Transparency International describes itself as a "global civil society organisation leading the fight against corruption." Their website is http://www.transparency.org/ and they publish an annual Corruption Perceptions Index.
The list for the year 2010 makes interesting reading. The top ten countries are Denmark, New Zealand, Singapore, Finland, Sweden, Canada, The Netherlands, Australia, Switzerland and Norway. The ten bottom countries down to 178 are Equatorial Guinea, Burundi, Chad, Sudan, Turkmenistan, Uzbekistan, Iraq, Afghanistan and Somalia. The first thing that springs to the eye is that the least corrupt countries are all rich and the most corrupt ones all poor or worse, downright miserable basket cases.
There is, then, an evident correlation between rectitude and wealth as well as between corruption and poverty. But there is no telling whether the correlation signals a causal connection or merely what David Hume would call a "constant conjunction". Nor can we easily tell which way the causation runs if there is one. Are countries rich because their people are righteous, demonstrating by their prosperity that honesty is really the best policy? Or are they righteous because sheer need does not press them and they can more easily afford self-respect and clean hands? By the same token, is it corruption that drags countries into backwardness and misery, or are they corrupt because poverty has few defences against corrupt practices just as a weak body has few defences against illness?
As is often the case when the direction of causation is not obvious, it is not obvious because it runs both ways at the same time along a feedback loop. Each point and each phase in the loop is both cause and effect. It looks tempting to think that they reinforce each other in a virtuous cycle for the clean countries and a vicious circle for the corrupt ones, but I think this would be jumping to conclusions. It would be ignoring all the other causes at work in a tangle of multiple causation, including history and its unpredictable turning points.
Italy and Spain are both Latin, Catholic and Mediterranean countries. Italy is richer than Spain and we might expect it to rank much higher than Spain. Instead, it is considerably lower in the list, presumably because the history of Naples and Sicily has brought into being strong organised-crime societies as defenders of the common people and they have remained, after all these stormy centuries, an integral yet intensely corrupt thread in the social fabric. No such criminal societies exist in Spain. Another historical twist accounts, at least in part, for the place of Russia as a very corrupt country in the list. The Soviet Union was no doubt a thoroughly corrupt place, but it devoted such an extraordinarily big share of its national income to internal controls and surveillance of everybody by everybody else that large-scale corruption remained limited (though petty corruption flourished as did petty crime). With the collapse of the Soviet system, organs of control turned themselves into organs of corruption. Russia is now corrupt almost beyond belief, much more so than its wealth and education would lead us to predict.
A few standard types of corruption might be distinguished for a better understanding of the whole phenomenon. One is simple parasitism, where a person or group uses its power to suck up resources that would otherwise accrue to those under their power. The case of Angola is a shining example. French and American oil companies discovered vast oil fields offshore Angola in the 1980s and '90s and obtained concessions to exploit them. An explicit clause in the concessions forbade them ever to publish or otherwise disclose the astronomical royalties and production-sharing profits they paid annually to the government. The latter used some of it to keep the army happy and run the state, and stole the rest; in fact, it is presumably still stealing it while suppressing the evidence that it is doing so. The Angolan people who own the oil remain dirt poor. A similarly parasitic practice on a less spectacular scale is the creation of sinecures and their allocation to friends of the ruling elite and their children.
In another and most widespread type of corruption, officials in some local or central government administration are supposed to render some service (to pay entitlements, certify facts of birth, death or property, award a licence permitting some business activity), but will only do so in exchange for bribes at tacitly understood tariffs. Such corruption may be petty, but in unique cases, such as the awarding of a large government contract to a favoured insider instead of to the lowest bidder, the excess cost and the mis-allocation of resources to inefficient uses is doing great damage.
Crowning it all is corruption within the police and the judiciary that are supposed to detect and punish, and hence to scare off, corruption by others. Alliances between corrupt officials (as well as common-law criminals) and the police, the public prosecutor and the judge appear to be standard practice in Russia today, and also occur in countries with heavy involvement in the drug trade.
All or nearly all forms of corruption can be reduced to a basic principal-agent problem. Instead of principals (families, owners of property and business) acting for themselves and dealing with other principals on terms of conventional bargaining equilibria or conventional rules (that may or may not involve payment), principals resort to agents who are supposed to represent and defend them and are better equipped to deal with the agents on the opposite side. The incentives that motivate the agents are never identical with the interests of their principals. The agents are under a constant temptation to abuse the bargaining and dealing powers delegated to them, and in effect to betray their principals.
The greatest agent, over-arching them all, is of course the sovereign state acting on behalf of its principal, the people. The great majority of less universal principal-agent relations are created by and derived from the many roles the state as the people's putative agent has been empowered to play. It is the ultimate creator of opportunities for corruption.
If a lesson can be drawn from such reflections, it is that fighting corruption by investing in police and judicial activity is unlikely to yield much of a result even against types of corruption where the perpetrator is not the government itself, but its lower-grade agents. The near-epic war against the drug trade in Mexico and Colombia has shown that as narco-traders are picked off by police and army, they are promptly replaced by others, for the size of the drug industry is dependent almost exclusively on American demand for drugs that, in turn, resists attempts to reduce it by police work. In like wise, corruption fills the scope that is inherent in the principal-agent relationship and its built-in ease of abuse. The reduction of corruption would follow as naturally from a reduction in the scope of government as the decline of the drug trade would follow from a fall in the demand for narcotics.