What Do Outside CEOs Really Do? Evidence from Plant-Level Data

John Bai is an Associate Professor at D’Amore-McKim School of Business, Northeastern University; and Anya Mkrtchyan is an Associate Professor at Isenberg School of Management, University of Massachusetts–Amherst. This post is based on their article forthcoming in the Journal of Financial Economics.


Whether a new CEO is promoted from within the firm or is hired from outside the company can have a profound impact on strategy-setting and firm performance. When debating the merits of promoting insiders, the typical assumption is that insiders have a deep understanding of a firm’s core competencies and established internal social networks. In contrast, outside CEOs are presumed to possess broader general skills and are believed to have an advantage in bringing bolder changes, as they have no emotional commitments to the firm’s status quo (e.g., Murphy and Zabojnik, 2007; Custódio et al., 2013; Parrino, 1997). However, changes initiated by an external CEO might not necessarily lead to performance improvements, since they may put significant strains on the firm’s managerial resources and organizational systems (Zhang and Rajagopalan, 2010). Hence, whether outside or inside CEOs transform companies better has been one of the most debated questions among academics, practitioners, and boards of directors.

In our recent paper, we contribute to this debate by analyzing the relative performance of outside and inside CEOs, using the U.S. Census Bureau’s rich microdata. This data enables us to analyze plant-level determinants of the choice to hire an outside CEO, estimate more accurately an alternative performance metric that can gauge performance changes taking place in the plants, and shed light on the mechanisms that could plausibly yield changes in performance.


We start our analysis by exploring factors determining CEO turnover and the decision to recruit an outsider. Our data shows a strong inverse relationship between plant-level performance and outside CEO hires, providing added support to the idea that poor performance increases the likelihood of appointing an outsider. Additionally, we show that firms operating in high-tech industries are more likely to opt for outsiders.

CEO origin and plant performance

We then turn to examine the relative performance of inside and outside CEOs. Here, the Census’s rich operational and financial information allows us to assess performance changes happening in the plants by estimating more precisely an alternative performance measure, i.e., total factor productivity (TFP).  We uncover several interesting findings using Census’s data. We show that although pre-turnover firms with outside hires are less productive, on average, relative to firms promoting insiders, they experience greater within-plant increases in productivity following the appointment of outside CEOs. The relation is economically significant: firms with outside CEOs increase TFP by 4.8% more, relative to firms with inside CEOs, which translates into an approximately 12% increase in profitability. Notably, we find that the positive association between outside successions and plant performance is stronger in firms with good prior performance (those with a firm-level three-year industry-adjusted operating margin at or above the sample median), rather than in poorly performing firms (those with a firm-level three-year industry-adjusted operating margin below the sample median). This finding contradicts the conventional wisdom that troubled companies, wherein major changes are required, need to hire outside CEOs, whereas well-performing firms, in which continuation of the current strategy is desirable, should promote internal candidates (e.g., Warner et al., 1988; Karaevli, 2007). Additionally, we observe that the relation between CEO origin and TFP growth is stronger in companies operating in industries facing rapid technological changes (i.e., high-tech firms) and in more complex firms, which challenges the notion that outside CEOs are handicapped in complex firms due to their lack of firm-specific knowledge (e.g., Dalton and Kesner, 1983).


Our next set of tests focuses on identifying the mechanisms underlying the changes in performance. Although the question of how CEO origin is related to actual post-turnover changes is important to understanding the factors influencing a CEO’s success or failure, it has been studied much less than questions related to the decision to hire an outsider or their relative performance. In this respect, the disaggregated nature of our data offers unique advantages, as it enables us to: i) shed light on within-firm changes in resource allocations that could be masked by the aggregated firm-level accounting metrics; ii) delve deeper into post-turnover transformations of firm-level concepts not typically disclosed via annual reports and other SEC filings (such as the composition of capital investment and the number of products); and iii) delineate the relation between CEO changes and outcomes related to other stakeholders, which have not been studied by prior literature, i.e., employees, as it provides more reliable and complete data on plant-level employment, workforce composition, and employee compensation.

We document that productivity improvements arise from eliminating pre-turnover inefficiencies in both extensive and intensive margins. Firms with outside CEOs enhance within-firm resource allocation by exiting low-performing, non-core, low-tech, and highly unionized plants at a higher rate compared to firms with inside CEOs. The performance of continuing plants is also improved. Our results show a substantial decline in capital expenditures, employment, average wage, and materials, as well as consolidation of the existing product offerings. Interestingly, we find that the decrease in inputs is accompanied by an increase in sales, which goes against claims that outsiders’ focus on cost-cutting prevents them from taking the steps necessary to facilitate growth (e.g., Karaevli and Zajac, 2012; Fanning, 2017). In addition, we show that firms with outside successions undergo significant changes in the composition of investment by increasing the fraction of capital expenditures spent on new machinery and information technology. These firms also shift to more capital-intensive production, increase the number of structured management practices, and improve labor productivity. When we compare the economic magnitude of the efficiency improvements via the continuing plants (intensive margin) to those via capital reallocation, such as divestitures of underperforming plants (extensive margin), we find that improvements at continuing plants account for approximately 82% of the firm-level differential in TFP, providing new evidence on the relative importance of the intensive versus the extensive margin in CEO turnovers.


Overall, our study offers a deeper understanding of changes following CEO turnover and uncovers a number of new facts about firms with outside CEOs. The thorough empirical evidence presented in our paper suggests that there are substantial differences between firms with internal and external successors. Our results do not imply that every firm should hire an outsider, but they suggest that inside successors might benefit from adopting an outside perspective, demonstrating a sensitivity to change, and challenging legacies and relationships that might diminish their effectiveness.

Both comments and trackbacks are currently closed.