Measuring CEO responsible leadership: Development and validation of a linguistic-based instrument

Responsible leadership (RL) has gained growing attention in both academia and organizational practice. However, extant research faces three key limitations: a lack of conceptual clarity (i.e., tautologies and overlaps with other leadership constructs), a conflation of leader behaviors with follower evaluations, and overreliance on surveys, which typically capture supervisory rather than top-executive perspectives. To address these issues, we introduce a novel behavioral measure of responsible CEOs based on linguistic markers. We leverage computer-aided text analysis (CATA) to develop and validate a keyword-based CEO RL measure, combining deductively derived words with inductively derived words from machine learning (ML) and human raters. We then empirically examine the relationship between CEO RL and firm Corporate Social Responsibility (CSR). Using a sample of 955 CEOs over 19 years, we find that CEO RL is positively associated with CSR. Finally, we derive a taxonomy of RL behaviors (RLBs) and show overlaps and distinctions with Ethical Leader Signals (ELS). We conclude with a roadmap for future research.