Вы находитесь на странице: 1из 17

Bribe is wealth earned by accomplishing a task for a party that should have been done without any

compensation (whether to bring benefit to the party or to avoid harm). In some situations, bribery
is similar to tips or rewards and distorted by some people who say that this is not bribery, but a

Bribery is clearly prohibited in the Al-Qur'an and Hadith. Allah s.w.t. mentioned:
Meaning : "And do not devour your property among yourselves by wrongful means, nor offer it as a
bribe to judges, with intent that you may unlawfully swallow up a portion of other people's
property, while you know." (Al-Baqarah:188)
It is mentioned in a hadith:
Meaning : "The Prophet s.a.w. has condemned the giver or receiver of bribe in decision making
(ruler, management, judges etc...)" (Narrated by At- Tirmidzi, 3/622: Imam Tirmidzi said: Hasan
1) Money, donation, loan, fee, gift, expensive collateral, properties (moveable or immovable),
rights in properties or any other similar benefits;
2) Any positions, titles, designations, jobs, contracts or services, and promise for any type of job or
3) Any payments, exemptions, settlement of loans, obligations and liabilities either in whole or in
4) Any type of compensation, discount, commission, rebate or bonus;
5) Any action that prevents oneself from claiming any money or monetary value or luxurious items;
6) Any type of service or help such as protection from any penalties, disciplinary penalties, civil or
criminal charges. This includes using one's authority or preventing from using one's authority to
provide that protection; and
7) Any offers or promises, with or without conditions, to give bribes such as any of the above items
Below are several types of bribery:
Requesting/receiving bribe
Any person or agent that request/receive/agree to accept any bribe (money/items/services etc.) as
a motivator or reward to do or not to do anything related to the transaction or related to a formal
transaction, principal/employer
Offering/giving bribe
Any persons or agents that offer/give/agree to give a bribe (money/ items/services etc.) as a
motivator or reward to perform or not to perform any related transaction or related to a formal
transaction, principal/employer.
Making a fake claim
Any person who gives to an agent, or an agent gives to his principal/ employer, a document that
he/she knows or realizes contains forged information, which is important to the
principal/employer, with the intention of deceiving his/her principal/employer.
Misusing title/position
Any public sector executive that use their title and position to receive bribery for any decisions or
actions related to himself, where the executive, his relatives or his cronies have an interest in
directly or indirectly.
Source: Anti-Corruption Agency, Malaysia
If you are involved in business or politics, its not uncommon to face one of the situations above. If
so, remind yourself of the following:

The Prophet s.a.w. said:

"Whoever that we take to do a job, and it is allocated for you a rizq (wages or rewards), then
whatever that he takes after that is ghulul (betrayal)" (Narrated by Abu Daud, no 2943, Albani:
Sahih; Ma'lim as-Sunan, Al-Khattabi, 3/8 Dar Kutub Ilmiah publication ]
2) A governor responsible for collecting zakat from the Azad tribe, when completing his job,
returned to the Prophets.a.w., keeping half of money and said
"This is for you (portion for Islam) and this has been given to me as a gift."
The Prophet stood up on the pulpit and said:
"Be known, leave and stay at your father's or mother's house to see whether you will be given gifts
or not, if you are really true (deserve a gift)." (Al-Bukhari & Muslim)
3) A narration mentions that Rasulullah s.a.w. sent Abdullah Rawahah to Khaibar (Domain of the
Jews that was under Islamic authority) to value the dates from the district since
Rasulullah s.a.w. has decided that the revenue of Khaibar should be divided into two; half for the
Jews and the other half for Muslims.
While Abdullah Rawahah was performing his tasks, the Jews came to him and brought a variety of
ornaments and said to Abdullah:
"These ornaments are for you, lighten us up and give us more than half." Abdullah replied, "Dear
Jews! For the sake of Allah s.w.t. you are indeed Allah s.w.t.'s creation that I hate the most. What
you are doing now makes me hate you more. The bribe that you offer is haram, and we as Muslims
will not have it!" By hearing the answer, they (the Jews) then said, "On this (attitude) the heavens
and earth stand!" (Imam Malik, Al-Muwattha':1450)
4) Rasulullah s.a.w. said:
"The gifts given to the enforcer are suht (haram) and the bribery received by a judge is
kufur" (Narrated by Ahmad).
Call it bribe or gift; both are obviously haram in Islam. All wealth gained from bribery is haram, to
keep them is haram and to expense them is haram. Hence, every receiver, giver and middleman
involved in this haram activity must be punished, and the punishment for them is ta'zir.
Heavy ta'zir has to be imposed on these individuals as a lesson to them and to discourage other
Muslims. Nowadays, without the application of the Islamic uqubat system, we can see people who
are involved in bribery still smiling and looking so calm while facing trial in court. They feel no guilt
or worry for the despicable things that they have done.
According to Islamic law, an individual who knows is obligated to report it to the authoritative
bodies. If not, he is considered abetting and will bear the sin as well. Hiding or letting this
epidemic continue will have a negative effect on Islam, corporations, organizations and the
country. In the end, all will be exposed and eventually collapse.
Unfortunately, I am sure the reasons below are among the main barriers to reporting any
1) Does not have written evidence or if he has, the document is private and confidential (P&C). If
he reports it, he will instead, be charged with the Official Secrets Act (OSA). This problem should
be reviewed by the authoritaties if they are really serious in fighting bribery.
2) Sympathy, because they are friends or close acquaintances. To those involved, remember that
you will also bear the sin of abetting.
3) A "don't care" attitude. This is also considered sinful in Islam.
4) Too lazy to go to court. Rightly, there should not be laziness in fighting what is wrong and evil.
This is an indication of a weak iman.
5) Worried that it will backfire if the whistleblower is not protected. Authorities must ensure
protection or no one will come forward.
6) No confidence in the authoritative bodies. Concerned that the authoritative bodies will be
bribed and the whistleblower is charged instead. The authoritative bodies need to earn the public's
7) Anxious because the individuals involved in bribery is well connected and powerful. These are
indications of a country awaiting its destruction.
The Prophet s.a.w. said:
"Indeed, the previous nations perished because if the respected people among them steal, they
will ignore it, but if lower class people steal, they will cut their hands. For the sake of God in
whose hand my soul resides, if Fatimah Muhammad steals, certainly I will cut her hands!"
Bribery is haram in Islam because it will corrupt society and the nation. I do not intend to give a
lengthy explanation regarding this, only a brief overview on the implications of bribery:

1) Easy ticket to hell. For Muslims especially, bribery will ease the path to hell.
2) Obliterate fairness in society. Protection of wealth, dignity, life and self cannot be insured. All
can be bought.
3) Producing an incompetent society. The number competent, skilful and knowledgeable people will
decline as skills have no more meaning. The skill to bribe is more important.
4) Selfish society. Individuals will only think about themselves without the negative effects on the
victims in the short term and to the country in the long-run.
5) Producing poor quality work. It threatens safety and peace. A job granted due to bribery exposes
the public to danger due to low quality of work/construction.
6) Increase in management cost. New entrepreneurs will face expansion problems due to cost of
7) Slow response and management. People will delay a task to wait for higher bribes. Approvals may
take longer if the executives at each level desire bribes before processing an application etc.
8) Snowball effect. The person who takes bribes enjoys it and feeds his family with it. Eventually
the whole family may be addicted to easy money. Allah s.w.t. said:
"In their hearts is a disease (of doubt and hypocrisy) and Allah has increased their disease. A
painful torment is theirs because they used to tell lies." (Al-Baqarah:10)
Hence, to those who feed his family with bribery, without sincere taubat, Allah s.w.t. may plant the
opportunity for his children to follow his footsteps that will eventually annihilate the family.
9) Imbalance in the distribution of wealth. The rich will become richer and the poor will become
poorer. The poor will face enormous difficulties in improving their lives since they have no
connections', and not enough money.

The Law and Economics of Bribery and Extortion

Annual Review of Law and Social Science
Vol. 6: 217-238 (Volume publication date December 2010)
First published online as a Review in Advance on August 6, 2010
DOI: 10.1146/annurev-lawsocsci-102209-152942
Susan Rose-Ackerman
Law School and Department of Political Science, Yale University, New Haven, Connecticut 065208215; email: Susan.rose-ackerman@yale.edu
FULL-TEXT| PDF (214 KB)| Permissions
Citation: Web of Science | Download| Email notification|
Web of Science : Related Records | Times Cited: 4

Corruption has serious economic and social costs and can undermine government legitimacy.
Economic analysis can help one understand the incentives for bribery and extortion and the deterrent
effect of the law. Such analysis suggests that the law in many jurisdictions ought to be redesigned.
Penalties are poorly tied to the marginal benefits of bribery. Small bribes often are more effectively
deterred than larger ones because penalties are not tied to the perpetrators' gains. Economic analysis
also highlights the tension between obtaining evidence to bring a case ex post and deterrence ex
ante. Furthermore, enforcement programs have not incorporated bureaucratic structure in a
sophisticated way, and in many countries the criminal law only applies to individuals, not firms. In
short, economic analysis can help guide the reform debate by proposing workable law enforcement
strategies for the control of bribery and extortion.

The economic analysis of bribery falls into two broad categories. First, economics can help isolate the
underlying factors that produce corrupt incentives, independent of the strength of bribery laws.
Empirical work builds on this theory to produce estimates of the social costs of corruption and its
impact under a range of conditions. The first section of this review summarizes work in this tradition,
both theoretical and empirical. Second, economic principles can help one assess the laws and
policies against bribery to see whether they deter payoffs effectively. The main goal of this review is to
summarize and critique research on this second topic. Thus, the subsequent section discusses
deterrence and argues that the criminal laws of most jurisdictions are blunt and ineffective in their
efforts to deal with bribery and extortion and with their accompanying social harms. The final section
embeds bribery in an organizational context, both for those who receive and those who pay bribes,

and it asks how the law ought to tailor deterrence strategies to take account of alternative
organizational forms.

Private actors frequently make payments to public officials to obtain a benefit or to avoid harm. If
these payments are pocketed by the recipient or used for partisan political purposes, most legal
systems classify some such payments as bribes and can impose criminal penalties on those who pay
and receive them. A few systems, however, treat payers and recipients differently depending on the
nature of the benefit.
National legal systems differ as to whether business firms or other organizations, as opposed to
individuals, can be guilty of bribery. They also differ in their treatment of payoffs that only involve
private actors, such as a payment from a salesman to a purchasing agent. In some legal systems,
business firms cannot be guilty of a crime but can be subject to fines, and, in others, commercial
bribery is a civil, not a criminal offense.
Principal-agent relationships are at the heart of any corrupt transaction. An employee or another
person acting as an agent for a government body or a private organization accepts a private benefit in
return for acting in the payee's interest. The bribe taker may be able to solicit bribes simply by dint of
his or her position of authority or may enhance that authority with threats. 1
Cross-country empirical research demonstrates that corruption is associated with lower levels of
investment, productivity, and growth and that corruption discourages both capital inflows and foreign
direct investment (Mauro 1995; Wei 2000; Graf Lambsdorff 2007, pp. 7179, 100107). Corruption
reduces the effectiveness of industrial policies and encourages business to operate in the unofficial
sector in violation of tax and regulatory laws (Ades & Di Tella 1997, Kaufmann 1997). Highly corrupt
countries tend to spend less on education, to overinvest in public infrastructure, and to have lower
levels of environmental quality (Mauro 1998, Esty & Porter 2002, Tanzi & Davoodi 2002). Overall,
corruption reduces the perceived legitimacy of democratic governments (Anderson & Tverdova 2003,
Seligson 2006). Of course, one can question the quality of the corruption data as well as the direction
of the causal arrow. Does corruption cause these outcomes or is corruption the result of these
underlying conditions? It seems likely that the causation goes both ways. 2
Furthermore, macro-level data are not very helpful in designing reform strategies. The social costs of
corruption differ in different situations (Rose-Ackerman 1978, 1999; Shleifer & Vishny 1993). Thus,
one needs detailed studies that explore the mechanisms at work in particular situations. For example,
a study of education in Uganda documented how central government subsidy funds disappeared at
the regional level before reaching the village schools. The level of transfers improved once local
parent-teacher associations were given information on funding levels. A study of hospital purchasing
in Argentina showed how information on the dispersion of prices across institutions helped rein in
corrupt purchasing (these situations are studied in Reinikka & Svensson 2004 and Di Tella &
Schargrodsky 2003; Rose-Ackerman 2004 provides additional references).
In doing such analyses, one should distinguish between whether or not the briber is entitled to the
benefit received and whether or not the benefit is scarce. There are four relevant categories: an illegal
benefit, a legal but scarce benefit that may be unfairly or inefficiently allocated, a legal benefit that is
not scarce if allocated honestly, and a legal benefit whose corrupt allocation is not inefficient but that
generates a wealth transfer.
If the benefit provided is illegalfor example, permission to import illegal drugsthe social harm is
the distortion introduced by corrupt payoffs. If the benefit is legal but scarce, the corrupt official gives
preference to bribers over other potential beneficiaries. The social harm is the net cost of allocating by
willingness to bribe instead of the stated criteria of the corrupted program. The first two categories
might combine in the case of an illegal and scarce benefit. For example, corrupt and budgetconstrained police authorities might only arrest those who do not make payoffs. In the third category,
the benefit may be legal and appear scarce only because of corrupt public officials. The social cost is
then the distortion created by the officials' efforts to create scarcity as a way to generate payoffs.
Finally, in some cases bribery might have no allocative effects and merely be a transfer from one
pocket to another. The only efficiency cost would be the time and trouble involved in keeping the
transaction secret. Such corrupt systems, however, especially if supported by credible threats, have
distributive consequences that are likely to be viewed as unfair and unjust (Rose-Ackerman 1978,
1999, pp. 926).
The social costs of bribery include the inefficiencies introduced by payoffs, the unfair distribution of a
public benefit, and the erosion of legal entitlements to act free of threats to person or property.
Corruption in law enforcement lowers the deterrent effect of sanctioning schemes. Systemic
corruption can undermine state legitimacy over and above its impact on the operation of particular
programs. In addition, corrupt schemes and other forms of self-seeking can be costly to implement as

public officials and bribe payers spend time and resources in an effort to keep their actions secret (the
literature on this issue is summarized in Rose-Ackerman 2004, Graf Lambsdorff 2007).
The economic analysis of bribery suggests a way to think about reform strategies. First, do the
payments produce harmful results? If not, the underlying policy environment should be reformed,
perhaps by legalizing payments. Second, if there are harmful consequences, what are the relative
merits of redesigning the program to limit bribery, of legalizing a formerly illegal activity, or of
implementing enhanced legal controls to catch and punish those who make and receive payoffs?
Different types of bribery have different costs that depend on background conditions. It is possible in
principle for some bribes to be beneficial, on balance, if they overcome unfair, inhumane, or inefficient
practices. For example, bribery might permit someone to escape an oppressive regime, or it might
allocate import quotas to the most efficient firms. However, bribery is always a second-best outcome.
If bribery overcomes inefficient rules, the best policy is either to repeal the rule or to legalize payments
to the state. Official tolerance of bribery leaves in place the transaction costs of corruption, favors
those who disrespect the law, encourages officials to create even more red tape and delay to increase
incentives for payoffs, and undermines state legitimacy (Bardhan 1997; Rose-Ackerman 1999, pp. 9
If a willingness-to-pay criterion is not itself objectionable, a state can introduce legal price incentives in
some situations. For example, it can sell import quotas, permit applicants to pay for speedy service
(for example, in obtaining a passport or receiving expedited customs clearance), and sell tradable
pollution rights. Alternatively, the state can repeal regulatory systems, such as complex procedures for
registering new businesses, that generate bribes, or it can legalize formerly illegal activities.
Whenever a business, such as the drug trade, gambling, or prostitution, is criminalized, this creates
corrupt incentives (Andrianova & Melissas 2009). The worldwide debate over legalizing drugs
depends, in part, on the feasibility of controlling the industry through the criminal law when law
enforcement authorities are vulnerable to corruption. Gambling, formerly outlawed in many American
jurisdictions, has become a legal business in many U.S. states, albeit under heavy state supervision
and even, at times, state ownership.
The bulk of the economics literature on corruption deals with government officials and politicians who
take or solicit payoffs, but corruption also occurs in private institutions. A sales agent may pay off a
purchasing agent in a large firm to get business or a labor union leader may extort payments from
corporate managers in return for labor peace. Some private-to-private corruption occurs because the
police are ineffective or corrupt. Then, organized crime groups may extort payoffs from shop owners
in return for protection from themselves and from rival gangs. See, for example, Chin (1996) on the
operation of gangs in New York City's Chinatowns and Varese (2005) on the operation of mafias in a
Russian city. Konrad & Skaperdas (1997, 1998) model these situations and analyze how police
monitoring can deter payoffs. However, as Gambetta & Reuter (1995) demonstrate, businesses may
benefit from organized crime connections if they succeed in frightening away potential entrants and in
producing monopoly profits to be shared between mafia members and firms.
The debate over anticorruption policy concerns not just the way corruption undermines efforts to
control illegal businesses but also the possibility that widespread corruption in one area will spill over
into other aspects of law enforcement, undermining respect for law enforcement authorities generally
and differentially attracting the potentially corrupt into a career in law enforcement (on the case of
Mexico see, for example, Brophy 2008). If the bureaucracy is highly corrupt, a state may not be able
to take on as many responsibilities as one with more honest and competent officials. Empirical
evidence shows that low levels of corruption are associated with high levels of public spending,
although, of course, the direction of causation is open to debate (Friedman et al. 2000). In any case,
the association is consistent with the hypothesis that the voting public will accept high taxes if the
government uses them effectively without extensive self-dealing.
Modern states, faced with corruption in their ranks and in the private sector, obviously cannot close
down entirely or legalize payoffs wherever they occur. The costs for society would be too high. Rather
they need to limit the corrupt incentives that produce harmful effects. This can be done through
program redesign, improved public oversight, and more effective law enforcement. Reform can limit
the discretion of officials through a move to clear, simple, and transparent rules combined with
monitoring by superiors and, crucially, a way for those harmed by payoff demands to complain to an
independent and honest official such as an ombudsman. 3
Many bribe payers, however, are better off in a corrupt than in an honest system. The briber obtains a
government contract even though his firm is a high-cost producer; a person obtains a driver's license
faster than his honest neighbor. In such cases, control must occur throughout the hierarchy, or the
program must be redesigned. For example, the government might purchase standardized items that
can be bought off the shelf without a specialized bidding process, or a government office might

streamline procedures to reduce everyone's wait. In addition, laws against bribery and extortion must
provide a credible deterrent to those who consider making and receiving payoffs.
Other work discusses the range of policy options besides law enforcement in some detail (Bardhan
1997; Campos & Pradhan 2007; Rose-Ackerman 1978, 1999, 2004, 2006). Here, I only wish to
highlight their diversity and interconnectedness. A narrowly focused reform may not limit corruption
unless combined with greater overall governmental transparency and outside monitoring. Particular
laws against bribery, extortion, and self-dealing will never be sufficient to deal with widespread
corruption. Fundamental redesign of the relations between the state and society will often be the only
way to control systemic corruption. Nevertheless, well-designed and enforced laws against bribery
and extortion are a necessary backup to any broader reform, and economic analysis can contribute to
the analysis of their operation and effectiveness. That is my focus in the remainder of this article.

All countries draw the line somewhere between illegal bribery and acceptable gifts of good will. Here, I
take that judgment as given and seek effective deterrence strategies using the criminal law. The
sanctioning strategies that are consistent with economic analysis often differ from the conventional
legal penalties, even in developed countries. A law and economics approach focuses both on
improving the deterrent effect of arrest and punishment and on providing incentives for people to
come forward with documentation of corrupt deeds. One conundrum for anticorruption efforts is the
possible tension between the goals of signaling credible expected punishments and using the law to
induce perpetrators to provide evidence.
Given the costs of law enforcement, the optimal level of corruption is not zero, even if one gives no
value to the benefits received by bribers.4 Once one takes the costs of prevention into account, the
level of deterrence expenditures should be set where the net benefits are maximized, that is, where
marginal benefits equal the marginal costs (Becker & Stigler 1974;Rose-Ackerman 1978, pp. 10919).
A higher level of deterrence would not be worth the extra costs; a lower level would sacrifice the net
benefits of increased enforcement. Economic analysis can aid in this assessment whether or not one
includes the benefits of malfeasance in the policy-analytic calculus.
The deterrence of criminal behavior depends on the probability of detection and punishment and on
the penalties imposedboth those imposed by the legal system and more subtle costs such as loss
of reputation or shame (Becker 1968). Low penalties upon conviction can still deter if the chance of
apprehension is high, and high penalties can compensate for weak enforcement as long as the
enforcement process itself is not unduly biased.
Successful detection of corruption depends upon insiders to report wrongdoing. Citizens and
businesses victimized by extortion demands may report bribery attempts, but they may not be able to
offer enough proof for prosecutors to act. Instead, effective law enforcement often requires officials to
promise leniency to one of the participants. This creates an important paradox for law enforcement
efforts. High expected punishments ought to deter corruption, but a high probability of detection may
only be possible if some are promised low penalties.
I begin by discussing deterrence based on expected punishment, measured by multiplying the
probability of apprehension by the punishment imposed. I then consider strategies that take account
of the interaction between punishment and the probability of apprehension. Then, I consider bribery
and extortion in law enforcement itself that may affect the enforcement of all types of law.
Because it takes two to enter into a corrupt deal, the transaction will not occur if the law can deter at
least one of the parties. Legal language frequently distinguishes between active and passive bribery
where the former refers to the briber and the latter to the bribee (see, for example,Council of Europe
1999, articles 2 and 3). This language seems to imply that bribe paying is worse than bribe
acceptance. However, both are generally criminal offenses, and most statutes impose parallel
punishments. National statutes and international conventions generally recognize that the distinction
between actively organizing a corrupt transaction and passively acquiescing is not a viable one.
Neither side is truly passive because both parties must agree before corruption can occur.
Furthermore, in practice, public officials might actively organize a corrupt bureaucracy that presses
citizens or firms to make payoffs.
However, in some countries asymmetries do exist. For example, in Taiwan paying off an official is only
a crime when the payment is made to obtain an illegal service. In all other cases, the payer is not
subject to criminal sanction.5 Under Romanian law, making a payoff is not a crime if the briber has
been coerced in any way by the one who received the bribe, and, in addition, such a briber can claim
restitution of his payments [Romanian Criminal Code, article 255(3)(5); discussed in Schroth & Bostan
2004, pp. 650, 661]. In other countries the reverse is true. For example, in Chile in the 1990s payment

of a bribe was a criminal offense, but accepting a bribe was not unless accompanied by other
wrongdoing (Hepkema & Booysen 1997, p. 415).6
Furthermore, the legal distinction between bribery and extortion is not straightforward, and in many
situations a person can be guilty of both (Lindgren 1988, 19921993). Statutes usually define
extortion without any specific reference to public officials. Coercive extortion can refer to payments
obtained by threats, whether made by an official, a mafia member, or a private individual. Violent
threats are often punished more severely than other types. 7 Extortion can occur in the United States
and England under color of office, a condition that need not involve an outright threat but is rather
associated with the bargaining power that comes from one's official position. Thus, in many cases the
official can be guilty both of accepting a bribe and of extortion (Lindgren 1988, 19921993).
The law and economics literature recognizes the two-sided nature of corrupt deals and does not refer
to active and passive bribery but rather refers to bribe payers and recipients. Some scholars
distinguish between the payer who receives better than fair treatment, on the one hand, or someone
who must pay to be treated fairly, on the other (Ayres 1997). More narrowly, Polinsky & Shavell
(2001) categorize extortion as a bribe paid to avoid being framed by an official for a trumped-up
offense. These bright-line rules are useful for analytic purposes, but they map imperfectly onto the
legal concepts. American law is quite confusing in defining these concepts (Lindgren 19921993), and
the proposed distinctions would be difficult to implement in practice. They require a clear benchmark
of fair treatment and raise difficulties when the person subject to extortion has, in fact, violated some
other law (Lindgren 1988, 19921993).
In deciding how to allocate law enforcement resources, the degree of social harm should be the key
variable, not the location of payoffs in the public or the private sector. In general, highest priority
should be given to preventing the allocation of illegal benefits or the imposition of illegal costs. For
legal benefits, the social costs depend on the damage done by using willingness-to-pay criteria and
the inefficiencies and inequities of officials' efforts to create bottlenecks and scarcity. These issues
loom especially large when officials or organized crime groups use threats of physical violence or
property damage (Konrad & Skaperdas 1997, 1998). Extortion that involves organized crime is
especially harmful for that reason. These groups seek to shift the reversion point for anyone who
resists to an outcome that is worse than the original status quo. Corrupt systems, especially if
supported by credible threats, have distributive consequences even if resources and services are
allocated efficiently. Corrupt officials share in the profits of private firms, and households may obtain
few benefits from a public program. The effects may be purely distributive, or they may have longterm impacts on entry into the corrupted businesses or activities.
A ranking of the social harm of different kinds of corruption, as outlined above, should help set
enforcement priorities. However, the penalties actually levied on the convicted need not be tied to
these social harms but rather should concentrate on the benefits received by the corrupt. To deter
bribery, at least one side of the corrupt transaction must face penalties that reflect its own gains.
Because the probability of detection and conviction is far less than 100%, those convicted should
sacrifice a multiple of these gains. From a pure deterrence point of view, either side of the corrupt deal
can be the focus of law enforcement efforts. From the point of view of public acceptability, however,
bribers who seek legal benefits may arouse public sympathy, not blame. Such offenses may be de
facto decriminalized through prosecutorial discretion. Whatever the focus, actors should face
expected penalties tied to their own benefit from corruption.
In practice, the briber and the bribee may bargain over the size of the bribe in light of the expected
penalty functions that each one faces. These functions depend both on the chance that the deal will
be uncovered and on the penalty levied on conviction. Most models of corruption do not include this
aspect of the problem and assume either a fixed bribe price or an equal division of the rents. For
example, if the briber faces a fixed maximum penalty X, while the bribee's expected penalty is an ever
increasing function of its gains, the division of the benefits will be affected by these differing
conditions. At some point the bribe recipient will reach his or her maximum bribe beyond which the
costs outweigh the benefits. In contrast, the briber in this example may be willing to contemplate very
large corrupt deals because, beyond some point, the penalties are not well tailored to the scale of the
deal. The law not only deters some bribery schemes altogether; it can also influence the division of
gains from corrupt deals.
To deter, officials' penalties should be an increasing function of the payoffs they receive and the
probability of detection. If expected penalties do not depend upon the size of the bribe, an
anticorruption drive would quickly confront a paradox. A high fixed penalty will lower the incidence of
corruption but increase the average level of bribes paid. Low bribes are not accepted, but once the
threshold is crossed, the penalty has no deterrent effect. If the penalty is high, officials must receive a
high return in order to be willing to engage in bribery. Thus, the expected penalty should increase by

more than a dollar for every dollar increase in the size of the bribe (Rose-Ackerman 1978, pp. 109
35; Shleifer & Vishny 1993). This could be done either by tying the penalty levied upon conviction to
the size of the bribe or by increasing the risk of apprehension as the size of the bribe increases.
However, if the probability of detection is lower for small payoffs, the penalty for each detected offense
must reflect that fact. This could mean that those convicted of petty bribery could face more severe
penalties than those found to have taken larger bribes. That outcome, however, is not likely to be
politically viable. Hence, two alternatives are possible: increasing surveillance and redesigning the
program to lower corrupt incentives by limiting the discretion of officials.
On the other side of the corrupt transaction, a fixed penalty levied on bribers will lower both the
demand for corrupt services and the incidence of bribes. However, as long as the probability of
apprehension does not depend on the size of the bribe, it will have no marginal impact once the bribe
passes the corruption threshold. If the probability of apprehension and/or the penalty rises with the
size of the bribe, then the level of individual bribes may fall as well. That result does not necessarily
follow, however, if higher bribes provide higher benefits. Suppose, for example, that the benefits to
bribery are an increasing function of the size of the bribe, such that a bribe of $1,000 provides benefits
of $1,500, but a bribe of $5,000 provides benefits of $20,000. Then expected penalties that are set at
twice the size of the bribe will deter the smaller bribe but not the larger one.
More fundamentally, bribes represent a cost to those who pay them; penalties for bribers should not
be tied to these costs unless they are a good proxy for the briber's benefits. In the example, they are
an imperfect proxy. To have a marginal effect, the penalties should be tied to the briber's gains (their
excess profits, for example), not to the size of the bribe. In the contracting area, if the potentially
corrupt firms are repeat players, one option is a disbarment procedure that prohibits corrupt firms from
contracting with the government for a period of years. To have a marginal effect, the disbarment
penalty should be tied to the seriousness of the corruption uncovered.
Under American law the maximum penalties are symmetric for those who make and those who accept
corrupt payments. The offender can receive a maximum sentence of three times the monetary
equivalent of the thing of value [given or promised to the official]or imprisoned for not more than
fifteen years, or both, and may be disqualified from holding any office of honor, trust, or profit under
the United States [18 U.S.C. 201(b)].8 Thus, the maximums do not explicitly recognize the
asymmetries in gains between bribe payers and recipients. However, the federal sentencing
guidelines do permit judges to incorporate the benefits received by bribe payers into their calculations.
According to the guidelines, the fine levied on an organization for bribery and related offenses should
be the greater of the value of the unlawful payment, the value of the benefit received or to be received
in return for the unlawful payment, and the consequential damages resulting from the unlawful
payment [U.S. Federal Sentencing Guidelines, 2C1.1(d)(1)]. Thus, it can either reflect the gain to
the firm or the loss to society. This seems a reasonable compromise with the principles outlined above
except for one glaring weakness. It does not account for the fact that, ex ante, the chance of being
caught is far less than 100%. To properly deter, the penalty should be a multiple of the gain to the firm.
The statute itself permits a fine that is three times the bribe paid, but it does not mention the benefit to
the bribe payers that may far exceed even that total, especially when viewed ex ante.
In the sentencing guidelines, the base penalty for individuals is several months in prison and is not
tied directly to the level of benefits received by paying or accepting a bribe. The penalty is increased if
a public official is bribed, if more than one bribe is involved, or if the value exceeds $5,000 [U.S.
Federal Sentencing Guidelines, 2C1.1(a)(c)]. In addition, civil fines can be imposed on those
convicted of bribery and related offenses. The fine can be either a payment of up to $50,000 or the
bribe amount if it exceeds $50,000. Thus, individuals can be punished with both fines and jail time, but
it does not seem that the marginal increases exceed the marginal benefits for large bribes once one
takes account of the low probability of apprehension. It may be that the penalties have little deterrent
effect on large bribes.
The law, however, does permit the president to rescind any contract or other benefit if there has been
a conviction under the statute governing bribery, graft, and conflicts of interest. The United States can
also recover, in addition to any penalties, the amount expended or the thing transferred or delivered
on its behalf, or the reasonable value thereof (18 U.S.C. 218). This right of recovery is designed to
avoid losses to the government. It is a weak deterrent to corrupt payoffs because the recovery is not
multiplied by a factor that reflects the probability of detection.
Outside the United States, the legal penalties bear only a weak relationship to the deterrence priorities
outlined above. Of the cases examined, admittedly not a comprehensive list, both those who pay and
those who accept bribes face the possibility of fines and imprisonment and, as in the U.S. statute, the
maximum penalties are generally symmetric for both groups. The statutes fail to link penalties either
to the social harm of corruption or to the private benefits obtained by those who engage in bribery. In

some statutes there are special increased penalties for aggravated instances of corruption, but these
are trigger strategies not explicitly tied to marginal gains and losses. It seems quite likely that largescale corrupt deals are only slightly deterred by the formal legal penalties. 9
If expected penalties do not increase along with the benefits of corruption for bribers and bribees,
government may be caught in a trap where high corruption levels beget high corruption levels. An
equilibrium with low corruption may also exist but be unreachable in small steps from the status quo.
High corruption can be a stable equilibrium when the net rewards of corruption increase as the
incidence of corruption increases. This might occur, for example, if law enforcement officials discover
a smaller proportion of corrupt deals when the incidence of corruption is high and if penalties levied
upon conviction are not adjusted to take account of that fact. Any multiple equilibria case, however,
can be converted into a single equilibrium, low-corruption case with the appropriate choice of law
enforcement strategy or a change in the information conditions. Strategies that tie expected penalties
to marginal gains can remove a society from a high corruption trap. Doing so, however, may require a
large increase in law enforcement resources to tip the system to a low-corruption equilibrium.
Fortunately, such a sharp increase in enforcement resources need not be permanent. It must simply
be sufficient to tip the system to a lower corruption level (Lui 1986, pp. 2122). The idea is to change
expectations. A concentrated cleanup campaign can change expectations about others' cooperation in
the corrupt system. Once a new low-corruption equilibrium has been established, it can be maintained
with reduced enforcement resources as long as the honest are willing to report corrupt offers and law
enforcement officials follow up on reports of malfeasance they do find (Andvig & Moene
1990; Bardhan 1997, pp. 133034; Cadot 1987; Rose-Ackerman 1978, pp. 13751). One problem
with this optimistic scenario, however, is the possibility that the corrupt will collude with each other to
lie low during the crackdown and reemerge later to resume their corrupt activities. The models
summarized here assume, in contrast, that there is no collusion and that both officials and bribers act
on the basis of their most recent past experience.
Finally, consider efforts to deter private-to-private corruption. Some legal systems criminalize privateto-private bribery, but in many jurisdictions such transactions are not against the law unless they
involve another illegal offense, such as extortion or operation of an illegal business. 10 One needs to
ask if private tort law or employment law is sufficient to deter such conduct. Is the cost of private-toprivate corruption merely the harm to the bribe taker's employer, and if so, why aren't private law
remedies, motivated by the employer's profit motive, sufficient? There are two linked issues. First, the
threat of criminal penalties may provide a more effective deterrent than civil penalties or a firm's
internal disciplinary procedures. The possibility of a prison sentence and the stigma of a conviction
may deter, even if the higher standards for proof and the defendants' greater procedural protections
mean that conviction is less likely. Second, some private-to-private bribery can have broader systemic
consequences. If widespread, it may spill over into interactions with government bureaucrats, and it
may provide a way for a firm to cement a monopoly position that harms other customers and
suppliers. Bribes may be paid not just to get business but also to dilute product quality, enforce
cartels, and limit entry. Then the motivation for enforcing the law in this area goes beyond its direct
impact on private business and merges with the discussion of public sector corruption.
Gathering Evidence: Promising Leniency and Rewarding Whistleblowers
Effective deterrence is impossible unless law enforcement authorities can obtain relevant evidencea
difficult task because often the participants are the only ones who know of the corrupt deal. In such
cases, the probability of detection is a function of whether any of the participants has an incentive to
report. Those subject to extortion may report attempts, but police promises of low penalties or even
rewards are often essential. However, such tactics are frequently criticized by anticorruption
commentators as inconsistent with the goals of the criminal law. For example, the Group of States
Against Corruption (GRECO), a European group, criticized a provision in Romanian law that
exempted bribers from punishment if they inform authorities of their activities before a formal
investigation begins [Romanian Criminal Code, article 255(3)(5)]. The GRECO report recognized that
this could be a means to gather evidence and initiate criminal proceedings against officials, but it
worried that it would weaken enforcement of the law against active bribery (GRECO 2002).
Standard work on the economics of crime does not confront the problem of obtaining evidence from
perpetrators. Detection is uncertain but independent of criminals' actions. That assumption does not
hold for bribery and extortion, for which a prime source of evidence is information obtained from those
engaged in corrupt transactions.
Suppose, first, that the benefit obtained in return for a bribe is legal and would be freely available in an
honest world to those who now pay bribes. Because the bribers receive benefits to which they are
legally entitled, they believe that they are extortion victims who would be better off in an honest world.
Such bribe payers are potential allies in an anticorruption effort and will likely cooperate in efforts to

eliminate payoffs. They should not be punished heavily because leniency will give them an incentive
to report corrupt demands and will encourage beneficiaries of public programs to demand services
that are free of payoff demands.
Consider next a scarce but legal benefit that is corruptly allocated to many individuals who would not
qualify if payoffs were eliminated. Neither those who pay nor those who receive bribes will voluntarily
report the corrupt transaction. Those shut out of the process, however, have a grievance. For
example, disappointed bidders for public contracts can facilitate efforts to limit corruption (Alam 1995).
They should be rewarded for coming forward with evidence even if the reason they lost the bid was
not moral scruples but their own unwillingness to make a large enough payoff. The reward offered
need not equal their lost benefits from losing the contract because one consequence of revealing
corruption will be a rebid contract in which the whistleblower can compete.
Bribes paid to obtain illegal services are likely to be the most difficult to control. Bribers are often also
engaged in other illegal activities, and those who fail in their corruption efforts can hardly come
forward to claim that they should have been the ones obtaining the illegal benefit. Nevertheless, the
very vulnerability of bribers can be used to uncover corruption. They may accept lenient treatment
with respect to, say, a violation of the drug laws, in return for providing evidence in a corruption trial.
Here, the law with respect to cartels in restraint of trade can provide useful parallels. Leniency for the
first firm that comes forward to report a cartel is a common feature of the law in the United States and
in Europe, and analysts study ways to structure the rewards to maintain the deterrence effect of
punishment. Thus, one study recommends increasing average penalties and making leniency
conditional on the quality of the evidence provided (Wils 2007).
An alternative system protects and rewards whistleblowers who come forward with evidence of
wrongdoing. Reporting the peculations of others can be dangerous. If corruption is systemic, one risks
being disciplined by corrupt superiors and attacked by coworkers. One study of corruption in China
suggests that this is a serious problem (Manion 2004). The whistleblower may even end up accused
of corruption. The United States has two statutes. The False Claims Act rewards those in the private
sector who report irregularities in government contracts and protects whistleblowers from reprisals (31
U.S.C. 37293731, Howse & Daniels 1995, Kovacic 1996). The act pays whistleblowers a share of
the total penalties and other damages levied against firms for wrongdoing that has injured the federal
government. The second protects whistleblowers inside government agencies from retaliation but
does not give them a financial reward [Whistleblower Protection Act, Public Law No. 101-12, 5 U.S.C.
2302(b)(8)]. Such statutes can help prevent malfeasance as long as they do not induce potential
whistleblowers to create compromising situations and as long as the search for misdeeds does not
lower the quality of public services. Thus, the value of such statutes depends both on the likelihood of
corruption and on the opportunity cost of whistleblowing activity. (For a somewhat different model of
whistleblowing that exploits a similar trade-off between ex ante effort and ex post revelation, see Ting
Sometimes public officials claim that firms virtually force bribes upon them. To the extent this claim is
credible, public officials could come forward with evidence of corrupt offers and seek protection under
the Whistleblower Protection Act. Firms would predictably defend themselves by arguing that the
official demanded the payoff. The distinctions in American law may be useful here. Under the False
Claims Act, the court can reduce the award for a whistleblower who was involved in wrongdoing, but
only if he or she planned or initiated the wrongful conduct. The award need not be eliminated,
however, unless the whistleblower is convicted of a crime [31 U.S.C. 3730(d)(3) (2006)].
Prosecutors with the authority to grant criminal immunity can thus set up a race in which the first to
report the corrupt transactions will be rewarded, while the others are punished. Alternatively, the law
might impose penalties on anyone who is part of a corrupt deal and who failed to report it promptly.
Then all participants would worry that the other participants will report their corruption to authorities as
a way of avoiding future penalties. Just as with a reward system, the idea is to create a race to the
prosecutors that will deter corruption ex ante.11
Both carrots and sticks, however, depend on the existence of a credible system of law enforcement
that might discover the corrupt deal on its own. The use of undercover sting operations can be
leveraged into a tool that encourages those offered or pressured for bribes to come forward. If they do
not, they know that the corrupt offer may be a trap set by law enforcement authorities. Although the
defendant may raise the defense of entrapment, this is rarely successful in corruption cases, at least
in the United States, and is especially unlikely to succeed when the defendant offers a bribe (69
A.L.R.2d 1397).12 The U.S. FBI has made good use of sting operations in its efforts to ferret out
domestic corruption and has recently done the same thing under the Foreign Corrupt Practices Act
(FCPA, in 15 U.S.C. 78). The most famous case is probably the Abscam sting, involving an
elaborate hoax that snared several members of Congress into corrupt transactions involving a

supposed Middle Eastern sheikh (for an overview that cites many of the background documents,
seeGrossman 2003, pp. 13). More recently, a sting led to 22 arrests in January 2010 for alleged
violations of the FCPA. The assistant attorney general in charge of the case stated that the
undercover techniques used in this case should cause all would-be FCPA fraudsters to pause and to
ask: am I really paying off a foreign government official or could this be a federal agent? (quoted
in Scarcella 2010). Carefully orchestrated sting operations can be a valuable tool as long as the
criminal behavior they target is relatively clear-cut.
Corruption in Law Enforcement
So far I have assumed an honest system of law enforcement that can deter corruption. Unfortunately,
corruption in law enforcement itself is widespread and can affect the incidence of all types of crime.
Survey evidence from Peru, for example, shows that the incidence of reported bribery is highest
among those who use the judiciary, followed by those who interact with the police (Hunt 2006).
Obviously, if such corruption reduces the expected costs of breaking other laws, it will encourage
criminal behavior (Becker & Stigler 1974, Bowles & Garoupa 1997, Polinsky & Shavell 2001). The
optimal deterrence strategy should take into account and seek to deter the corruption of enforcers,
such as the police (Bowles & Garoupa 1997). This could be done through a mixture of organizational
and personnel reforms combined with better oversight.
However, it is at least possible that a corrupt system operates so that the cost of breaking the law is
higher under a corrupted law enforcement system as officials exercise their opportunities for illicit
private gain. Potential corrupt payments can be thought of as a common pool in which the police all
try to fish for private gain. No individual policeman takes account of the fact that his or her extraction
of payoffs limits those available to others. They all race to collect payoffs, and their uncoordinated
actions may lead them to overfish for corrupt rents, thus deterring other forms of illegality (Pashigian
Polinsky & Shavell (2001) model law enforcers as having three options: first, soliciting bribes from
offenders not to report a violation (or to reduce the sanction); second, extorting a payment from an
innocent person by threatening to frame him; and, third, actually framing the innocent. Thus, their
distinction between bribery and extortion, although it is too precise to reflect the actual use of these
terms in the criminal law, captures the distinction between potential payers who are innocent or guilty
of the underlying offense. The control of corruption in law enforcement is justified in their model
because corruption limits the deterrent effect of the law. They follow work on the law and economics
of crime that recommends the use of maximal fines up to the criminal's wealth constraint (Becker &
Stigler 1974). They recommend the use of fines because they are much less resource intensive for
the government than putting people in prison and should have equivalent deterrent effects for
defendants who are not judgment proof. Becker & Stigler demonstrate the social value of maximally
fining both law enforcers and offenders who engage in bribery as well as enforcers who frame the
innocent.13 Of course, law enforcement officials, especially in developing countries, are not well paid
and cannot be deterred by fines that exceed their family's wealth.
Their most surprising conclusion is that pure extortion should not be punished. This is because its
punishment will push corrupt enforcers either to frame the innocent or to demand higher extortion
payments from the innocent. This result arises from the structure of their model, in which the victims
of extortion have no escape route. They are caught in a trap where outside law enforcement can do
no more than raise the costs to extortionists, hence encouraging them to demand higher payoffs.
For Garoupa & Klerman (2004), bribery is a second-best way to introduce monetary penalties when
nonmonetary sanctions (e.g., imprisonment) predominate. They show that, in the presence of
corruption, nonmonetary sanctions perform the useful function of generating high bribes that are paid
to avoid prison. In practice, ordinary crime is deterred by the expectation of bribe payments. In a fully
corrupt system no one goes to prison, and the state benefits from the resulting cost savings. This
model, of course, assumes that bribe demands actually deter crime rather than simply being a way of
sharing the monopoly profits of criminal activity between the police and the criminals. As Garoupa
himself recognizes in later work, it neglects the possibility that the unequal opportunities for payoffs
across different types of policework will distort enforcement priorities (Echazu & Garoupa 2010).
The economic models of corruption in law enforcement are also consistent with research that looks
favorably on privatizing law enforcement. Under this view, tolerating the bribery of enforcers is worse
than policies that legally incentivize enforcers by giving them private incentives to work diligently to
enforce the law. For example, Becker & Stigler (1974) andPolinsky & Shavell (2001) suggest the
potential of bounties paid to enforcers for successful law enforcement activity. However, privatizing
law enforcement through corruption could be a desirable second-best result in a world where
enforcers must be paid a fixed salary. That tolerant view, however, seems extremely optimistic. If
penalties are low, bounty hunters might induce people to submit to arrest in return for a share of the

bounty.14 If one adds in the possibility of coercive extortion and intimidation combined with threats of
violence, the quasi-privatization of enforcement, as Polinsky & Shavell recognize, increases the
incentives to extort and frame the innocent. Their analysis points to the importance of establishing
credible checks so that those subject to extortion or framing have a place to complain. Moves toward
a legal bounty system or toleration of corruption could tilt the entire system into a violent, rent-seeking
free-for-all. The system could end up dominated by powerful criminal mafias able to organize corrupt
payoffs and intimidate rank and file officials and ordinary citizens, whether or not they engage in
criminal activity. An extreme alternative, of course, is to legalize the criminal activity, thus removing the
incentive to corrupt the police and the courts. More modestly, privatization does not appear to be a
desirable option given the downsides highlighted here.

Bribery frequently occurs in organizational hierarchies and in structures with multiple potential payoff
sites. These organizational complexities add an important dimension to the analysis. One needs to
consider not just the relationship between a single agent, a single honest principal, and a potentially
corrupt outsider but also the possibility that individual bribe payments are part of a complex system. I
consider government hierarchies first and then analyze public systems with multiple potential
corruption points. Next, I assess the impact of middlemen who are formally outside the public sector
but perform an intermediary role. Finally, I consider the organizational issues that arise in a firm when
an employee bribes a public official.
Government Hierarchies
Some corruption occurs at the top of a government hierarchy where officials sign major procurement
contracts, award concessions, and privatize state firms. In those cases the analysis above applies,
but deterrence can be especially difficult because the corruption involves high-level political actors
with the power to set enforcement priorities and create corrupt opportunities (Rose-Ackerman 1978,
pp. 10936). Bribe demands not only determine how officials and firms divide a fixed pool of rents, but
also influence firm behavior. The contracting process may be less competitive as honest firms stay
away and corrupt firms collude to push up the cost of contracts. Furthermore, firms that face
extortionary demands may adjust their investment choices away from highly capital-intensive
techniques that will leave them subject to future bribe demands (Rose-Ackerman 1999, pp. 27
38; Choi & Thum 2004).
In another common pattern, bribes are paid at the bottom of the bureaucratic pyramid, but low-level
officials share them with superiors. Those at the top organize the entire hierarchy as a bribe-sharing
machine and may pressure those farther down the chain both to become part of the corrupt system
and, once recruited, to pass on a share of their take up the hierarchy. Low-level officials, who may be
difficult to monitor in their day-to-day activities, may be given bribery quotas that must be paid to
superiors as a condition of maintaining employment or in return for good working conditions.
Superiors can subvert standard anticorruption proposals to their own interests. For example, the civil
service might have a policy of rotating officials to different locations throughout their careers as an
anticorruption strategy that prevents them from establishing close local ties. In a corrupt hierarchy, in
contrast, higher-ups can use this personnel policy to punish those who do not join the corrupt system
by sending them to poverty-stricken and remote locations. 15
In such a system, a reforming chief executive might remove and punish the official at the top of the
corrupt hierarchy and install an honest agency head. For example, a newly elected mayor might
remove a corrupt police chief and install a person of known integrity. Will such an action limit
corruption or will it simply decentralize bribery and make it less visible? The answer depends on the
nature of low-level corruption and on the role of those higher up the hierarchy. A corrupt police chief
may have more power to extract rents and better information about their level than any other
individual official. In that case centralization will be especially harmful (Rose-Ackerman 1978). Such a
top official may be able to increase overall payoffs by, for example, giving low-level officials petty
monopoly power, increasing the level of red tape or the discretion of officials, issuing threats to
potential bribers that enhance the extortionary power of underlings. Then, removing a corrupt head
ought to limit the level of corruption and the harm that it causes. A new, honest agency head can
spearhead reforms to limit corrupt opportunities.
In contrast, if the corrupt top official simply presides over a bribe extraction machine, then replacing
him will not change the underlying structural features that produce corrupt incentives. However, the
policy could still have a positive effect if an honest chief can instill norms of honesty in subordinates
and improve monitoring and oversight. In the worst case, suggested by Olson (1993), Shleifer &
Vishny (1993), Choi & Thum (2004), and Olken & Barron (2009), deterring corruption at the top might
increase it at the bottom as officials compete and overfish the pool of rents, leading to greater social
harms. Centralized corruption produces an inefficient and perhaps unfair system as the bribery

monopolist maximizes rents, but it avoids the risk of complete breakdown. Under a decentralized
system, where officials do not have to share their gains with superiors, more street-level officials might
become corrupt. This assumes, of course, that corrupt superiors were simply rent extractors who
provided few benefits to inferiors and merely threatened and intimidated them.
Multiple Corruption Points
But pyramids are not the only organizational form relevant to the control of corruption. A second
possibility is an activity that requires the potential briber to interact with different public officials. These
may be in a fixed order, as when a trucker faces multiple checkpoints on a highway, or they may not
have a fixed order, as when a business needs multiple permits to operate. Call these sequential and
fragmented systems, respectively (Rose-Ackerman 1978, pp. 16773).
The operation of a sequential corrupt system where applicants must approach officials in a fixed order
depends on the precise model specification. Thus, one model assumes that applicants Nash bargain
with officials and share the surplus in a manner determined by their exogenously determined
bargaining power that assures them a share x of the total. In that model, if all officials have the same
bargaining power, the last official can extract the highest bribe because that is where the surplus is
highest. Olken & Barron (2009) illustrate the basic intuition with a simple example using their empirical
study of truckers in Indonesia. Suppose that there are two checkpoints and that the value of a
successful delivery is V. At any checkpoint, the official can confiscate the shipment at zero benefit to
both. At the last checkpoint, all past bribes are sunk costs, and the choice is between paying a bribe
of xV or getting zero. Moving back one step, the first official can only extract a bribe of x(1 x)V. In
other words, the trucker and the first official, looking ahead to future checkpoints, anticipate the
second official's bribe demands. Their empirical work confirmed this prediction for one of the trucking
routes in their study (Olken & Barron 2009).
Lambert-Mogiliansky et al. (LMR) (2007, 2008), using a Nash equilibrium model of sequential bribe
demands, show that, under some conditions, no applicant even starts the process because all believe
that they will lose money (see also Yoo 2008). Olken & Barron only study the behavior of truckers
already on the road. They do not try to determine if the volume of trucking has been affected. Sunk
cost, however, plays the same role in both models; at each stage in the sequence, the
applicant/trucker looks ahead to future checkpoints to determine the level of rents at any checkpoint.
However, in the LMR model officials are more powerful but less knowledgeable. They do not know the
value the applicant places on bureaucratic approval, but they will try to extract the entire surplus in
expected value terms, and this produces cases where the bribe demand exceeds the surplus. There
are no honest bureaucrats in the LMR models, so the applicant cannot gamble on the chance of
confronting officials who do not demand bribes. Hence, in the one-shot case no project is ever
approved, and all applicants, anticipating what is to come, simply stay home. From the point of view of
conformity with the law, this can be beneficial if applicants are not legally qualified to obtain the
benefit, but it is harmful in case they are qualified. Olken & Barron's truckers are, in contrast, all
qualified to use the roads, although most trucks can be legally fined for being overweight, a factor that
influences the size of their bribe payments. If the process is repeated, however, players remember
their own actions and those they dealt with, and they learn what other actors have done. Here, there
can be equilibrium paths of normal bribes where any defection would cause the whole corrupt system
to unravel (LMR 2007, Yoo 2008). These equilibria impose different degrees of social harm depending
on the pattern of bribes.
In contrast, under other model specifications, the first official might have the most bargaining power
and extract most of the rents. That result could occur if the bargaining power of subsequent officials is
itself a function of the remaining level of positive profits. Officials and applicants suffer from the sunk
cost fallacy in the sense that once the first official has taken most of the surplus, subsequent officials
accept the argument that the applicant has no more profits to share. This argument might work
especially well if the applicants are repeat players who can credibly threaten not to return to the
corrupt system if charged too much.
In the fragmented case in which the order is not fixed in advance and applicants can keep the order
secret, this will reduce the take of officials. In the Olken & Barron model, no official would know that
he is all that stands between the applicant and final approval, and hence he cannot extract the last
checkpoint rent. Similarly, in the second, sunk cost fallacy case, if the order is not predetermined, no
official will know that he is the first in line. The result could be a complete breakdown in the system as
officials compete for bribes and discourage applicants.
This theoretical and empirical research sheds some light on two contrasting policy proposals. The first
advocates a one-stop-shop for the registration of businesses or other bureaucratic approval
processes. This is equivalent to the centralized solution in a hierarchical system. In both cases, as
discussed above, a single official replaces a multiplicity of potentially corrupt individuals, and the key

question is whether overfishing occurs in a decentralized system. If so, a centralized system, even if
corrupt, will be less harmful. Alternatively, if centralization goes along with greater power over
applicants for legal services, the one-stop-shop may simply permit greater rent extraction in a way
that distorts the allocation of resources and the fairness and legitimacy of public programs. In the
intermediate case analyzed by LMR (2007), a one-stop-shop may permit officials to consolidate their
extortionary power. Thus, a one-stop-shop may either decrease or increase bribe payments
depending upon the equilibrium situation that prevails in the absence of that reform. Recall, however,
that in their model, officials are corrupt and face no risk of punishment for their actions. They extend
their analysis to the closely related case where the applicant hires an intermediary to obtain the
needed permits (LMR 2009).
The second proposal goes in the other direction and advocates the appointment of multiple officials
only some of whom must be approached. This structure contrasts with models such as those
developed by LMR (2007, 2008) and Olken & Barron (2009) where the applicant-trucker must satisfy
a fixed sequence of officials. In the competitive model, officials compete for a share of the bribes, and
this behavior may constrain the overall level of payoffs. Even in that case, of course, the
anticorruption benefits of competition among officials depend on the nature of the public benefits
(Rose-Ackerman 1978, pp. 13766). Overlapping jurisdictions can limit the level of bribes paid to get
legal benefits and hence discourage officials from asking for handouts. Similarly, if bribes are paid to
avoid arrest for engaging in illegal activity, overlapping jurisdictions both lower the bribes that any
official can demand and increase the chance that lawbreakers will be caught. Rather than needing a
series of officials to approve before an activity can go forward, any one of several officials can stop an
illegal activity. For example, police may have overlapping beats. Then no one police patrol can obtain
a high bribe from an illegal gambling operation because another patrol may come along later and
either demand a second bribe or arrest the operators. However, in that case the police themselves
have an incentive to collude to extort payoffs from gambling establishments.
Agents and Middlemen
A familiar feature of many corrupt systems is the private middleman with connections who promises to
smooth one's route through the bureaucracy for a price. The payments are used both to bribe public
officials and to compensate the agent. Many countries' bribery statutes criminalize payments to such
agents as well as the acceptance of payments by agents. However, there is considerable crosscountry variation.16
Why are such agents so common even when everyone recognizes that they are paying bribes? Why
not eliminate the middleman and pay the official directly? The key point is that middlemen are repeat
players compared to most of those seeking the benefit. They function in systems in which applicants
either have little or no recourse to honest complaint mechanisms or are seeking something valuable
to which they are not entitled. The middleman is often either a former official or a current one on an off
day. He or she knows the going rate for the service, can save the applicant time by eliminating a wait
in line and/or the need for bothersome extra visits to government offices. Especially in fragmented
systems, they can cut through complex official procedures. They seem to be performing a useful
service by speeding up bureaucratic processing and reducing the time and hassle for citizens. Notice,
however, that the better this corrupt system works, the greater the incentive of the officials and the
middlemen to work together to increase the time and trouble imposed on honest citizens as an
inducement to corruption. They also have an incentive to refuse service to the honest but qualified
applicants or even to invent offenses.
A recent paper by Hasker & Okten (2008) draws on earlier efforts by Bayar (2005), Graf Lambsdorff
(2002), and Oldenburg (1987). In their model, intermediaries undermine such standard enforcement
techniques as increased monitoring or higher penalties. Rotating bureaucrats through different offices,
far from limiting corruption, can increase the impact of intermediaries, resulting in increased corruption
(Hasker & Okten 2008). Because they are not civil servants, they may be especially hard to control. A
policy of firing corrupt officials may backfire. For example, as described in Fjeldstad (2003, p. 172),
the Tanzanian government launched an anticorruption campaign by firing one-third of the bureaucrats
in the tax administration. Private businesses hired these former bureaucrats because of their
knowledge and insider contacts. New corrupt networks soon emerged. What seemed like a simple
solution increased the problem because the government ignored the market for intermediaries
(Hasker & Okten 2008, p. 114).
Two experimental studies in India illustrate how such systems can work. One studied the issuance of
drivers' licenses, often to people unable to drive, and the other the issuance of ration cards for
subsidized food (Bertrand et al. 2006, Peisakhin & Pinto 2010). Both showed how the use of
middlemen speeded up processing, but the ration card experiment had an additional twist. Applicants
who filed a Freedom of Information Act request concerning their application under a new Indian law

received relatively speedy service and did not need to pay a middleman. This result, if it can be
reproduced in other contexts, suggests a promising avenue for reform.
Business Firms
Now consider the other side of the corrupt transaction in which employees of business firms pay
bribes to public officials for benefits such as contracts, lower tax bills, or exemptions from health and
safety regulations. These transactions all benefit the firm by increasing its profits and may be
worthwhile even if detected as long as expected penalties are sufficiently low. Thus, one cannot rely
on the firm's owners and top managers to control this type of bribery. They may express strong ethical
principles, but they only have a strong incentive to control commercial bribery that damages the firm's
bottom line, such as payments to their own employees from sales agents. If payoffs help a firm obtain
business, managers and owners may hope to facilitate their subordinates' bribery while remaining
ignorant of the details (Rose-Ackerman 1978, pp. 189209). If corporations are held criminally liable
for the corrupt acts of their employees and agents, top management may not support an effective
monitoring system. However, if firms escape criminal liability, it may be difficult for the criminal law to
impose credible punishments.17
Arlen (1994, Arlen & Kraakman 1997) analyzes this problem for the general case of corporate crimes
and concludes that various alternative rules are superior to present U.S. law that imposes pure strict
liability on firms. One possibility is a negligence rule under which firms are only liable if they have
neglected their internal enforcement responsibilities. For such a rule to be workable, however, courts
must be able to evaluate internal firm behavior, a difficult task. One solution may be quite precise
directives stating what type of internal monitoring is required with checks to be sure it is carried out in
good faith.
The UK will face exactly this challenge under the new 2010 Bribery Act that permits commercial
organizations to defend against allegations of bribery by showing that they have adequate
[preventive] procedures in place [United Kingdom Bribery Act 2010, chapter 23, 7(2)]. Similarly, the
U.S. FCPA supplements its prohibitions against paying bribes to obtain business abroad with
accounting provisions that apply to firms within the jurisdiction of the Securities and Exchange
Commission. These firms must establish accounting systems that accurately reflect transactions
involving the firm's assets, and they must have an effective system of internal accounting controls.
Firms and their managers can be subject to both civil and criminal penalties for violating these
accounting provisions [FCPA, 15 U.S.C. 78m(b) (1988 & Supp. IV 1992), sect. 102; Jadwin &
Shilling 1994, pp. 67980; Nobles & Maistrellis 1995, pp. 9, 19; Pickholz 1997, p. 237]. Unlike the new
UK statute, the FCPA has no formal due diligence defense. But in practice, firms that establish and
enforce effective internal control systems appear to experience more lenient treatment. The Federal
Sentencing Guidelines also reward internal firm efforts to detect and punish violations of the law
(Nobles & Maistrellis 1995). Thus, both the American and the UK laws attempt to counter
management's incentives to insulate itself from the profit-maximizing malfeasance of employees and

Law and economic analysis can help set priorities for anticorruption campaigns, suggest reforms in
programs riddled with corruption, recommend law enforcement priorities, and help to design workable
law enforcement strategies. Economic analysis counsels reformers to concentrate on the underlying
conditions that produce rents. Criminal law should not be the first or the only line of attack. Rather,
programs should be redesigned or eliminated, and government should operate in an accountable and
transparent manner. That said, it remains true that law enforcement against bribery and extortion has
a backup role to play. Economic analysis can help with the design of law enforcement strategies.
Looking at a range of cases, it appears that the laws on the books are often quite far from the
recommendations of law and economics. Penalties seem poorly tied to the marginal benefits of
bribery both to those who pay and to those who receive bribes. Small bribes seem to be more
effectively deterred than larger ones, unless prosecutorial discretion makes up for the legal language.
The penalties levied on bribers are not well tied to their gains. The tension between obtaining
evidence to bring a case and deterrence ex ante has seldom been recognized and has been
imperfectly resolved. We do not know if giving corrupt individuals leniency in return for their evidence
and testimony limits corruption or encourages people to participate ex ante. The role of organizational
structure, including the status of middlemen and agents, has not been well incorporated into
enforcement programs, and many countries have difficulty deterring organizations as opposed to
individuals, in part because the criminal law only applies to individuals. Although the fundamental
tasks for reformers are the redesign of public programs, improvements in the motivations and
incentives facing public officials, and greater government transparency and accountability, reform of

the law of bribery and extortion remains a necessary, if not sufficient, area of reform where economic
analysis can help guide the debate

The legal framework

3.30 Domestic legislation, making bribery and corruption of public officials a
crime, was a well-established
feature in many countries from the end of the 19thCentury onwards14. For a
large part of the 20th
Century, however, bribery and corruption of foreign public officials by companies,
could be considered as
an issue for the country concerned and, if accepted practice in that country,
considered part of the cost
of doing business. It could be, if not openly, then tacitly condoned by
companies home countries.
3.31 With the recognition of the insidious damage bribery and corruption causes
to nations, particularly
developing nations, there were considerable legal developments through the
latter part of the 20th
Century. The US was the prime mover with the passing of the Foreign and
Corrupt Practices Act in
1977. This was followed by the Organisation of Economic Cooperation and
Development (OECD)
Convention on Combating Bribery and Foreign Public Officials in 1997, which has
so far been
implemented by 37 countries. The prohibition of bribery and corruption has thus
become a well
established, and increasingly important, global standard15
3.32 Today, the ethical and reputational risks associated with bribery are at acute
levels. Any defence
company (indeed any company) must have a clear and explicit policy against
bribery and the control
procedures, training, audit and investigatory measures in place to prevent and
detect bribery by, or
of, an employee.

3.33 Increased attention, particularly in the defence industry, has been focused
on indirect corruption by third
parties with whom a company has a contractual relationship. Any analysis of past
and present cases of
transnational bribery demonstrates that third parties have very frequently been
instrumental to the
corruption. There is a distinction between a third party who is actively
encouraged to bribe by the
company on its behalf, or to whom a blind eye is turned, and a third party who
acts covertly and in
breach of the companys rules without its knowledge. Nevertheless, the clear
trend in regulatory
requirements, particularly in the US, is to impose responsibility, and by
implication ethical and
reputational risk, upon the contracting company irrespective of whether it knew,
or ought to have
known, of corrupt practices by third parties.