Syndicate content

How do you measure corruption?

Deborah Perlman's picture

A coworker recently emailed me an article about corruption in Vietnam.  Both that article, which talks about the arrests of several journalists who had done extensive reporting on corruption, and some others I’ve read about anticorruption efforts in China have made me wonder whether it’s possible to really measure corruption.  Most of the existing, accepted measures, as the Washington Post noted in an editorial, rely on “perceptions” of corruption or on individual experiences of bribery. There are several well-known surveys, including the frequently-cited Transparency International Global Corruption Barometer, that ask business people or households questions such as whether they think the government is corrupt, whether they have ever paid a bribe or how much they think businesses in general spend on bribes.

These measures are interesting, but what do they really tell us?  It’s hard to know if we can generalize about levels of corruption based on whether individuals in a country will (or won’t) admit to paying a bribe, or if business owners or managers have accurate perceptions about the bribes other companies pay or general levels of corruption.  What these recent articles have reminded me is that other potential measures are even more problematic.  Does the increasing number of stories about a government’s crackdown on corrupt officials point to a large amount of corruption in the country, or to the increasing success of admirable anticorruption efforts?  If the government’s anticorruption activities are less than transparent, could seemingly impressive reports of crackdowns actually be signs of the scape-goating of lower-lever officials while higher-level corruption goes unpunished? 

What is your experience?  Are more prosecutions a sign of progress, or repression, or an increase in corruption?  Are greater public perceptions of corruption a sign of more corruption, or of a successful public awareness campaign about the ill effects of corruption?


Submitted by Casey on
Has enough of the aid provided after the hurricane been given to the people affected to survive the loss of their homes & jobs? If they had one or any? Has the ruling junta kept all of the rice, water and others supplies for themselves?

I'm not sure if you intended your comment as a response to this post or perhaps to one of the others on the blog that discusses Myanmar specifically. I'm afraid I don't have enough information to really answer your question -- since the Bank isn't providing material assistance to Myanmar directly, we have no way of tracking exactly how aid is being managed or channeled. The World Bank is supporting ASEAN's efforts to aid cyclone victims, though, so you may be able to find more information through

Submitted by Positive-Contributor on
Vietnam - public perception or knowledge does not change anything unfortunately, mainly because most people will not do anything about it since they do not feel they can make any difference or change anybody who is in power and chooses to abuse it; also because of the historical repression, it is accepted as a way of life; and if anybody wants anything done quickly that is important for their lievelihood, then to save time, travel and money (as a bottom line) they will become a part of the is short and there is only one life. I have lived in Vietnam and as most of us know, it is not different in Kenya, India, Indonesia, Ethiopia and as a recent report I heard on BBC news, in the G8 countries business transaction!! There is of course what I call intangible corruption, where somebody does something because...not because it is a part of her/his duty SK Addis

Submitted by JHC on
Citing the Post's piece(*), you bring up the topic of measures of perceived corruption, electing to highlight the Global Corruption Barometer of Transparency International. It would have been more appropriate, however, to cite TI's annual Corruption Perceptions Index, since it not only is still their best known tool, but is based on both surveys of business executives and assessments by country analysts from independent and reputable institutions as well (as opposed to 'households,' as you write a bit quaintly). In relation to your questions, as TI has noted many times in explaining why its corruption indices are based on perception, apparently objective data such as number of prosecutions are not univocally a function of the degree of corruption, but also depend on factors that vary among countries, such as the degree of legal enforcement and what is deemed illegal or not, and on the ability of different prosecutors. Lastly, in the same way that prosecutions may be used for political purposes (as you point out), perceptions surveys may be potentially affected by political interference, as evidenced by the controversy surrounding the 2008 report on Promoting Revenue Transparency of TI (, accused of having an anti-Chavez bias ( Cheers (*) This editorial of May 13 last actually focused on the World Bank's report on its anti-corruption efforts, criticising them as well as aspects of the report itself, and only in passing touched on surveys of perceived corruption.

In response to the last two comments: it’s true that there are problems with both “objective” measures of anticorruption action, such as number of prosecutions, and subjective measures such as public perception. At the Bank, our expectation is not that looking at measures of perception (whether of individuals or businesses) will change anything, but rather that if our work is successful those measures might change. We are continually searching for better ways to evaluate our governance interventions, and a key problem with perception measures is that if government or Bank anticorruption interventions are successful then the public may also be more aware of corruption. So when a survey finds an increase in the public perception of the level of corruption, it’s not clear whether corruption has actually increased or whether a successful public education campaign or media coverage of anticorruption efforts has simply made people more aware of the issue and therefore more likely to identify corruption as a problem in their survey responses. Also, thanks for mentioning the recent World Bank Independent Evaluation Group report on public sector reform. More information on the report is available here:

Submitted by Anonymous on
How to measure corruption is a multi-billion dollar question. The issues involved are probably less about perception, change in perception, or public acknowledgment of corruption; because if things do not change, then all of the above are meaningless. The difficulties are not merely about definition, objective or subjective, or about how to gather data. People may know well what corruption is but many interpretations can be used, objective or subjective, even with a given definition, no matter how imperfect it is. Perception and proofs can be manipulated. When the system is corrupt to the core, perhaps one way to measure it is looking a way to find what is not corrupt. One particular issue is whether a change in the legal system, awareness, activities, commitment can actually improve the whole economy, not just some short-term successes for a show. I also work on the same question, but I wonder if a need for a practical way as an ad hoc application is desired, or if there should be a more systematical approach. An article may help to see the inseparable of social rules and power vise:

Illicit Financial Flows from Developing Countries, 2002-2006 by Dev Kar and Devon Cartwright-Smith takes a different approach. The Abstract says: The objective of this study is to estimate the volume and pattern of illicit financial flows—money that is illegally earned, transferred, or utilized—from all developing countries based on a critical review of competing models. Through a process of testing various combinations of these models and employment of a two-stage filter for elimination of “spurious data,” this study presents a range of estimates of illicit financial flows from developing countries from 2002 through 2006. While all developing countries which report relevant data to the International Monetary Fund (IMF) and the World Bank are included in the study, salient deviations from the regional classification used in the IMF’s International Financial Statistics are noted in the paper. Overall findings indicate that illicit financial flows are growing in volume on a yearly basis with the largest recorded outflows coming from Asia and Europe. The Middle East and North Africa regions demonstrate the fastest yearly growth. While the methodology employed in this study has produced reliable estimates on illicit financial flows based on the most recent data available, the authors note that estimates of illicit financial flows based on existing economic models are likely to understate the actual problem because official statistics cannot capture all of the conduits for sending capital out of a country.

Add new comment