by Elias L Ohneberg, CERF Research Associate, Cambridge Judge Business School, University of Cambridge
In a 2012 survey by Penn Schoen Berland and Georgetown University, 75% of senior executives responded that they witnessed favouritism at work, and 83% stated that it can lead to the wrong people being promoted. More recently, an article by the New York Times uncovered blatant favouritism by Bank of America’s chief operating officer. His favoured employees received bonuses and promotions, while disfavoured employees were put up for pay reductions. Given the prevalence of favouritism, it is important to understand its effect on employee incentives and on-the-job performance.
The effects of favouritism on employee incentives and behaviour
Causal empirical evidence on the effects of favouritism on employee incentives and behaviour has so far been sparse. The field of psychology offers some correlative evidence that personal relationships between supervisors and employees can impact performance appraisals and firing and promotion decisions. Causal empirical research in the field of economics and finance is practically non-existent. Work by Prendergast and Topel (1993), nevertheless, provides theoretical support for a negative effect of favouritism on employee effort provision and forms a basis for my empirical work.
To investigate the effects of favouritism on employee incentives and behaviour more generally, I study the effect of workplace connectedness on 13,347 portfolio managers in the mutual fund industry over a quarter of a century. The mutual fund industry lends itself well to studying the effect of workplace connections on employee incentives. It is highly human capital intensive and sizable, managing $60.1 trillion worldwide.
First, I construct a measure that quantifies mutual fund managers’ connectedness within their company. Two mutual fund managers are defined to know each other if they have a past work relationship – they have co-managed a mutual fund in the past. Connections between 2 employees can have different importance. Being more connected to a senior employee is likely more important than to a junior employee. Therefore, the measure places a higher weight on connections to more senior employees.
To link the notion of favouritism to the connectedness of employees I need to establish that well-connected employees receive some sort of preferential treatment. I do this by investigating whether better-connected employees are treated differently in firing and promotion decisions, controlling for objective performance measures that should impact these decisions. My findings suggest that better-connected mutual fund managers are less likely to be fired for poor performance and more likely to be promoted despite lack of good performance. Surprisingly, my analysis suggests that past performance does not matter at all for well-connected managers. These results are consistent with preferential treatment towards well-connected managers.
The threat of being fired and the reward of being promoted are important incentive mechanisms. They punish people for shirking and reward them for a job well done. Thus, my finding that past performance has a smaller impact on the promotion and firing probabilities for well-connected managers suggests that the incentive effects typically provided by firings and promotions are hampered. If performance does not matter as much for firing and promotion decisions, do well-connected mutual fund managers exert less effort in managing their funds? This is the next question I investigate.
To ascertain effort provision by mutual fund managers, I can look at different measures. The job of a portfolio manager is to generate profitable investment ideas and to implement them successfully to generate a high return on investment. Coming up with these investment ideas requires time and effort spent on researching novel investment ideas and implementing them into the portfolio. Instead of exerting effort to generate a novel investment approach, mutual fund managers could simply copy what their peers are doing. Thus, my first measure of effort is how distinctively a mutual fund is managed in comparison to its peer group. I find that mutual funds managed by well-connected managers are less distinct in their sector allocations. This finding suggests that well-connected managers indeed expend less effort in managing their funds.
Moreover, if well-connected mutual fund managers are exerting less effort in managing their funds, we would also expect their mutual funds to perform worse. Therefore, I next investigate whether mutual funds managed by better-connected managers perform worse than their counterparts. I find that mutual funds managed by better-connected managers exhibit poorer risk-adjusted investment performance than mutual funds managed by worse-connected managers.
Why is favouritism tolerated?
My findings, thus far, suggest that connectedness at the workplace can induce favourable treatment in promotion and firing events. This favourable treatment hampers the incentivisation usually offered by these career-altering events, and induces lower effort provision and, ultimately, poorer mutual fund performance.
Given the negative effects uncovered in the analysis, there is a final question that needs to be addressed. Why is it that mutual fund companies allow this sort of favouritism to exist? It is somewhat puzzling that this value-destroying behaviour is not stopped. There are essentially two reasons why this behaviour would persist. First, there is simply nothing the mutual fund company can do to prevent the favourable treatment of well-connected managers. Second, the mutual fund company does not have an incentive to prevent it. Due to the difficulty of directly empirically investigating the first question, I will focus on the latter.
To investigate if mutual fund companies have an incentive to tackle favouritism, we have to look at how mutual fund companies make money. Mutual fund companies are compensated through fee revenue. This fee revenue comes directly from annual fees paid by investors to the mutual fund company. Mutual fund fees are typically not tied to performance but instead consist of a fixed annual percentage fee of the money invested in the fund. Therefore, the aggregate fee revenue of a mutual fund is the size of the fund multiplied by the annual percentage fee. I focus my efforts on investigating how the connectedness of mutual fund managers impacts the size and net money inflows into the fund. I find that connectedness does not impact the overall size of the fund or the ability of the mutual fund to attract new money from investors. This would suggest that mutual fund investors are not aware of the negative effects of connectedness on fund performance or are not able to observe the connectedness of fund managers. Furthermore, it offers a rationale as to why mutual fund companies may not prevent the favourable treatment of well-connected managers despite its negative consequences on performance and effort provision, more generally.
Overall, this study uncovers the notion of favouritism in the mutual fund industry. It shows that connected managers receive advantageous treatment in career-altering decisions unrelated to objective performance measures. This favourable treatment hampers incentivisation mechanisms and induces lower effort-taking by mutual fund managers. Mutual fund companies do not bear the costs of this value-destroying behaviour, however. Instead, mutual fund investors suffer through poorer investment performance.
CERF post-doctoral researcher Kevin Schneider studies the effect of seasonal output prices on a firm’s inventory and stock returns. He finds that seasonal inventory holdings help explain several stock return anomalies.
Marwin Mönkemeyer, Research Associate at Cambridge Centre for Finance (CCFin) and Cambridge Endowment for Research in Finance (CERF), blogs about the role of social trust in the portfolio allocation decisions of global institutional investors.