Corporate criminal liability and optimal firm behavior: internal monitoring versus managerial incentives

Abstract

Legislation addressing corporate criminal liability has been the subject of worldwide debate ever since the financial scandals of the early 2000s. Under current regimes, firms must observe such compliance requirements as internal monitoring mechanisms, the purpose of which is inducing firms to detect the wrongful conduct of their agents. We develop an analytical framework for identifying when, and to what extent, firms may find it beneficial to adopt these regulatory devices. We conclude that more productive firms, those operating in sectors with more market power, and firms whose managers have more opportunities for criminal activity are more likely to prevent wrongful conduct—either through monitoring or the payment of efficiency wages. When the potential returns to illegal activities are high or the firm is large, internal monitoring is probably the optimal strategy of crime prevention; in contrast, smaller firms typically proceed by paying efficiency wages (or ignoring crime). This paper also analyzes the role of the State’s legal capacity as well as the effects of interactions between the structure of reputational losses and the firm’s market power.

Appendix

1.1 Proof of lemma 3

We start determining the conditions under which $S=M$ is a MPE when paying efficiency wages is dominated by monitoring, i.e. $x\ge x^{*}$ (see Lemma 2). In doing this we use the one-shot deviation principle (Fudenberg and Tirole 1991) which requires that the value of the firm from the prevention strategy with monitoring at all dates (i.e., $S_{s}=M$ for all $s\ge t$) is higher than the value from a deviation where the firm adopts the strategy of non-prevention at date t and then switches back to monitoring forever from date $t+1$ (i.e., $S_{t}=N$ and $S_{s}=M$ for all $s>t$):

$$\begin{aligned} W(M,G_{0})\ge & {} \alpha \left\{ p\left[ x\left( y-w_{0}-F\right) +\beta W(G_{1})\right] +\left( 1-p\right) \left[ x\left( y-w_{0}\right) +\beta W(M,G_{0})\right] \right\} \nonumber \\&+(1-\alpha )\left[ x\left( y-w_{0}\right) +\beta W(M,G_{0})\right] . \end{aligned}$$

(47)

The left hand side of (47) is given by (16). The right hand side comes the fact that the strategy of non-prevention is chosen at date t: crimes are feasible and undertaken with probability $\alpha$ and, then, are detected with probability p; when crimes are detected the firm suffers a loss F and his continuation value is $W(G_{1})$ reported in (22). With probability $1-p$ crimes are not discovered when committed and with probability $1-\alpha$ they are not feasible; therefore, in both cases, the firm gets the static payoff $x\left( y-w_{0}\right)$ and the value of prevention with monitoring $W(M,G_{0})$ from date $t+1$ onward.

Substituting (16) and (22) into the last expression, we obtain that condition (47) can be rewritten as $\phi \ge \phi _{M}^{*}$ with $\phi _{M}^{*}$ defined in (25). This proves the result that $S=M$ is a MPE for all $\phi \ge \phi _{M}^{*}$.

We now apply the one-shot deviation principle to determine when the non-prevention strategy dominates the monitoring one. This will also allow us to show the uniqueness of the MPE. In particular, we determine the conditions under which the value of the firm from non-prevention at all dates (i.e., $S_{s}=N$ for all $s\ge t$) is higher than the value of a deviation where the firm adopts the monitoring strategy at date t and then switches back to non-prevention forever from date $t+1$ (i.e., $S_{t}=M$ and $S_{s}=N$ for all $s>t$):

$$\begin{aligned} W(N,G_{0})\ge x\left( y-w_{0}\right) -k+\beta W(N,G_{0}). \end{aligned}$$

(48)

Substituting (23) into (48) and solving for $\phi$ leads that $S=N$ is a MPE for all $\phi <\phi _{M}^{*}$ with $\phi _{M}^{*}$ in (25). This also shows that the MPE at point (i) of Lemma 3 is unique.

Point (ii) of Lemma 3 can be shown in a similar way. We first determine the conditions under which $S=E$ is a MPE when the payment of efficiency wages is always preferred to monitoring, i.e. $x<x^{*}$. Again, using the one-shot deviation principle we find conditions under which the value of the firm from the prevention strategy with efficiency wages at all dates (i.e., $S_{s}=E$ for all $s\ge t$) is higher than the value from a deviation where the firm adopts the strategy of non-prevention at date t and then switches back to paying efficiency wages forever starting from date $t+1$ (i.e., $S_{t}=N$ and $S_{s}=E$ for all $s>t$):

$$\begin{aligned} W(E,G_{0})\ge & {} \alpha \left\{ p\left[ x\left( y-w_{0}-F\right) +\beta W(G_{1})\right] +\left( 1-p\right) \left[ x\left( y-w_{0}\right) +\beta W(E,G_{0})\right] \right\} \nonumber \\&+(1-\alpha )\left\{ x\left( y-w_{0}\right) +\beta W(E,G_{0})\right\} . \end{aligned}$$

(49)

From the substitution of (18) and (22) in the last expression we obtain that (49) holds when $\phi \ge \phi _{E}^{*}$ with $\phi _{E}^{*}$ reported in (26).

We now determine the condition under which the non-prevention strategy dominates the efficiency wage payment. Hence, using the one-shot deviation principle, we compute under which conditions the value of the firm from non-prevention at all dates (i.e., $S_{s}=N$ for all $s\ge t$) is higher than the value of a deviation where the firm adopts the strategy of efficiency wage at date t and then switches back to non-prevention forever from date $t+1$ (i.e., $S_{t}=E$ and $S_{s}=N$ for all $s>t$):

$$\begin{aligned} W(N,G_{0})\ge x\left( y-w_{e}\right) +\beta W(N,G_{0}). \end{aligned}$$

(50)

From the substitution of (23) into (50) it follows that $S=N$ is a MPE for all $\phi <\phi _{E}^{*}$ with $\phi _{E}^{*}$ in (26). This also implies that the MPE at point (ii) of Lemma 3 is unique.

Finally, it is worth noting that the same results are obtained by comparing the values of the strategies assuming these are followed at each date, i.e. comparing (23) with respectively (18) and (16).