AVRAMOVICH MARÍA CECILIA
Libros
Título:
ESSAYS ON CARTEL BEHAVIOUR AND DETERRENCE POLICY
Editorial:
UNIVERSIDAD CARLOS III DE MADRID
Referencias:
Año: 2012 p. 144
ISSN:
0-0000-0000-0
Resumen:
This book consists of three chapters on the impact of deterrence policies for corporate crimes in the strategic behavior of economic agents. It combines topics from Industrial Organization and Law and Economics, with theoretical grounds from Microeconomics. The first chapter analyzes the impact of the antitrust policy on deterring collusion and the productive efficiency of surviving cartels when concealing collusion is costly. This chapter considers an antitrust policy based on fines, inspections and leniency programs. The second chapter analyzes the impact of disclosing information on the likelihood of inspections on the sustainability of cartels. The third chapter analyzes the effectiveness of rewarding whistle-blowing to deter occasional corporate crimes, as well as its impact on the optimal contract between the firm owner and the whistle-blower.
In the first chapter, entitled "What if Fines on Collusion are not high enough? Implications on Deterrence and Productive Efficiency when Concealment is Costly", I analyze the impact of the antitrust policy on the firm´s decisions over crime and production. In the model, cartel firms devote costly effort to activities related to productive efficiency and concealment, as the former reduces marginal costs from production and the latter reduces the probability of detection. Effort is costly, and so that limited; thus cartel firms have to decide on how to allocate it among productive efficiency and concealment. In this context, a first analysis refers to how the antitrust policy distorts firms´ interest in productive efficiency with respect to that in concealment. I show that when fines are low and/or the probability of inspection is low, firms find it profitable to allocate all effort to productive efficiency. However, as fines or inspections go up, firms substitute effort from productive efficiency to concealment. This reallocation of effort makes collusion sustainable in industries where it wouldn´t be otherwise and create inefficiencies not considered in standard models of collusion.
In the light of these results, a second analysis refers to the impact of a policy change in deterrence and welfare. I show that a fine increase can have two opposite effects on welfare, while it can improve welfare through fewer cartels, it can also reduce it through more inefficient surviving ones. Particularly, for intermediate fine levels, a fine increase can imply a welfare gain from fewer cartels that does not compensate the welfare loss from more inefficient surviving ones. This result favors setting very high fines such that no cartel survives. However, in practice this is not always credible or possible to implement. In this context, I show that deterrence is non-monotonic neither in the level of fines nor in that of inspections individually considered: an increase in any of them may enhance collusion sustainability rather than deterrence if the other instrument is not changed accordingly. Therefore, the main policy recommendation is that the antitrust policy has to be carefully designed, such that combining both instruments, fines and inspections, conveniently.
Finally, the analysis favors the use of leniency programs. Since leniency programs demand full collaboration from the reporting firm, a leniency application implies no effort on concealment from that firm, and thus an efficiency gain under deviation.
In the second chapter, entitled "An Inspector Calls: On the Optimality of Warning Cartel Firms before an Inspection", I extend the previous model by introducing a new policy in the game: the Antitrust Authority can credibly disclose information on the likelihood of a current inspection before firms take strategic decisions on price, production and cartel activities. I consider that the Antitrust Authority performs this policy by sending warnings to firms with a high probability of a current inspection.
The possibility to take decisions on the basis of a more accurate probability of inspection affects cartel firms incentives to collude in two opposite ways. On the one hand, it raises collusion profitability: with accurate information on the likelihood of a current inspection, firms can minimize profit losses from (i) devoting costly effort to concealment anytime that they do not receive a warning, and from (ii) an unprofitable effort allocation towards productive efficiency each time that they receive a warning. In this case, incentives to collude go up and collusion is facilitated. However, on the other hand, the deviants also benefit from this information; this enhances incentives to deviate and makes collusion harder to sustain.
In this context, the main question is whether the introduction of a warning program can improve deterrence (i.e., whether the latter effect described above can prevail over the former). I show that disclosing information on the likelihood of a current inspection improves deterrence when both the fine and the probability of inspection are high. For such parameter values the program does not distort much the behavior of a firm that follows the collusive agreement, but it does so for a deviant. Indeed, faced with high expected costs, the program does not imply much for the former: this allocates all its effort to concealment (a) always, if there is not a warning program, and (b) almost always, if there is. However, for a deviant the program implies a lot. Since periods of no inspection are few, accurate information about when one is in one of these is of huge relevance: in these periods deviation not only implies the standard higher gains from sales, but also minimum (no) detection costs. Hence, in this context, whereas firm´s benefits derived from the program are little under collusion, there are huge under deviation; and so the program implies an improvement in deterrence.
In the third chapter, entitled "Rewarding Whistle-Blowers: Implications on Deterrence and on Principal-Agent Contracts", I analyze the impact of rewarding whistle-blowers on the deterrence of corporate crimes and the internal efficiency of firms. In the model, there is a principal that owns a firm with two employees. Both employees are hired to devote costly effort to production, but one of them can secretly devote effort to commit corporate crime ("the offender") and the other to gather evidence of such an act ("the agent"). The crime not only yields private gains to the offender employee, but also an externality to the firm owner.
In this context, I demonstrate that whistle-blower programs that offer a reward to non-offender employees for the public exposure of corporate crimes can improve deterrence: rewards induce non-offender employees to gather crime evidence, which raises the probability of crime detection; faced with a higher probability of detection (and thus with lower net gains from crime), the offender´s willingness to commit crime go down. However, I also show that this reward policy has side effects on the productive efficiency of firms: if the principal´s net externalities from crime are very high, this will overpay effort devoted to production in an attempt to bias the agent´s effort allocation away from crime detection. The opposite will hold when the principal´s net externalities from crime are very low (highly negative). Two extreme cases stand out: one in which the principal hires the agent only for his activities related to gathering crime evidence, and other in which, to minimize crime detection, he does not hire the agent. The latter one implies that rewards create a total loss of welfare.
Finally, I demonstrate that it may be in the principal´s interest to create a reward-program private to the firm, regardless of the existence of a (public) whistle-blower program. A program of this type arises when crime is highly detrimental to the principal and, if there exists a whistle-blower program, when also the public reward is not high enough to deter crime as much as the principal would like. Moreover, I demonstrate that the use of a private reward-program implies maximum deterrence. This result suggests the implementation of public whistle-blower programs to prosecute crimes that work in favor of the interest of firm owners (e.g., tax-frauds, collusion, environmental crimes, etc.).
In the first chapter, entitled "What if Fines on Collusion are not high enough? Implications on Deterrence and Productive Efficiency when Concealment is Costly", I analyze the impact of the antitrust policy on the firm´s decisions over crime and production. In the model, cartel firms devote costly effort to activities related to productive efficiency and concealment, as the former reduces marginal costs from production and the latter reduces the probability of detection. Effort is costly, and so that limited; thus cartel firms have to decide on how to allocate it among productive efficiency and concealment. In this context, a first analysis refers to how the antitrust policy distorts firms´ interest in productive efficiency with respect to that in concealment. I show that when fines are low and/or the probability of inspection is low, firms find it profitable to allocate all effort to productive efficiency. However, as fines or inspections go up, firms substitute effort from productive efficiency to concealment. This reallocation of effort makes collusion sustainable in industries where it wouldn´t be otherwise and create inefficiencies not considered in standard models of collusion.
In the light of these results, a second analysis refers to the impact of a policy change in deterrence and welfare. I show that a fine increase can have two opposite effects on welfare, while it can improve welfare through fewer cartels, it can also reduce it through more inefficient surviving ones. Particularly, for intermediate fine levels, a fine increase can imply a welfare gain from fewer cartels that does not compensate the welfare loss from more inefficient surviving ones. This result favors setting very high fines such that no cartel survives. However, in practice this is not always credible or possible to implement. In this context, I show that deterrence is non-monotonic neither in the level of fines nor in that of inspections individually considered: an increase in any of them may enhance collusion sustainability rather than deterrence if the other instrument is not changed accordingly. Therefore, the main policy recommendation is that the antitrust policy has to be carefully designed, such that combining both instruments, fines and inspections, conveniently.
Finally, the analysis favors the use of leniency programs. Since leniency programs demand full collaboration from the reporting firm, a leniency application implies no effort on concealment from that firm, and thus an efficiency gain under deviation.
In the second chapter, entitled "An Inspector Calls: On the Optimality of Warning Cartel Firms before an Inspection", I extend the previous model by introducing a new policy in the game: the Antitrust Authority can credibly disclose information on the likelihood of a current inspection before firms take strategic decisions on price, production and cartel activities. I consider that the Antitrust Authority performs this policy by sending warnings to firms with a high probability of a current inspection.
The possibility to take decisions on the basis of a more accurate probability of inspection affects cartel firms incentives to collude in two opposite ways. On the one hand, it raises collusion profitability: with accurate information on the likelihood of a current inspection, firms can minimize profit losses from (i) devoting costly effort to concealment anytime that they do not receive a warning, and from (ii) an unprofitable effort allocation towards productive efficiency each time that they receive a warning. In this case, incentives to collude go up and collusion is facilitated. However, on the other hand, the deviants also benefit from this information; this enhances incentives to deviate and makes collusion harder to sustain.
In this context, the main question is whether the introduction of a warning program can improve deterrence (i.e., whether the latter effect described above can prevail over the former). I show that disclosing information on the likelihood of a current inspection improves deterrence when both the fine and the probability of inspection are high. For such parameter values the program does not distort much the behavior of a firm that follows the collusive agreement, but it does so for a deviant. Indeed, faced with high expected costs, the program does not imply much for the former: this allocates all its effort to concealment (a) always, if there is not a warning program, and (b) almost always, if there is. However, for a deviant the program implies a lot. Since periods of no inspection are few, accurate information about when one is in one of these is of huge relevance: in these periods deviation not only implies the standard higher gains from sales, but also minimum (no) detection costs. Hence, in this context, whereas firm´s benefits derived from the program are little under collusion, there are huge under deviation; and so the program implies an improvement in deterrence.
In the third chapter, entitled "Rewarding Whistle-Blowers: Implications on Deterrence and on Principal-Agent Contracts", I analyze the impact of rewarding whistle-blowers on the deterrence of corporate crimes and the internal efficiency of firms. In the model, there is a principal that owns a firm with two employees. Both employees are hired to devote costly effort to production, but one of them can secretly devote effort to commit corporate crime ("the offender") and the other to gather evidence of such an act ("the agent"). The crime not only yields private gains to the offender employee, but also an externality to the firm owner.
In this context, I demonstrate that whistle-blower programs that offer a reward to non-offender employees for the public exposure of corporate crimes can improve deterrence: rewards induce non-offender employees to gather crime evidence, which raises the probability of crime detection; faced with a higher probability of detection (and thus with lower net gains from crime), the offender´s willingness to commit crime go down. However, I also show that this reward policy has side effects on the productive efficiency of firms: if the principal´s net externalities from crime are very high, this will overpay effort devoted to production in an attempt to bias the agent´s effort allocation away from crime detection. The opposite will hold when the principal´s net externalities from crime are very low (highly negative). Two extreme cases stand out: one in which the principal hires the agent only for his activities related to gathering crime evidence, and other in which, to minimize crime detection, he does not hire the agent. The latter one implies that rewards create a total loss of welfare.
Finally, I demonstrate that it may be in the principal´s interest to create a reward-program private to the firm, regardless of the existence of a (public) whistle-blower program. A program of this type arises when crime is highly detrimental to the principal and, if there exists a whistle-blower program, when also the public reward is not high enough to deter crime as much as the principal would like. Moreover, I demonstrate that the use of a private reward-program implies maximum deterrence. This result suggests the implementation of public whistle-blower programs to prosecute crimes that work in favor of the interest of firm owners (e.g., tax-frauds, collusion, environmental crimes, etc.).
