Board Diversity Appearance and Firm Performance: An Image-Based Deep Learning Approach

Greger L, Scholz H, Stiller A, Webersinke N (2024)

Publication Language: English

Publication Type: Other publication type

Publication year: 2024

URI: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4947640

Abstract

We introduce a new measure of board diversity appearance leveraging machine learning methods. Using portrait pictures of board members from S&P 100 firms, we receive a score that quantifies the level of visually observable diversity on a board. We argue that stakeholders may use this board diversity appearance to approximate unobservable attributes of board members that affect firm performance. Further, we suggest that this signal is particularly valuable for firms that do not receive much public attention and could positively alter the behavior of stakeholders towards a firm. Using factor model analyses, we find indication that f irms with more diverse appearing boards and less public attention indeed tend to outperform. In addition, we also find support for improved firm performance of these firms based on accounting measures.

Authors with CRIS profile

Lukas Greger Lehrstuhl für Betriebswirtschaftslehre, insbesondere Finanzierung und Banken Hendrik Scholz Lehrstuhl für Betriebswirtschaftslehre, insbesondere Finanzierung und Banken Anja Stiller Lehrstuhl für Betriebswirtschaftslehre, insbesondere Finanzierung und Banken Nicolas Webersinke Lehrstuhl für Betriebswirtschaftslehre, insbesondere Finanzierung und Banken

How to cite

APA:

Greger, L., Scholz, H., Stiller, A., & Webersinke, N. (2024). Board Diversity Appearance and Firm Performance: An Image-Based Deep Learning Approach.

MLA:

Greger, Lukas, et al. Board Diversity Appearance and Firm Performance: An Image-Based Deep Learning Approach. 2024.

BibTeX: Download