Do Good People Make Bad Innovators? Managerial Prosocial Preferences and Automation Innovation

Abstract

Automation, even while increasing aggregate employment in the long run, can displace and cause significant harm to incumbent employees. We propose that managers’ prosocial preferences, specifically their desire not to displace and harm their employees, deter investing in automation innovation. Using both novel and previously-used proxies of prosocial preferences (e.g., CEO’s use of ‘we’ vs. ‘they’ pronouns during earnings calls, employee-prosocial vs. shareholder-centric language in annual reports, Machiavellianism score, and charity engagement), we show that managerial prosociality decreases automation innovation in US public firms while having a weaker effect on non-automation innovation. The negative effect is stronger when financial slack provides managers with greater discretion and weaker when social safety nets reduce the harm from losing a job. In a complementary laboratory experiment, we show that aversion to harming employees underpins the negative relation. Our study highlights prosocial preferences as a novel source of heterogeneity that shapes the direction of firm technological investment and provides a richer psychological foundation beyond self-interest and career concerns.

Keywords: managerial prosocial preferences, innovation, automation, CEO characteristics

JEL Classification: D91, O31, O33