighting corruption, but merely to point out that there are other factors that are much more crucial to welfare and stability in a state. It could even be the case that people doing business in developing countries with weak institutions are benefiting from the option to pay bribes, because without this grease on the wheels it would be impossible to get through the bureaucratic jungles that these states often times house. In strongly institutionalized states the opposite happens. Rather than an alternative to bureaucracy, corruption represents a burden. On a side note, it seems that especially those societies suffering from social revolt and financial crises put the blame on corrupt politicians and institutions, even though objectively they might be relatively clean. The fact that people are critical or careless however doesn´t make these countries more or less corrupted. Transparency International also makes this point regarding public opinion, but misses the point when it regards the World Economic Forum executive opinions.

The second reason is that the CPI uses a several sources that only compare competitors within a certain category. The CPI then combines these indicators as if they came from the same league, creating a distortion. To put it in simple terms, the reason that Liberia scored a lot of points on some indicators is because it was regarded among the best in Africa. Bulgaria and Romania ranked high because they were regarded among the best on the Transformation Index. At the same time, Italy and Greece scored badly on other indicators in which they were only compared to other rich countries, but not to Liberia, Rwanda, or Bulgaria.

The question remains what it means to receive a higher or a lower ranking on the Corruption Perceptions Index. Does the index measure to what extent states are perceived as corrupted? And if so, perceived by whom? Or does it measure to what extent corruption is perceived a problem in these states? And if so, a problem for what? If the answer is that the CPI measures the former (perceptions of corruption levels) my answer would be that the ranking is flawed because of its methodology. If the answer is that it measures that latter (corruption as a perceived problem), the counter-intuitive rankings of some developing countries with weak institutions could be justified. However, fighting corruption in these countries might not be the solution as long as there is no institutional capacity to take away the incentives for corruption.