"Do CEOs matter in the long run? British CEOs in the twentieth century"

Paper

The extent to which corporate leaders matter to the companies they lead is widely debated. An extensive literature has emerged around the measurement of the impact CEOs have on the companies they lead (Lieberson and O’Connor, 1972). This is called the ‘CEO effect’ and measures how much of the variation in a firm’s performance (positive and negative) is explained by leader, firm, year, and industry effects.

Previous research (Quigley and Hambrick, 2015) has shown that in the United States the CEO effect doubled between 1950 and 2009. This change was due to long run trends that empowered the CEO role. These include the emergence of shareholder capitalism (Lazonick and O’Sullivan, 2000), increased competition and globalization. These findings have significant implications for understanding how the role and importance of corporate leaders has changed, the evolution of corporate governance, and claims underpinning the enormous increases in CEO pay.

We address these debates by measuring the CEO effect of British CEOs between 1900 and 2009, using a hand collected database of 407 firms and 1,558 CEOs. To measure corporate performance, we collect annual data for the firm’s return on assets and total shareholder return. We find that there is a CEO effect in the UK which increased by fifty percent across the century. The effect is lower than in the US and grows more slowly over time. We consider how changes in government economic policies, corporate governance, company ownership and organisation, and characteristics of CEOs, explain these findings. This is a novel addition to the rich history examining the impact of management and leadership on the performance of British companies (Chandler, 1990; Elbaum, and Lazonick, 1984; Keeble, 1992) and to the growing Strategy literature (Crossland & Hambrick, 2007, 2011) seeking to understand differences in the CEO effect across nationalities and time.