All discussions

August 28, 2005

Economics of Corruption

Economics of Corruption--Posner

The word "corruption" is extraordinarily vague, and, in part for that reason, ubiquitous. Charges of corruption are everywhere. Notably, economically booming China is nevertheless said to be seething because of the corruption of local officials, and Chicago's mayor is being questioned by federal investigators about corruption in his otherwise very successful administration.

The problem with the word is twofold. First, identical practices sometimes are "corruption," sometimes not. Second, despite the pejorative connotations of the word, the normative signficance of corruption is not always clear.

Fifty years ago it was common in nightclubs in New York (maybe it still is--I haven't been in a nightclub in New York in 48 years!) to have to give the headwaiter a tip in order to get a table, even if there were many empty tables. This was a form of bribery, but accepted as proper. Management knew about the practice and condoned it. The headwaiters were doubtless paid less than if they had been forbidden to accept these bribes, but that of course was not a clear gain to the nightclub; the nightclub presumably charged customers less because the full cost of the entertainment to them was greater by the amount of the bribe. So far, a wash; but if the bribes induced the headwaiter to be friendlier and more helpful to the clientele, the nightclub was better off. Likewise with tips for waiters and waitresses, despite the possibility that a generous tipper will get better service at the expense of other customers, to the harm of the latter.

So what is wrong with bribing public officials to obtain public services, provided the practice is known and wages are adjusted accordingly? In effect, bribes shift the financing of public services from taxes to a combination of taxes and fees for service. By injecting a market element into public services, bribes can actually improve efficiency when used to get around rigid or inefficient rules. To recur to the 1950s in New York, municipal ordinances forbade contractors doing construction work to obstruct sidewalks and streets, but often it was impossible to do such work without creating at least minor obstruction and so contractors bribed police to look the other way. The net effect on social welfare was probably positive.

But there are several problems that together make bribery of public officials on balance inefficient--and thus "corrupt" in an unequivocally bad sense. First, not all rules are inefficient, and bribes are bad from an economic standpoint when they subvert an efficient rule, as when a building inspector accepts a bribe to overlook a serious fire hazard. Second, without competition among bribe takers (in the New York nightclubs this was secured by the competition among the nightclubs themselves, which limited the amount of bribes that management permitted its headwaiters to receive), the bribe will exceed the cost of the public service being purchased with it, distorting the allocation of resources (though the higher taxes that would be required to compensate public employees who did not have bribe income would also have distortionary effects--all feasible taxes do). Third, delay and uncertainty are created when multiple officials must be bribed. And fourth, a bribery culture reduces pressure to repeal inefficient laws--in fact, it creates in public officials a vested interest in preserving such laws. In that respect, it is a protection racket.

Since public corruption seems on balance inefficient, the question arises why it is so common. The answer is that corruption flourishes where the economy is heavily regulated but the legal framework is weak. The more heavily regulated the economy, the more irksome restrictions there are that will create a demand for methods of avoiding compliance with them, and bribery of the enforcers of the restrictions is one such method. The weaker the legal framework, the more difficult it will be for the government to prevent bribery, a classic "victimless" crime because bribery is a voluntary transaction; and it requires a sophisticated legal machinery to detect and punish such crimes.

There is another and subtler effect of the legal framework. Unless there is an effective machinery for the impartial enforcement of contracts, people will be reluctant to do business with strangers. Economic activity will tend rather to be organized on the basis of familial and other personal relationships. In such a culture it will seem perfectly natural for public officials to exhibit favoritism toward friends and relatives, including persons who purchase their friendship with a generous bribe. Nepotism, clientalism, and bribery become substitutes for contract when the enforcement of contracts is undependable. In contrast, corruption should be rare in a free-market system with courts that enforce contracts honestly and dependably.

So how to explain public corruption in America's big cities today? It seems less common than a half century ago, and perhaps that is because there is somewhat less economic regulation and also a somewhat greater professionalism in civil services, police, and the judiciary. Another factor is that most big cities have Democratic mayors, and the Presidency has been in Republican hands for almost two-thirds of the period since 1969; Republican attorneys-general are more likely to investigate and prosecute public corruption in Democratic-controlled cities than Democratic attorneys-general are. Becker discusses other causes of the decline in U.S. corruption in his comment.

The persistence of corruption in some of our big cities may reflect the presence of immigrant communities in these cities, in which barter and other forms of reciprocal dealing based on (and constructing) relations of trust, extended family relationships, clan ties, and the like continue to organize significant economic activity and make it natural to think of public officials as "selling" public services to their friends and relatives.

The problem of corruption underscores the importance of the legal framework to economic development. An honest, incorruptible police, criminal law enforcement machinery, and judiciary can increase economic efficiency by greatly reducing the amount of corruption (as well as in other ways), though it is equally important to have a commitment to free markets and a workable legislative and regulatory machinery to prevent economic activity from becoming encrusted with inefficient restrictions.

Comment on Corruption-BECKER

I agree with Posner's basic approach to corruption, so I will elaborate on some ways to reduce corruption, refer to evidence where corruption actually helps performance, and offer a suggestion for why big-city corruption in America appears to have declined over time. I confine my comments to corruption in the public sector, although for every public official who is bribed, there is always a businessman, union official, or someone else in the private sector that is doing the bribing.

By corruption I simply mean that public officials accept payments that violate some laws in order to affect the implementation of other laws or regulations. Corruption so defined is bad if it lowers efficiency in the economy or society- that is, if the cost imposed on everyone else exceeds the gain to an official. Good corruption raises efficiency, so while the corrupt official may gain, so does the economy and society as a whole.

Clearly, in a country with bad laws, corruption by officials that enable businessmen and others to get around these laws may be helpful. The Soviet Union, for example, had terrible economic and other laws, and performed badly. Still, the widespread corruption that existed helped it to do much better than it would have if all officials followed the letter of the law. A preliminary study by a graduate student at the University of Chicago, Maxim Mironov, analyzes the effects of corruption on economic growth in 140 nations during the past decade. He finds that corruption in countries with weak institutions, defined by government effectiveness, the rule of law, and the quality of regulations, appears to help countries grow faster, whereas corruption in countries with good institutions slows down economic growth.

For the remainder of my comment I concentrate on corruption that on balance is bad. Posner points out that corruption flourishes with a weak legal system, and with larger government. Obviously, if governments strongly regulate many activities, then companies, unions, and other groups that are regulated can do better if they can "bribe" officials to overlook or relax these regulations. So the wider is the reach of governments, the greater is the corruption potential. There was relatively little corruption in the Federal government of the US in the early 19th century primarily because the government did so little then.

Other than narrowing the scope of government and strengthening legal institutions, what can be done to reduce (bad) corruption? One simple step is to improve the incentives of officials to act honestly. The incentive to be honest would be stronger when officials are better paid, and if they are fired from their well-paying jobs, and sometimes also punished rather severely if they are caught engaging in corrupt behavior. A few studies do support this conclusion that corruption thrives more in environments where officials are badly paid, such as policemen in Mexico.

Corruption is reduced by greater competition between separate political jurisdictions and stronger competition for political leadership. This implies that corruption is lower in decentralized political systems compared to centralized systems. Various studies do indicate that democracies generally appear to have less corruption than totalitarian systems, although some of the corruption in totalitarian systems like the Soviet Union may be of the good kind because the laws are so bad.

Corruption is reduced when information is more easily disseminated to the public. That is why a free press is such an important protector against greater corruption. The press is more effective in better educated societies, and various studies have shown that corruption is lower when education is greater. Education also helps cut corruption by improving political institutions, so part of the positive relation between the amount of corruption and the weakness of institution is the result of the positive connection between education and good institutions.

Work in progress by Professor Edward Glaeser and others at Harvard University suggests that corruption in the US declined over time in part because education increased. This helps answer Posner's question about why corruption in big American cities appears to have been declining. The steep growth in regulations over time would suggest growing, not declining, corruption. I believe the increase in education, combined with more vigorous competition among print and other media to disclose information about corrupt officials, and greater geographical mobility of many types of business have all contributed to the apparent decline over time in the amount of corruption in big cities.