Strains of corruption
Nov 6th 2006
their verdict on the honesty
of governments around the world
IT IS a common fate of analysts and pundits to enjoy knowledge without power; to be informed yet impotent. But once a year some students of world affairs see that their opinions actually count for something. Such experts provide most of the raw material for Transparency International’s annual corruption index, which ranks governments (public officials and politicians) around the world for their honesty or impropriety.
Defined as the abuse of public office for private gain, corruption, as they perceive it, is worse in Haiti than in any of the other 162 countries considered this year (see chart). Of the cleaner parts of the world, Finland, Iceland and New Zealand are reckoned to be the squeakiest, although sticklers will note that both Singapore and Denmark fall within the margin of error of the top spot.
or countries like Italy, ranked 45th, the index provides an annual rebuke, and confirmation that corruption can thrive even without the alibi of poverty. For America, which is now no cleaner than Chile, this year’s ranking was a fresh embarrassment.
But for poorer countries that jockey for position nearer the foot of the table, money as well as pride is at stake. Donors, fed up with losing aid to corruption, eye rankings like these as a guide to where to put their money. For example, seven of the nine institutions that Transparency International (TI) leans on also inform the aid-giving of the American government’s Millennium Challenge Corporation. This week the corporation unveils the countries eligible to pitch for its pot of money, which amounted to $1.8 billion last fiscal year. Haiti one suspects need not apply.
Can these expert perceptions bear the weight that is now being put on them? TI itself “does not encourage” the use of its corruption indicators as a condition for aid. This is not because it doubts the veracity of its measures, but because it insists that corrupt countries should not be written off.
Others are more critical of the measures themselves. The subjective views of outsiders may be seen as mere “groupthink”, with commentators simply echoing each other’s views about a country. Some note that certain countries prosper despite corruption, and these seem to enjoy a “halo” that blinds experts to embezzlement and self-dealing. It is only when the economy falls that these same sages suddenly decide that a country’s angel-wings were moulting all along.
TI points out that some of its sources are based in the country they judge, and others were born there. Foreigners’ views also tally well with those of residents. But the most rigorous defence of subjective measures of corruption comes from Daniel Kaufmann, Aart Kraay and Massimo Mastruzzi of the World Bank Institute, who compile their own indicators of misrule, drawing on many of the same sources as TI.
Experts’ verdicts often do chime with each other, but that may just be because they are all looking at the same country. Mr Kaufmann and his co-authors find that expert opinions are more tightly correlated with the impressions of businessmen than they are with each other. They also show that if a source goes out on a limb one year, he or she is more likely to edge even further out the following year, rather than seek the safety of conformity. “Halo effects”, as they are called, may be more of a problem. Perceptions of corruption in countries like South Korea and Thailand deteriorated along with their currencies in the wake of the Asian financial crisis in 1997-98.
More worrying than bias in such measures is the lack of discrimination. Corruption has many different strains: sometimes politicians are bent, and judges are straight; in some countries, civil servants are honest executors of a corrupt politician’s will; elsewhere crooked administrators may be the main problem. Two countries holding the same place in the table, therefore, may suffer from rather different problems. TI's rankings show the analysts' disapproval, but do not show what governments must do to impress them.