اثرات تشویقی مالیات پاداش در مدل عامل اصلی

Several countries have implemented bonus taxes for corporate executives in response to the current financial crisis. Using a principal-agent model, this paper investigates the incentive effects of bonus taxes by analyzing the agent's and principal's behavior. Specifically, we show how bonus taxes affect the agent's incentives to exert effort and the principal's decision regarding the composition of the compensation package (fixed salary and bonus rate). We find that, surprisingly, a bonus tax can increase the bonus rate and decrease the fixed salary if the agent is highly risk averse. Additionally, a bonus tax can induce the principal to pay higher bonuses even though the agent's effort unambiguously decreases. Nevertheless, a bonus tax reduces the overall salary of the agent. Further results are derived with respect to the existence and uniqueness of the equilibrium for a general effort cost function.

Despite the large body of literature and numerous theoretical and empirical studies on executive compensation, only a few papers have addressed the consequences of executive compensation regulation in general and the effects of bonus taxes in particular. For example, Dew-Becker (2009) reviews the history of government rules and regulations in the US that affects executive compensation. By discussing disclosure rules, advancements in corporate governance, and say-on-pay, Dew-Becker analyzes the evolution of pay regulation and concludes that mandatory say-on-pay could be the most effective and least harmful measure of controlling executive compensation. Knutt (2005) examines diverse regulatory issues from a legal point of view. He claims that the various attempts to regulate executive compensation, such as the disclosure and tax regulations, have not yet been effective. Hall and Liebman (2000) analyze the extent to which tax policy influences the composition of executive compensation and discuss the consequences of rising stock-based pay.1 Their empirical study shows that the dramatic explosion in executive stock-option pay since 1980 cannot be attributed to tax rate changes. Moreover, the so-called million dollar rule induced a substitution from fixed salary toward performance-related pay. Unlike Hall and Liebman (2000), who concentrate on a tax on stock-based pay, we study a tax that is levied on the agent's bonus. Radulescu (2010) analyzes the effects of bonus taxes in a two-country, principal-agent model with relocation possibilities for the managers. The paper focuses on tax incidence and analyzes the effects of bonus taxes on firm profits, dividends and welfare in the case of a quadratic effort cost function. The paper shows that a bonus tax induces lower profits and dividends so that the incidence is borne by the shareholders. The welfare implications of bonus taxes depend on the relocation possibilities for the managers. In contrast to our model, in which we focus on the incentive effects of bonus taxes for a general cost function and derive an ambiguous effect of a bonus tax on the fixed and variable salary, Radulescu (2010) finds that the effort-based compensation component (bonus) unambiguously increases with a higher bonus tax.2 Finally, there is now a growing literature on regulating incentive pay in the financial sector.3 For example, Hakenes and Schnabel (2012) show that the presence of bailout guarantees induce bankers to increase their risk-taking behavior and lead to a steeper compensation scheme. An upper limit on the bonus could alleviate these problems. Bolton et al. (2010) develop a theoretical model to show that the credit default swap reduces risk taking of executives at highly levered financial firms. Based on a model of workers in the financial sector, Besley and Ghatak (2011) show that bailouts induce lower effort and higher risk taking.

This paper has investigated the consequences of taxing the agent's bonus in a principal-agent model. A bonus tax has different effects on the fixed salary and the bonus rate because the principal sets an optimal compensation package on the conditions that the agent is expected to reach at least his reservation utility and the package is incentive compatible. To determine the composition of the package, the principal anticipates that a higher bonus tax will increase the leakage in the contracting environment because she incurs the full cost of the bonus payment, while the agent receives only a fraction of it. This leakage effect makes the use of bonuses more costly for the principal. However, a higher bonus tax decreases the risk premium required to compensate the agent for the uncertainty in his expected salary due to a lower salary variance. This risk effect provides the principal with incentives to increase the bonus rate. Whether the bonus rate increases or decreases through the introduction of bonus taxes depends on the relative size of these two effects. Our model further shows that a bonus tax does not necessarily induce the principal to substitute variable salary with fixed salary for a general effort cost function. In particular, it is possible that a higher bonus tax simultaneously increases the fixed salary component and the bonus rate. Moreover, we derive that the agent unambiguously reacts to a higher bonus tax with lower effort. For quadratic effort costs, despite a tax-induced effort reduction, a higher bonus tax induces the principal to pay higher bonuses if the agent is sufficiently risk averse and/or the variance in the firm value is sufficiently large. In this case, the increase in the bonus rate overcompensates for the decrease in the agent's effort. Therefore, it is not guaranteed that firms will have to pay lower bonuses after implementing a bonus tax. Nevertheless, a bonus tax reduces the overall salary of the agent so that a bonus tax proves to be an effective policy instrument in reducing the agent's salary. Hereby, the reduction of the agent's salary is more pronounced the lower the risk parameter. To sum up, a bonus tax influences both the overall size and the structure of the agent's salary. For example, if the agent is highly risk averse, then the reduction in the agent's gross salary induced by a bonus tax is relatively small and, in addition, the principal shifts the compensation package from the fixed salary to the variable salary. However, if the agent is not too risk averse, then a bonus tax induces a relatively large reduction in the gross salary and also shifts the compensation package towards the fixed salary. Our model yields potentially testable comparative-static results. In particular, our model might help to predict the sectors and firms in which bonuses would decrease or increase. In uncertain economic environments and/or in firms, in which the monitoring and evaluation of the manager's performance is comparatively hard, we expect that bonus taxes will induce a firm to pay higher bonuses. Additionally, we do not expect bonus taxes to have the effect of shifting the bonus rate towards a fixed salary. That is, we anticipate low fixed salaries and high bonus rates. Moreover, our model predicts that the overall salary will only experience a relatively small reduction. We can derive the following examples of firms for these predictions: (i) new-economy firms: These firms tend to operate in more uncertain economic environments than old-economy firms. In addition, it might be more difficult to observe the manager's marginal contribution in new-economy firms because these firms grow faster, are more R&D intensive and have larger market-to-book ratios than old-economy firms (see Ittner et al., 2003). (ii) Large firms: according to Schaefer (1998), large firms have more noisy measures of individual performance than small firms. Moreover, in large firms, one manager's action has less influence on the firm value than it might have in small firms. (iii) Privately held firms: Marino and Zabojnik (2008) suggest that it is harder for privately held firms to evaluate their managers because a public firm's stock price provides an informative measure of performance which is less available in privately held firms. Our simple model may serve as a basic framework to further analyze bonus taxes in a principal-agent model. There is a broad range of further applications and model extensions. For instance, an interesting avenue for further research could be the extension of our model to more than one period. An agent's effort decisions are often connected over time, and working contracts extend over several time periods. The implementation of these dynamics in the model could shed more light on the impact of bonus taxes on executive pay. Furthermore, interesting extensions would be to analyze the effects of bonus taxes on executives’ risk-taking behavior or to study the welfare implications of bonus taxes.