Malcolm E. Fabiyi, PhDTuesday, October 28, 2008
[email protected]
Chicago, IL, USA



ational Decision Theory is a standard approach used by economists and other social scientists to explain why individuals make the decisions that they do. The field of decision theory has spawned several Nobel Laureates, and a healthy body of experience has developed to support its fundamental premise. The behavior of stockholders, economic cycles of boom and bust, criminal behavior, and a host of other behaviors have been explained by this theory. In this article, I apply simple rational decision theoretic rules to explain why corruption thrives in Nigeria. I also apply the theory to identify some critical areas where public policy efforts should be focused if corruption is to be tamed.

At its core, decision theory is based on simple and intuitive concepts. It assumes fundamentally that given an option, W, which has a likelihood or probability of occurrence, P, the expected benefit from taking that choice is given by the multiplication of the option, W and the probability P. In simple mathematical terms, we would say that:

Expected Benefit = W x P.

Of course any choice options naturally carry corresponding costs. Therefore, given a Cost, C with a probability of occurrence P', the expected cost of the option is given as:

Expected Cost = C x P'.

In effect, the expected net benefit of any option W given a cost C and respective probabilities of occurrence P, and P' is given as:

Expected Net Benefit = Expected Benefit - Expected Cost = W x P - C x P'.

Whenever the Expected Net Benefits are greater than zero, the choice is one which a rational decision maker is expected to make. If the Expected Net Benefits are negative, the rational thing to do is to turn down the option. Now that we have a decent handle on rational decision theory we will apply the concept to two hypothetical corruption cases.

Case 1: Low Enforcement regime

A 45 year old Director General in the Nigerian Civil Service earns N8m annually (including allowances, rent and home ownership subsidies and other emoluments). He has the opportunity to corruptly enrich himself to the tune of N10 m. If caught, he will suffer reputational damage for the rest of his life, spend some time in jail and lose his pension entitlements. He estimates that the losses he will suffer if caught and prosecuted will come to about N5 m annually. These losses can be calculated as a perpetuity, which at an assumed 10% interest rate gives estimated lifetime losses of about N50 m. He knows that there is a 10% chance that he will be caught, and if caught, there is a further 10% chance that he will be prosecuted. He expects that he will be found guilty if prosecuted. Should he corruptly enrich himself?

Answer: Yes. The Expected Net Benefit is positive (+N9.5 m).

  1. Probability of getting embezzled amount (P1)= 100%

  2. Probability of getting caught (P2) = 10%

  3. Probability of getting prosecuted (P3) = 10%

  4. Probability of being found guilty if prosecuted (P4) = 100%

  5. Estimated Lifetime Losses (C) = N50 m

  6. Gains from Corrupt enrichment (W) = N10 m

Expected Net Benefit = P1 x W - P2 x P3 x P4 x C

Expected Net Benefit = 100% x 10 m - 10% x 10% x 100% x 50 m = 10 - 0.5 = N9.5 m.

Case 2: High Enforcement Regime

Consider the case of the same civil servant. Assume that now the enforcement and judicial environment is radically different. The government has instituted a policy that rewards whistleblowers who report corrupt officials with 2.5% of the total embezzled amount. There is also a 2.5% bonus reward that goes to the lawyers who prosecute the cases if there is a successful conviction. These developments have caused the probability of getting caught to increase to 50% and the probability of getting prosecuted to increase to 50%. Should the Civil Servant still corruptly enrich himself in this new scenario?

Answer: No. The Expected Net Benefit is negative (-N2.5 m).

  1. Probability of getting embezzled amount (P1)= 100%

  2. Probability of getting caught (P2) = 50%

  3. Probability of getting prosecuted (P3) = 50%

  4. Probability of being found guilty if prosecuted (P4) = 100%

  5. Estimated Lifetime Losses (C) = N50 m

  6. Gains from Corrupt enrichment (W) = N10 m

Expected Net Benefit = P1 x W - P2 x P3 x P4 x C

Expected Net Benefit = 100% x 10 - 50% x 50% x 100% x 50 m = 10 - 12.5 = -N2.5 m.

Drawings Public Policy Lessons from Rational Decision Theory

While these examples are hypothetical, the implications are clear. The self interest of people and the proven rational basis for much of human behavior can actually be used as tools for cracking down on corruption. Rational decision theory suggests that if corruption is to be tackled effectively public policy must focus on (i) increasing the likelihood of getting caught, (ii) increasing the likelihood of prosecution and (iii) raising the penalties borne by corrupt officials.

(i) Encourage whistleblowers so that the likelihood of catching embezzlers increases. There is always a long line of clerks, messengers, cashiers, accountants, personal bankers and auditors that are privy to every single act of corruption. Right now, these people keep quiet because there is no incentive to make them come forward with the information that they have. Imagine what would happen if a law was enacted that guaranteed a reward of 2.5% of the recovered proceeds of corruption to a whistleblower (the payout should probably be capped at N10m or so, otherwise the incentives become perverse). I am confident that the authorities would be inundated with all types of information about corrupt activities. As a consequence the likelihood that corrupt officials will be reported should increase, thereby increasing the costs to the perpetrators of corrupt activities.

(ii) Provide Incentives to Lawyers so that the likelihood of prosecution increases: Even where corrupt officials are arrested, we have seen situations where trials drag on forever, or the officials are let off the hook because of the ineptitude and sloppiness of the lawyers charged with their prosecution. To get prosecution rates up, it will be prudent to provide an incentive that rewards the successful and timely prosecution of cases. Guaranteeing some portion of the recovered sums to successful prosecutors could help in this regard. In the corporate world, where shareholders are faced with what is called the Principal-Agent problem, the commitment of the CEO and other executives to making sure that their (the executives') focus is on growing shareholder's wealth is secured by making stock options part of the compensation package. If the stock does well, then the executives benefit as well.

(iii) Increase the penalties associated with corruption: As the cases suggest, a critical element that determines the expected net benefit (or loss) of corruption is the magnitude of the lifetime loss suffered. In all the instances considered, if the lifetime losses had been more extreme, the expected net benefits of corrupt activity would have greatly reduced. Some of the steps that could be taken from a public policy perspective to increase the costs to the perpetrators of corrupt activities includes severely restricting the economic opportunities available to people convicted of corruption. For instance professionals like lawyers, doctors, and engineers who have been convicted of corruption could be stripped of their charters and/or licenses. Convicted embezzlers could be excluded from participation in the stock market, or politics, and onerous forfeiture requirements could be imposed to make the lifetime losses large and significant.

Dr Fabiyi is a former President of the University of Lagos Students' Union (ULSU)