Open-ended advice is ubiquitous, yet we know little about who is good at giving it. We introduce a method for measuring the causal effect of individual advisors on the performance of the people they advise. In a preregistered online experiment, 157 advisors learned hidden information about an abstract task and provided audio advice to 64 advisees (\(N=10,048\)). We find that a one SD increase in advisor quality raises advisee performance by 0.05 SD (p=0.015) and that advisor effects predict the performance of held-out advisees. This 'good advisor' effect size is twice as large as the difference in task performance between college-educated and other advisees. We find that being good at the task is unrelated to being good at advising on it. Moreover, advisees' impressions are informative: more effective advisors are rated as sounding more trustworthy. Methodologically, we find that LLM rubric scores of advice quality are positively correlated across two diverse advice-giving tasks, but we detect no relationship between rubric scores and advisors' causal impact. Our results highlight the value of outcome-based measures of advising skill. Substantively, we use our approach to explore gender heterogeneity and find evidence that the quality of advice matters more for mixed-gender pairs.
Feedback plays a critical role in shaping beliefs, guiding decisions, and improving performance. We conduct an online experiment to study the nature and effectiveness of qualitative feedback. Although qualitative feedback is widely used, it has received little attention in experimental economics, where the primary focus has been on quantitative feedback. We find that subjects interpret the open-ended qualitative feedback they receive properly and update their beliefs appropriately. In particular, they do not incorporate the feedback in a way that is unduly favorable to themselves. This contrasts with previous work on quantitative feedback which finds an upward distortion in updating. The providers of written feedback exhibit a kindness effect: their feedback is more positive than their true opinions. We find no difference in how feedback is given to men and women, unlike prior studies. We explore two channels through which feedback influences decisions: a belief-updating channel and a content channel. Women respond to both, while men are less responsive to the content channel. We find improved performance for individuals who receive feedback that includes more concrete advice. Our experiment uses the entire performance-feedback sequence.
This paper analyzes the relative efficiency of horizontal and vertical organizational structures modeled as communication networks under limited managerial attention. In this framework, workers receive noisy signals about the state of the world and communicate reports to a decision-maker whose meeting time relatively more constrained. The clarity of each report depends endogenously on meeting lengths and the workers' heterogeneous communication and processing abilities. When workers possess identical communication abilities, the horizontal structure is always optimal because it allows the decision-maker to weight signals optimally without compounding transmission noise. Conversely, when workers have asymmetric communication abilities, the attention constraint of the decision-maker induces a structural switch. Under these constraints, the vertical structure becomes preferred because the decision-maker can delegate the time-intensive task of information aggregation to the superior communicator, balancing the loss from transmission noise against the premium on the decision-maker's scarce time. Under the assumption of homogeneous processing abilities I derive the analytical threshold for the n-worker model.
Not only do men and women choose different professions at dissimilar rates, within many professions, men and women often specialize in different jobs. In fact, there are many jobs that, in the minds of many, are clearly associated with one gender. One potential explanation for these differences is that women and men possess different innate abilities, they specialize accordingly, which then impacts employers' hiring decisions. Yet, there is mounting evidence that differences in ability cannot account for the observed degree of specialization along gender lines. An alternative explanation is that gender stereotypes impacts both workers and employers decision-making, although identification of this effect is difficult with observational data. To address this in a labor market experiment with candidates and employers, we randomly induce one of two arbitrary stereotypes about gender differences in performance and a control with no performance information. In the experiment candidates' specialize and then employers assign roles. Employers cannot observe the candidates' abilities but can infer them through the candidates' specialization decision and their gender. In the first part of the experiment when there is no performance information gender is ignored, candidates specialize according to their ability and employers hire according to specialization. However, with arbitrary gender stereotypes, gender is no longer ignored, employers faced with equally specialized candidates base their assignment decisions on the prevailing stereotype. Anticipating this, candidates tend to specialize in the role stereotyped for their gender, and thus, the initially arbitrary stereotype becomes self-fulfilling. In the second part of the paper, we study whether the cycle induced by self-fulfilling stereotypes can be overcome through better information about the underlying ability distribution. We find that even though beliefs about ability shifts towards gender parity, stereotypes continue to drive specialization and assignment decisions.
This paper summarizes the theoretical and empirical research in economics on the impact of employee diversity on organizational performance, where diversity is predominantly viewed through the lens of gender and ethnicity/nationality. The literature has studied this topic through two types of interactions: horizontal interactions between workers who are peers and vertical interactions between managers and workers. The theory of horizontal interactions highlights the conditions under which diversity is beneficial, such as when different groups bring complementary knowledge, but also when it is costly, as in the case of intergroup communication frictions. The theory of vertical interactions focuses on discrimination against different social groups due to the hierarchical nature of these interactions. Discrimination can result from preferences that favor or disfavor certain groups, or from imperfect information and the use of beliefs about group productivity to infer individual productivity. For both horizontal and vertical interactions, the empirical findings on the impact of diversity are mixed, with varying effects on organizational performance, ranging from positive to negative to no impact. Although this may suggest the field has little to say on the subject, the impact of diversity is often in line with the theoretical predictions. This suggests that for organizations to reap the potential rewards from diversity, they must consider how their context relates to the theory and act accordingly.