Tag Archives: bribes



A. Shleifer and R. W. Vishny

The Quarterly Journal of Economics, Vol. 108, No. 3 (Aug., 1993)

A Summary


In a Nutshell

The extent of corruption very depends upon the structure of the corruption network, and this has implications for growth within an economy. The paper is concerned with the sale of government property for personal gain such as passports, licences etc. which enable private agents to pursue economic activity. Insofar as officials have control over these goods they can collect bribes. The level of bribe they are able to extract depends upon the structure of the bureaucracy. In particular the presence of competition within government agencies will tend to drive the level of bribes toward zero.

In many ways bribes behave like taxes, with one critical difference: corruption is generally illegal. This facet has important implications for resource allocation and growth because officials will generally divert resource allocation decisions toward an allocation where it is easier to maintain secrecy regarding corruption. This distortion has negative affects regarding the potential for growth.

 [The article is primarily concerned with lower levels of corruption and as such may not be a hugely useful tool for analysing high level corruption by presidents etc. who are subject to different constraints.]

Basic Model

  • Government produced goods  are sold by officials who can restrict quantity supplied. An important reasons many permits/licences exist is to give officials the power to deny them (de Soto). [The implication being that if regulation was relaxed there would be less room for corruption to exist. Under de Soto’s theory regulations were put in place to restrict the ability for the new economic migrants to urban areas in Latin America from claiming a slice of the economic pie that belonged exclusively to the elites.]

The objective of the official is to maximise the level of bribes he can collect. The goods cost him nothing to provide personally. How much does he collect? Two possibilities

  1. Without Theft – here the official receives the official price of the good, plus a bribe. He pay the official price to the government. The marginal cost then of providing the good is the official price (that he must pay over). This raises the price of the good.
  2. With Theft – the official receives money but pays nothing to the government. The price to the consumer may be lower than the official price. Here the marginal cost of provision is zero. This potentially lowers the cost of the good, and is thus more attractive to consumers.
    • If the official does not price discriminate between buyers he is a monopolist and simply sets MR=MC.
    • If the official can be penalised , and if the P(detection) and the punishment are independent of the bribe and the number of people that pay it, then the official will charge the same until the expected cost of provision no longer makes corruption profitable [indicating that severe penalties for corruption could be a tool to curb it]. However, if the punishment increases with the level of the bribe he may reduce the bribe and increase output. If it increases with the no. of bribes, then he may restrict supply and increase the level [which would be worse for growth].

This model suggests that corruption spreads because of competition between officials and consumers:

  1. Officials – if jobs are distributed by auction then an official not willing to take bribes once in office simply cannot afford to get the job.  Those who are willing to collect more once installed will pay more to get the job. Competition between officials for jobs ensures that maximal bribes are collected. [Thus policies that make accountable the awarding of government jobs should be a top priority.]
  2. If buyer A can get the government service more cheaply than B, then his business is at an advantage; A can outcompete B. Thus if A bribes the official then all other competitors must in order to stay in business. This is not relevant for the spread of corruption without theft. In other words, observance of the law cannot survive in a competitive environment. As corruption with theft more closely aligns the interests of the officials and the consumers it will be more pervasive and harder to eradicate. What is needed is a good accounting system to check on officials.

Industrial Organization of Corruption

    • The above model is most relevant to monarchies, dictatorships where it is always clear who should be bribed, and the bribe is distributed amongst the relevant bureaucrats and they do not return to ask for more. In reality the ability or willingness of agencies to collude in price setting will be relevant for determining the extent and level of corruption.

If two different agencies supply one of two different complimentary licences needed for a business. Possibilities:

  1. The agencies collude on price:   they take into account the cross price elasticity of demand for the goods. In other words agency one keeps the price of good 1 down in order to expand the demand for complementary good 2, and vice-versa.
  2. Agencies are independent: both agencies ignore the cost that raising prices has on the other agency, so they set higher bribes which diminishes demand for both and results in a lower aggregate level of bribe [and production]. This is the highest level of bribe, and indicates how growth retarding free entry into bribe collection can be (e.g. setting up a toll on a road).
  3. Both agencies can supply both goods: if a consumer is bribed he will simply go to the other official. The Bertrand model of competition suggests that this will tend to drive the price of each bribe down to zero. This effect can also occur simply by the threat of new entrants to the sale of government goods. This will tend to keep corruption to zero [meaning that the provision of goods should not be restricted to one agency alone.

Different Structures of Corruption

    • Collusion is more likely to be enforced when price cutting can be easily detected and punishment severe, so i.e. when the government has a strong policing machine such as the KGB. Also relevant is the size of the elite (smaller = better) i.e. oligarchies etc.  The flip side of this is that the state must be sufficiently strong to punish those who get greedy and ask above the agreed limit.
    • The weakness of the government in African states in part explains why this collusive activity is absent, it cannot enforce bribe collusion indicating that independent monopolist structure will dominate and have terrible consequences for development.
    • Huntington noted that as states transition toward democracy corruption usually increases. This could be explained by this model whereby the state loses its strong hold over the bureaucracy machine and the system of bribes, so new entrants mean multiple agencies now take bribes where previously it was only one. This is consistent with Africa after the colonial structure broke down.
    • Countries with more political competition, media etc. are less likely to try to maximise bribes [consistent with the findings in Ferraz and Finan as summarized this week.]
    • The model indicates the best method for solving corruption is to increase the competition between officials [perhaps not that relevant for high level corruption, see Tangri & Mwenda this week.] This is the idea behind decentralization, although this applies only to corruption with theft, for corruption without theft better accounting is needed.

Corruption and Secrecy

  • Corruption is detrimental to development [see Will Easterly summary]. Mauro finds that countries with higher corruption have lower levels of investment. The independent monopolist model has the potential to reduce output to zero.
  • The fact that corruption is illegal means that people go to great efforts to keep it secret. It is much easier to take bribes on some goods rather than others, and so allocation decisions can be distorted. Government officials will use their power to ensure that people substitute goods that are easier to take bribes on for other goods that may have improved efficiency and growth.
  • The social costs of misdirection, in any of the models, can vastly dwarf the value of the bribes themselves.
  • This may explain why many nations prefer to spend on big infrastructure and military budget items rather than health and education, as the level of and ease of receiving bribes in the former, make them much more attractive purchases.