Abstract We find that division managers who are connected to the CEO are substantially less likely than others to depart from the firm and are more likely to be promoted. Connected managers are protected when performance is poor, and they display no special ability to improve performance given this protection. Connections matter more in weak governance/incentive environments, and the external labor market and stock market appear skeptical of connected managers’ talents. While much of the evidence suggests inefficient favoritism, connected managers are protected more in peripheral segments, suggesting a possible efficiency benefit in helping to resolve intrafirm information problems.
The "Lieutenant" Problem
While the careers and incentives of CEOs have been studied extensively, we know surprisingly little about the "lieutenants" who run the actual divisions of large conglomerates. These division managers are critical to a firm's success, yet their career paths are often opaque. This study tackles a fundamental tension in organizational economics: are personal connections between a boss (the CEO) and a subordinate (the division manager) efficient or inefficient?
- The Efficiency Hypothesis: Connections foster trust and better information flow, allowing CEOs to be "patiently" supportive of talented managers during temporary downturns.
- The Inefficiency Hypothesis: Connections lead to favoritism and "cronyism," where CEOs protect incompetent friends at the expense of shareholders.
Data and Methodology
The study utilizes a robust, hand-collected dataset covering S&P 1500 firms from 2000 to 2015. The researchers hand-matched division managers to specific business segments using 10-K filings and proxy statements, resulting in nearly 6,000 manager-year observations. Personal connections were defined using three criteria: shared educational background (alumni), shared past employment (outside the current firm), or shared social/professional organizations (clubs, charities, etc.).
The researchers employed logit regression models to estimate the probability of departure or promotion, event studies to analyze stock market reactions, and post-separation tracking to assess the external labor market's view of these managers.
The "Long Leash" and Promotion Premium
The findings reveal a stark difference in career outcomes based on social capital. For unconnected managers, there is a strong link between poor performance (measured by ROA) and getting fired. For connected managers, that link effectively disappears; they are "protected" from the consequences of bad results.
- Turnover: Connected managers are 40% less likely to depart the firm involuntarily compared to unconnected managers, holding performance constant.
- Promotion: Sharing a connection with the CEO increases the likelihood of internal promotion by approximately 75%.
Market Skepticism and the Stigma of Firing
The external world appears to view these connections with deep skepticism. Announcements of connected promotions resulted in a -1.33% average 3-day abnormal stock return, compared to a neutral reaction for unconnected promotions. This suggests investors view these moves as agency-driven favoritism rather than merit-based rewards.
Furthermore, if a connected manager actually does get fired, they find it much harder to get a high-level job elsewhere. The external market assumes that if even your "friend" the CEO gave up on you, your talent level must be exceptionally low. Managers who resurface at other public firms typically experience a 14% decrease in base salary.
The "Peripheral" Exception
Interestingly, connections provide the most protection in business segments that are unrelated to the firm's core focus. This suggests that when a CEO doesn't understand a division's business well (high information asymmetry), they rely more heavily on the "trust" of a personal connection to manage it. This is the only area where an efficiency benefit—helping to resolve intrafirm information problems—seems plausible.
Conclusion
While personal connections might help bridge information gaps in complex business segments, the overall evidence points toward inefficient favoritism. Connections matter more in weak governance environments, and both the stock market and external labor markets remain wary of the "friendship premium" in corporate hierarchies.
Summary updated