![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
This paper offers a simple strategic framework to help governments identify various policy mechanisms to minimize public sector corruption. The paper offers a formal model that blends the economics of crime with identity economics and money laundering. It presents a partial equilibrium framework that focuses on a representative public official engaged in a mix of legal and illegal effort. The model introduces various levers a government might use to impact the costs and benefits of illegal effort. The ultimate goal is to help turn volatile vicious cycles of political instability, into steady virtuous cycles of stability, growth, and sustainable development.
Acknowledgments
We would like to thank Craig Stone and other session participants at the 2016 Western Economics Association International meetings for valuable comments. We are also grateful to Ryan Sullivan and Todor Tagarev for feedback on earlier drafts. Finally, we would like to thank two anonymous referees for their valuable suggestions. The usual disclaimers apply. Views expressed in this paper belong to the authors and do not necessarily represent those of NATO, the Naval Postgraduate School, the U.S. Navy, or U.S. Department of Defense.
Notes
1. In reviewing common characteristics of countries with the highest levels of corruption, Svensson (Citation2005) finds that ‘all are developing or transition countries’ and that ‘with few exceptions, the most corrupt countries have low income levels [and] are considered closed economies.’ (24).
2. These also coincide well with the Transparency International (See Transparency International’s 2013 ‘Government Defence Anti-Corruption Index’ http://ti-defence.org/wp-content/uploads/2016/03/gi-2013-mainreportphase04.pdf).
3. Note that Bardhan (2006), Transparency International, and others have found that excessive laws, rules, and regulations actually promote corruption (i.e. excessive red tape promotes bribery and extortion). A recent econometric study concludes: ‘We do not need more regulation, but smart regulation … adapted to local conditions, and a reduction in rent-seeking opportunities provided by regulation … ’ (Mungiu-Pippidi and Kossow Citation2016) Another challenge for governments is to deter corruption in the most benign way possible to preserve the efficiency and effectiveness of organizations. While neither of these issues is explicitly addressed in this paper, they offer important avenues for future research.
4. The authors conclude:
If inculcating identity is cheap, if there is much uncertainty, if workers’ effort is hard to observe, if output depends upon special exertion at peak times, if workers are especially risk averse, if high effort is critical to the organization’s output, we would expect a firm to find it more profitable to use an identity-oriented incentive scheme. (Akerlof and Kranton Citation2005, 15)
5. Block and Heineke (Citation1975) extend Becker’s framework to explicitly include ethical costs of crime.
6. Public officials in our model engage along a continuum of legal and illegal effort. For example office supplies ‘borrowed’ from work and used at home for personal use could be considered a minor form of illegal effort (embezzlement) … resulting in modest savings (benefits) to the individual, and involving limited risk (expected costs/penalties). In contrast to Becker and others who model a binary and mutually exclusive choice of legal or illegal activity, and focus on finding corner solutions where policing and punishment eliminates corrupt effort (such that ‘crime does not pay’), our model views public officials more pragmatically. The goal is to nudge those on the margin into more legal effort and less illegal effort.
7. The NATO Building Integrity framework similarly views public officials as utility maximizers that evaluate costs and benefits of engaging in corrupt activities. Citing two eminent philosophers, Emmanual Kant and John Stuart Mill, the framework principally focuses on increasing the costs of corrupt efforts. This can be accomplished in two ways: By increasing the moral burden of bad behavior (Kant), or the expected penalties (Mill). Expected penalties consist of two components: the probability of detection, and the consequences. ‘Building Integrity’ increases the moral costs of corrupt actions. Raising expected penalties requires some combination of ‘Increasing Transparency’ – increasing the probability of detection, and ‘Improving Accountability’ – increasing the consequences (Tagarev Citation2010, Chap 2).
8. We assume the fine, F(.), is monotonic increasing in corrupt earnings, and if a penalty includes incarceration, that this could be monetized as the discounted present value of any lost wages or other possible income-generating opportunities.
9. A third possible revenue stream – ‘clean’ utility from laundered dirty income – is introduced later.
10. ‘When the agent sees herself as an insider, she maximizes her identity utility by exerting … high effort … ’ ‘When an agent sees herself as an outsider, she requires a higher wage differential to compensate her … ’ (Akerlof and Kranton Citation2005, 15) In our model, the more a public official relates as an insider, the greater (relative) utility they derive from legal income, and the less from expected illegal income.
11. The diminishing returns to corrupt income captures Akerloff and Kranton’s (Citation2005) notion of an ‘ideal.’ The authors postulate that the further an individual strays from their notion of an ‘ideal,’ the less utility they derive from any given effort: ‘ … workers have an ideal level of effort [and] lose utility insofar as they deviate from this ideal’ (Kranton and Akerlof 2005, 24). We also expect the more an individual cares about this ideal (i.e. the more they are an insider) the less utility of income they enjoy from engaging in additional illegal activity. Thus as illegal income increases, the utility from dirty income diminishes at the rate of (1 – β).
12. From the Implicit function theorem, totally differentiating (7) at the optimum (letting = U(E*,k) = 0), yields dE/dk = –U
k
/U
E
Ξ ∂E*/∂k, which reveals the marginal change in optimal legal effort with respect to a change in any parameter, k (where U
E
=
< 0). From the Envelope Theorem (Silberberg 1978, 168–171), taking partial derivatives of (6), the objective function, with respect to any parameter, k, yields the change in the overall value function, V*, with respect to a change in k. Detailed calculations are available upon request.
13. Note the comparative statics for illegal/corrupt effort, e*, are easily obtained for any parameter, k, by simply switching the signs for E*, since from (2), e*=E T – E*, implies ∂e*/∂k = –∂E*/∂k (for all k ≠ E T ).
14. The model becomes significantly more complex if the fraction of laundered income is treated as a second decision variable. For example, maximizing the value function (6′) for alpha instead of effort yields the optimal fraction to launder of: . This could offer an avenue for future research.
15. One can think of transaction costs as including the costs of avoiding detection. A useful extension of the model could unbundle the transaction cost, c, to include explicit assessment of a new risk being introduced. That associated with the probability of detection and punishment not just from illegal effort, but now also from money laundering (See Ferwada Citation2009).
16. Note that in the best case, where β = 1, a public official is a fully committed (100%) insider who completely identifies with goals of the organization, so that incentives to serve the public are perfectly aligned. In this case the individual derives no utility from either dirty money or laundered income, leaving clean money (after-tax legal income) as the only revenue stream that generates any utility.
17. Note that for an increase in the probability of detection to reduce the marginal change in expected dirty income remaining after laundering, E( from an increase in
or
=
< 0, this requires
> ∝ (i.e. the fractional penalty is greater than the exogenous fraction laundered). This is not necessary for the overall effect on dirty effort to be negative, however, as increasing P also decreases the money available to launder, as our representative public official only has the option to launder if undetected.
18. From the Implicit function theorem, totally differentiating (7′) at the optimum (letting = U(E*, k) = 0), yields dE/dk = U
k
/U
E
. Ξ ∂E*/∂k, or the marginal change in optimal legal effort with respect to a change in any parameter, k (where U
E
= V
EE
< 0). Note that for all k ≠ (
), ∂E*/∂k =
. From the Envelope Theorem (Silberberg 1978, 168–171), taking partial derivatives of (6′), the objective function, with respect to any parameter, k, yields the change in the overall value function, V*, with respect to a change in k.
19. ‘The lack of commitment to public service has been cited as an important determinant of the pervasiveness of corruption in the bureaucracies of many developing countries. In many of these economies … public service is seen as an opportunity of personal enrichment.’ ‘An important implication … is that … positions in the civil service be awarded only to individuals who have the ability to efficiently perform … duties … and not as political rewards or in exchange for bribes … ’ (Mbaku Citation1998, 204) This argues for screening, selection and retention/promotion decisions based more on merit and less on political connections.
20. Interestingly, Andreoni (Citation1991) and others find that raising expected penalties may reduce the probabilities of apprehension, and of conviction given apprehension. The former is attributed to more effort made to avoid detection and apprehension, while the latter is attributed both to greater spending to avoid conviction by the criminal, and also to a juror’s threshold level of reasonable doubt rising with the degree of punishment. Not wanting to make a type I error and convict an innocent man when penalties are particularly harsh, they may instead be more accepting of a type II error, allowing a guilty man to go free.
21. Law enforcement typically consists of four systems designed to: (i) monitor violations of the law, (ii) identify and arrest offenders, (iii) assess guilt, and (iv) administer punishments.
22. Under certain conditions, raising the cost of money laundering, particularly in conjunction with increased transparency and accountability, could also be used to reduce corrupt effort. This supports Ferwerda’s (2008) claim that ‘anti-money laundering policy deters potential criminals to commit not only the illegal act of laundering money, but crimes in general’ (1).