CEO Behavior and Firm Performance

Abstract

We measure the behavior of 1,114 CEOs in Brazil, France, Germany, India, UK and US using a new methodology that combines (i) data on every activity the CEOs undertake during one workweek and (ii) a machine learning algorithm that projects these data onto scalar CEO behavior indices. Low values of the index are associated with plant visits, and one-on-one meetings with production or suppliers, while high values correlate with meetings with high-level C-suite executives, and several functions together, both from inside and outside the firm. We use these data to study the correlation between CEO behavior and firm performance within the framework of a firm-CEO assignment model. We show results consistent with significant firm-CEO assignment frictions, which appear to be more severe in lower-income regions. The productivity loss generated by ine cient assignment is equal to 13% of the productivity gap between high-and low-income countries in our sample.⇤ This project was funded by Columbia Business School, Harvard Business School and the Kau↵man Foundation.We are grateful to

Management and Shocks to Worker Productivity

“…Managerial quality is highly associated with firm productivity (Adhvaryu, 2018;Adhvaryu et al, 2018;Bloom and Reenen, 2011;Bloom and Van Reenen, 2007;McKenzie and Woodruff, 2016), including in experimental studies (Bloom et al, 2013(Bloom et al, , 2018bKarlan et al, 2015;McKenzie and Woodruff, 2013). A few recent studies have sought to understand which particular practices matter most, though much work remains to be done in this regard (Bandiera et al, 2017;Bloom et al, 2018aBloom et al, , 2016Campos et al, 2017). We add to this recent strand of work by showing one mechanism through which two important dimensions of managerial quality -monitoring (Bloom et al, 2012;Halac and Prat, 2016) and active personnel management (Aghion et al, 2017;Bresnahan et al, 2002) -generate productivity impacts by mitigating the negative effects of shocks.…”

Data science for entrepreneurship research: studying demand dynamics for entrepreneurial skills in the Netherlands

“…This time-consuming procedure, however, creates the problem of small sample sizes and suffers from selection issues. Bandiera et al (2017) tackle the issue by developing new methodology: First, via daily phone calls with 1114 CEOs or their assistants, they collected 42,233 data points about the decision makers' diaries. Then they employed an unsupervised learning algorithm (a latent Dirichlet allocation, LDA), which provides them with a complete probabilistic description of time-use patterns, despite the high dimensionality of their data set.…”

How Does Scientific Progress Affect Cultural Changes? A Digital Text Analysis

“…We also contribute to the growing use of "text as data" in the social sciences. In economics, for example, the use of text corpora is developing especially in such fields as political economy, the study of media, and innovation (Balsmeier et al 2018;Bandiera et al 2017;Catalini et al 2015;Gentzkow et al 2018, Iaria et al 2018Jelveh et al 2014;Kelly et al 2017). In business studies, there are several applications to marketing and finance (see Kearney and Liu 2014 for a review).…”