文摘
This dissertation is comprised of three essays. In the first essay, titled, "Do Bonuses Enhance Sales Productivity? A Dynamic Structural Analysis of Bonus-Based Compensation Plans," we look at how various components of the compensation plan affect the sales performance of heterogeneous sales agents.
This research has two main methodological innovations: First, we implement empirically the method proposed by Arcidiacono and Miller (2011) to accommodate unobserved latent class heterogeneity with a computationally light two-step estimator. Second, we estimate discount factors in a dynamic structural model using field data. The key to identification of discount factors is that bonuses affect only future payoff in non-bonus periods providing exclusion restrictions on current payoffs. Further, we exploit differences in predicted effort over time from the exponential and hyperbolic discounting models to identify present bias.
Substantively, this study sheds insights on how different elements of the compensation plan enhance productivity. We find evidence that: (1) bonuses enhance productivity across all segments; (2) overachievement commissions help sustain the high productivity of the best performers even after attaining quotas; and (3) quarterly bonuses help improve performance of the weak performers by serving as pacers to keep the sales force on track to achieve their annual sales quotas. We also find clear evidence of hyperbolic discounting by salespeople.
In the second essay, titled, "The Effects of Incentive Framing on Productivity: Evidence from a Field Experiment," we look at the effectiveness of incentive framing and gift exchange on worker productivity. We conducted a six month-long field experiment, varying incentives at a Chinese high-tech manufacturing firm. We model and estimate a dynamic structural model of worker behavior, with the experimental treatments allowing its to generate variation to identify incentive framing effects. In addition, to empirically test the effectiveness of gift exchange, we examine the value of unconditional incentives and its effect on individual worker performance. In order to adequately account for individual heterogeneity, we estimate the model using a hierarchical Bayesian framework with an embedded dynamic programming problem within the likelihood to infer the choice of effort by individual workers.
Overall, incentives increase productivity. We find significant evidence of lass-aversion with incentives being 6.7-percent more effective when framed as a penalty for not achieving quota, than as a bonus for reaching quota. We also find concrete evidence of dynamic behavior on behalf of the agents with a mean daily discount rate of 1.6-percent. We find unconditional incentives to he effective, increasing average productivity of workers, but find the effect to he heterogeneous across individuals. Specifically, we find agents with low initial goodwill to be more affected by unconditional incentives and find that goodwill created by unconditional incentives substantially decays with time. We find that under the first best scheme, allowing different incentive plans for individuals, the firm can increase profits as much as by 35-percent.
In the third essay, titled, "The Dynamic Advertising Effect of Collegiate Athletics," we measure the spillover effect of intercollegiate athletics on the quantity and quality of applicants to institutions of higher education in the United States, popularly known as the "Flutie Effect." We treat athletic success as a stock of goodwill that decays over time, similar to that of advertising. A major challenge is that privacy laws prevent us from observing information about the applicant pool. We overcome this challenge by using order statistic distribution to infer applicant quality from information on enrolled students.
Using a flexible random coefficients aggregate discrete choice model - which accommodates heterogeneity in preferences for school quality and athletic success - and an extensive set of school fixed effects to control for unobserved quality in athletics and academics, we estimate the impact of athletic success on applicant quality and quantity.
Overall, athletic success has a significant long-term goodwill effect on future applications and quality. However, students with lower than average SAT scores tend to have a stronger preference for athletic success, while students with higher SAT scores have a greater preference for academic quality. Furthermore, the decay rate of athletics goodwill is smaller for students with lower SAT scores, suggesting that the goodwill created by intercollegiate athletics resides more extensively with low-ability students than with their high-ability counterparts. But, surprisingly, athletic success impacts applications even among academically stronger students. Finally, athletic success effects are greater at public schools than at private schools.
