Manipulating Statistics to Justify Self-Serving Recommendations

Abstract

We investigate how the ability to manipulate statistics affects advisors’ willingness to provide self-serving recommendations to advisees. We build on theories in deception and persuasion stating that people have preferences to both be honest and appear honest, which can sometimes conflict with their financial self-interest. We present several findings using an abstract experiment. First, when given the opportunity, advisors actively manipulate statistics to support recommendations that benefit themselves. Second, more favorable statistics are associated with more self-serving recommendations (i.e., advisors rely on favorable statistics to justify their recommendations), and this association holds regardless of whether or not the statistics are shown to advisees. Furthermore, we find that statistics manipulation is associated with advisors’ tendency to deceive themselves, which may facilitate their ability to internally rationalize providing a self-serving recommendation. Overall, the findings from our study have theoretical and practical implications for various behavioral accounting contexts. Data Availability: The data used in this study are available upon request. JEL Classifications: D8; M4.