Discrimination in Hiring: Evidence from Retail Sales
Alan Benson (University of Minnesota)
Simon Board (UCLA)
Moritz Meyer-ter-vehn (UCLA)

Abstract : Using data from a major U.S. retailer, we find that black, white and Hispanic managers within the same store are 2-3% more likely to hire workers of their own race. This segregation may be caused by taste-based discrimination, whereby managers intrinsically prefer same-race applicants, or by screening discrimination, whereby managers have better information about same-race applicants. To separate between these hypotheses we use the productivity distributions of commission-based salespeople. We find workers are generally more productive when hired by a same-race manager, and that white and Hispanic workers also have lower productivity variance when hired by a same-race manager, suggesting that screening discrimination is more important than taste-based discrimination.

Do Ceos Know Best? Chinese Ceo, Manager and Worker Survey Data
Nicholas Bloom (Stanford University)
Hong Cheng (Wuhan University)
Mark Duggan (Stanford University)
Hongbin Li (Stanford University)
Franklin Qian (Stanford University)

Abstract : We analyze a new management survey for around 1,000 firms across two large provinces in China. The unique aspect of this survey is it collected management data from the CEO, a random sample of senior managers and workers. We document five main results. First, management scores, much like productivity, have a wide spread with a long left-tail of poorly managed firms. This distribution of management scores is similar for CEOs, senior managers and workers management, and appears broadly reasonably compared to US scores for similar questions. Moreover, for all groups these scores correlate with firm performance, suggesting all employees within the firm are (at least partly) aware of the their firms’ managerial abilities. Second, the scores across the groups are significantly cross-correlated, but far from completely. This suggests that while different levels of the firm have similar views on the firms management capabilities, they do not fully agree. Third, we find that the CEO’s management scores are the most predictive of firm performance, followed by the senior managers and then the workers. Hence, CEOs do appear to know best about their firms management strengths and weaknesses. Fourth, within-firm management score dispersion is negatively correlated with investment and R&D intensity, but uncorrelated with other performance measures. Finally, when the management score dispersion is large, the impact of management score on firm performance becomes smaller.

Analyzing the Aftermath of a Compensation Reduction
Jason Sandvik (University of Utah)
Richard Saouma (Michigan State University)
Nathan Seegert (University of Utah)
Christopher Stanton (Harvard Business School)

Abstract : Wage rigidity creates real and financial frictions, though the real-world drivers of rigidities remain largely unstudied. We use staggered commission reductions at a sales firm to estimate effects on worker turnover and effort. In response to an 18\% decline in commissions, we find turnover increases for the most productive workers. We detect limited effort responses, and find no evidence of different effects based on workers' expectations of fairness or future promotion. The findings suggest that adjustment constraints stem primarily from adverse selection concerns on the extensive (turnover) margin as opposed to asymmetric effort responses on the intensive margin.