THE CEO EFFECT: What the Research Actually Says About Leadership and Firm Performance - CEOWORLD magazine
Ask any shareholder why a struggling company suddenly turned around, or why a thriving one collapsed, and the answer is almost always the same: the CEO. We credit leaders with near-mythic powers to shape corporate destiny. Executive search firms, compensation committees, and the business press reinforce this narrative daily. But what does the empirical evidence […]
Ask any shareholder why a struggling company suddenly turned around, or why a thriving one collapsed, and the answer is almost always the same: the CEO. We credit leaders with near-mythic powers to shape corporate destiny. Executive search firms, compensation committees, and the business press reinforce this narrative daily. But what does the empirical evidence actually say? Considerably less than the mythology suggests, and the gap between perception and reality has serious implications for how we lead, govern, and invest. The Numbers: How Much Does the CEO Actually Drive?Four decades of management research have tried to isolate the ‘CEO effect’: the measurable portion of firm performance variance attributable to the individual at the top. The findings are illuminating, and more than a little humbling. Early studies were skeptical. Lieberson and O’Connor (1972) found the CEO explained roughly 14% of variation in company sales and earnings, a real but far from dominant effect. Subsequent researchers refined the methodology. Wasserman, Nohria, and Anand (2010), working from Harvard, landed in similar territory: the average CEO accounts for 10 to 20% of the variability in firm performance. Hambrick and Quigley (2014), using a more contextualised modelling technique on a 20-year data set, pushed the upper estimate to approximately 33% in high-discretion industries, but closer to 17% for large public companies with powerful boards and entrenched organisational structures. More recent work using machine-learning methods has generally confirmed the range of 15–25%, with the balance of performance variance attributable to industry conditions, firm-level culture, and year-on-year macroeconomic forces. At best, the CEO directly drives roughly one-quarter of what happens to performance. The other three-quarters lives inside the organisation itself. These numbers should not be read as evidence that leaders don’t matter. They clearly do. But the research underscores a crucial point: the CEO is one input among many, not the master control. The organisation (its culture, systems, structures, and embedded practices) is doing most of the heavy lifting. Context Is Everything: When CEOs Matter More (and Less)Not all CEOs are equally impactful. The research consistently shows that CEO influence is contingent: it expands or contracts depending on structural factors largely outside the leader’s control. Managerial DiscretionThe single biggest moderator is discretion: how much latitude does the CEO actually have to make decisions? Hambrick and Finkelstein’s theory of managerial discretion (1987) identified three drivers: industry dynamics, internal organisational factors, and individual characteristics. A CEO running a capital-intensive mining or utility company faces enormous structural constraints: regulation, commodity prices, long depreciation cycles. An advertising agency CEO, by contrast, can pivot the entire firm with a single creative hire. Research confirms that CEO effects are significantly larger in high-discretion environments (professional services, technology, early-stage companies) and far smaller in regulated, capital-heavy, or highly bureaucratic ones. Board Power and GovernanceWhen boards are powerful and engaged, they constrain individual CEO influence. This is not necessarily bad. Strong governance tends to distribute strategic judgment more broadly, reducing the organisation’s vulnerability to a single personality. Conversely, founder-CEOs with controlling stakes (think Elon Musk at Tesla or Mark Zuckerberg at Meta) show substantially higher CEO effects, for better or worse. Tenure and TimingResearch on CEO turnover (ScienceDirect, 2024) consistently shows that new leaders have their strongest measurable impact in the first two years of tenure. Impact then diminishes, as the CEO’s strategic choices become embedded in structure, culture, and habit, and the organisation, not the person, carries them forward. This temporal dynamic is systematically underappreciated in boardrooms. The Overlooked Variable: What’s Actually Driving the Other 75%?If the CEO explains only 10–33% of performance variance, what accounts for the rest? Three categories dominate the research:Industry and Macroeconomic ForcesStudies routinely show that industry-level factors (sector growth rates, competitive intensity, regulatory environment, interest rate cycles) explain a substantial and arguably underestimated share of firm performance. A mediocre CEO in a booming sector often outperforms an excellent CEO in a structural decline. Industry tailwinds and headwinds are powerful. We consistently misattribute their effects to the leader. Firm-Level Culture, Systems, and PracticesResearch by Hartnell, Kinicki, and colleagues (Journal of Applied Psychology, 2016) found that organisational culture and human resource practices (reward systems, performance appraisal fairness, work design) explain significant variance in performance independently of leadership behaviour. Critically, in strong-culture organisations, culture can substitute for leadership: employees behave consistently with shared values whether or not the CEO is actively directing them. This is the ‘flywheel effect’: culture, once embedded, runs on its own momentum. A landmark study across 3,245 respondents in 424 organisations spanning 12 industries found that while industry-level factors showed no significant relationship to cultural norms, leadership behaviours and HR practices together explained meaningful variance. In other words: internal organisational design matters far more than which sector you’re in. Middle Management and Distributed ExecutionHarvard Business School’s Raffaella Sadun and colleagues have demonstrated that how CEOs allocate their time varies widely across firms, with direct but lagged consequences for productivity, typically emerging in years three to five after appointment. But Sadun’s work also highlights that the mechanisms of CEO influence run through middle management and organisational processes. The CEO sets direction; the organisation translates it into action. That translation layer, or its failure, is where most performance is won and lost. Boeing’s catastrophic safety failures in the 737 MAX programme were not simply a CEO failure. They were the product of years of embedded cultural drift: cost pressures institutionalised into engineering decisions, speak-up culture suppressed by hierarchical norms, and risk management systems progressively hollowed out. The Real-World Evidence: Leadership Myths and RealitiesApple after Steve Jobsis the standard counter-argument: surely Jobs’ removal in 1985 and return in 1997 prove individual CEO primacy? But look more carefully. The product systems, design philosophy, and engineering culture Jobs built persisted until he returned and demanded they take precedence over quarterly performance. And the Apple culture compounded even after his death in 2011. Tim Cook’s Apple has generated more revenue than Jobs’ Apple ever did, by running the machine Jobs constructed. GE under Jack Welchis an even more instructive case. Welch was lauded as the defining CEO of his era, and GE’s stock soared during his tenure. But the management systems he embedded (financial engineering, short-termism performance culture, aggressive earnings management) ultimately contributed to GE’s near-collapse as the business environment changed. The organisational culture Welch built outlasted him, but did not match the new world of business. Microsoft’s transformation under Satya Nadella(from 2014) is often cited as a pure CEO turnaround story. But Nadella’s genius was largely to surface and amplify existing technical talent and capability (the engineers who built Azure were already inside Microsoft) while dismantling cultural pathologies (stack ranking, internal competition) that had suppressed performance. He changed the culture; the culture changed the results. Implications: Rethinking the CEO PremiumThe evidence suggests several conclusions that should challenge prevailing boardroom orthodoxy:First, CEO compensation packages that treat the individual as responsible for 80–90% of firm value creation are not supported by research. The pay premium accorded to ‘star’ CEOs reflects mythology more than measurable impact. Second, organisations that are excessively CEO-dependent are structurally fragile. Boards that obsess over succession planning while neglecting the systems, culture, and practices that carry performance between tenures are solving the wrong problem. Third, the question boards should be asking is not merely ‘Do we have the right CEO?’ but ‘Do we have the right organisation?‘ That means investing in culture diagnostics, work practice design, middle management capability, and the embedded systems that produce results independent of who sits in the corner office. The CEO effect is real, meaningful, and worth noting. But it operates within, and through, the organisation. Leaders who understand this invest as much in the machine as they do in the drivers. Boards that understand this evaluate the system as rigorously as they evaluate the individual. And analysts who understand this look beyond the leadership announcement to ask what kind of organisation the new CEO is actually inheriting and is there evidence he/she has the courage to reshape the culture. The best CEOs are not the ones who dominate their organisations. They are the ones who build organisations that perform beyond them. Key ReferencesLieberson, S. & O’Connor, J. F. (1972). Leadership and organizational performance. American Sociological Review, 37(2), 117–130. Hambrick, D. C. & Quigley, T. J. (2014). Toward more accurate contextualization of the CEO effect on firm performance. Strategic Management Journal, 35(4), 473–491. Wasserman, N. , Nohria, N. & Anand, B. N. (2010). When does leadership matter? In Handbook of Leadership Theory and Practice. Harvard Business School Press. Hartnell, C. A. , Kinicki, A. J. , Lambert, L. S. & Fugate, M. (2016). Do similarities or differences between CEO leadership and organizational culture have a more positive effect on firm performance? Journal of Applied Psychology, 101(6), 846–861. Dahlstrand, A. , László, D. , Schweiger, H. , Bandiera, O. , Prat, A. & Sadun, R. (2025). CEO-Firm Matches and Productivity in 42 Countries. NBER Working Paper No. 33324 / Harvard Business School Working Paper No. 25-033. Quigley, T. J. & Hambrick, D. C. (2015). Has the ‘CEO effect’ increased in recent decades? Strategic Management Journal, 36(6), 821–830. Salvi, A. , Tron, A. & Colantoni, F. (2024). The impact of CEO turnover on firm performance and insolvency risk: A global analysis. Finance Research Letters, pii S1544612324001235. Written byJ. R. Childress. Have you read?Legal AI Has a Trust Problem—The Firms That Solve It Will Win the Decade. AI Is a Performance Multiplier: The Two Drivers of Elite Success. Why Recovery Alone Won’t Deliver High Performance. The Height Standard: Redefining Excellence in Leadership and Performance. How Briana Hetherington Is Transforming Executive Leadership Through Mindset.