What Do Good Managers Do?

On December 5, 2017, Tuesday, at 4:15 pm in Pettengill G65, Samarth Gupta, an Economics PhD Candidate from Boston University presented “What Do Good Managers Do? Evidence from an Insurance Firm in India.”

Abstract:   Do managers within a firm differ in productivity? If so, what do good managers do? In this paper, I use a panel dataset of 211 manager-led teams of salespersons from 2012-2015 in an insurance firm in India. I find large performance differences: the ratio of 90th and 10th percentile of team output is close to 20 and this ratio for output per worker is 5:1. Further, these two metrics are positively correlated. To find what good managers do, I use output per worker of each team in 2012 as a metric of managerial ability and correlate it with the outcomes of the managerial tasks in from 2013-2015. Managerial ability is positively and significantly correlated with the output of their new recruit. When an agent, whose manager exited from the firm, is assigned to a new manager quasi-randomly, his output increases only when joining a manager in the upper 10th percentile. Further, agents within or across teams do not differ in the nature, cost and value of products they sell, ruling out product or market heterogeneity. Skill differential, thus, appears to be selection of recruits for most managers and guidance & supervision  for high performing managers. To explore the implications of differences in selection, I develop a model where managers differ in the precision of a signal they receive of a candidate’s productivity before recruiting him. The model provides the following implications: a) negative relation between exit rate and managerial ability, b) positive span of control with respect to managerial ability & tenure and c) increasing returns to scale of team output with team-size. Empirical results confirm all the implications. The paper demonstrates how skill differential across managers may drive productivity variation.

Presented by The Casey Lecture Fund – Economics.