The Pay-for-Performance Promise
Every proxy statement tells the same story: CEO compensation is designed to pay for performance. Bonuses reward achieving goals. Stock grants align CEO and shareholder interests. Long-term incentives focus executives on sustainable value creation.
The logic is compelling. The evidence is troubling.
What boards claim:
- CEO pay is tightly linked to company performance
- High pay reflects high performance
- Performance-based compensation motivates better results
- Shareholders benefit from incentive alignment
What research shows:
- The correlation between CEO pay and company performance is weak and inconsistent
- CEO pay increases regardless of performance in most years
- Incentive effects are smaller than commonly believed
- Much of what looks like pay-for-performance is actually pay-for-luck
After 40 years of academic research on executive compensation, the pay-for-performance promise remains largely unfulfilled.
The Research Evidence
Finding 1: Weak Pay-Performance Correlation
The fundamental relationship between CEO pay and company performance is surprisingly weak.
The research findings:
Classic studies: Jensen and Murphy's landmark 1990 research found CEO wealth changed by only $3.25 for every $1,000 change in shareholder value—a sensitivity they called "low" relative to incentive theory predictions.
Recent updates: Despite dramatic increases in equity compensation since 1990, pay-performance sensitivity has increased only modestly. CEOs capture a tiny fraction of the value they create (or destroy).
Cross-company comparison: Companies with higher-paid CEOs don't systematically outperform companies with lower-paid CEOs. Within industries, pay levels don't predict performance differences.
Year-over-year analysis: CEO pay increases most years regardless of whether performance improved. The "ratchet effect" pushes pay upward independent of results.
What this means:
If pay were truly tied to performance, we'd see strong correlations—high performance companies paying more, low performance companies paying less, pay rising with performance and falling with underperformance. We don't see this pattern consistently.
Finding 2: Pay for Luck, Not Skill
Much of what appears to be pay-for-performance is actually pay-for-luck.
The luck problem:
Industry effects: When oil prices rise, oil company CEOs get paid more—even though they didn't cause oil prices to rise. Industry tailwinds (and headwinds) affect CEO pay, unrelated to CEO actions.
Market effects: During bull markets, CEO equity compensation increases simply because stock prices increased broadly. The CEO may have done nothing to cause the increase.
Asymmetric filtering: Boards don't symmetrically adjust for luck. Positive luck flows through to higher pay; negative luck often gets filtered out through discretionary adjustments.
The research evidence:
Bertrand and Mullainathan's influential research showed CEOs are rewarded for "lucky" performance they don't control—commodity price movements, exchange rate changes, overall market returns. Boards rarely adjust for these factors.
What this means:
A significant portion of "performance-based" pay isn't based on performance at all—it's based on external factors the CEO didn't influence. The pay-for-performance label often masks pay-for-luck.
Finding 3: Incentive Effects Are Smaller Than Assumed
The behavioral response to performance-based pay is weaker than incentive theory predicts.
The incentive assumption:
Economic theory assumes performance pay motivates effort and attention. Higher performance-pay sensitivity should produce more performance.
The research reality:
Effort effects: Research struggles to document significant effort effects from executive compensation. CEOs already work hard; marginal incentives may not increase effort meaningfully.
Attention effects: CEOs may focus attention on measured metrics—but this can distort behavior toward measurable activities away from unmeasurable but important ones.
Risk effects: Heavy equity compensation can encourage excessive risk-taking, value-destroying acquisitions, and short-term stock price manipulation.
Retention effects: Compensation may matter more for retention than incentive. CEOs who feel underpaid leave; well-compensated CEOs stay—but this is different from performance motivation.
What this means:
The incentive story—pay more for performance and get more performance—is weaker than commonly assumed. Pay affects retention and recruitment more clearly than it affects effort and results.
Finding 4: The Comparison Group Problem
Much of CEO pay reflects comparison group dynamics rather than performance.
The comparison problem:
Benchmarking behavior: Boards set CEO pay by comparison to peer companies. Every board targets median or above. The result: continuous upward pressure regardless of performance.
Peer group selection: Companies can select peer groups that justify desired pay levels. "Aspirational" peers, larger companies, and higher-paying industries all enable higher pay.
Consultant incentives: Compensation consultants, paid by companies, have incentives to recommend higher pay. "Below market" findings create client dissatisfaction.
The research evidence:
Faulkender and Yang's research documented systematic peer group manipulation—companies select peers that justify higher pay. Bizjak, Lemmon, and Naveen showed peer group benchmarking drives much of CEO pay, independent of performance.
What this means:
The pay-for-performance narrative provides cover for pay determined by other factors. Comparison-based pay isn't performance-based pay—it's market-matching pay, regardless of performance.
Finding 5: Governance Weakness Predicts Pay Levels
CEO pay is higher when governance is weaker—the opposite of what pay-for-performance would predict.
The governance finding:
Board independence: Companies with less independent boards pay CEOs more, controlling for performance and company characteristics.
CEO power: When CEOs have more power (combined CEO-Chair role, longer tenure, weaker board), pay is higher.
Shareholder monitoring: Companies with concentrated ownership or activist shareholders pay CEOs less.
The research evidence:
Bebchuk and Fried's influential work documented how CEO power affects compensation. Core, Holthausen, and Larcker showed governance weakness predicts higher pay. The pattern: weaker oversight, higher CEO pay.
What this means:
If pay were truly performance-based, governance quality wouldn't matter—good performers would be paid well regardless of governance. The governance-pay relationship suggests pay reflects bargaining power, not just performance.
Why the Myth Persists
Reason 1: Disclosure Complexity
Compensation disclosure is technically compliant but practically incomprehensible.
The complexity problem:
- Proxy statements run hundreds of pages
- Compensation tables require sophisticated interpretation
- Performance conditions are buried in technical language
- Year-over-year comparisons are difficult
The effect:
Shareholders, media, and even directors struggle to understand actual pay-performance relationships. Complexity enables disconnect between narrative and reality.
Reason 2: Board Capture
Boards that set CEO pay are often influenced by the CEO.
The capture mechanisms:
- CEOs influence director selection and retention
- Directors depend on CEO for information and access
- Social relationships between CEO and directors
- Directors' own compensation affected by CEO decisions
The effect:
Boards may genuinely believe they're implementing pay-for-performance while actually implementing pay-for-CEO-preference. Capture is often unconscious.
Reason 3: Consultant Conflicts
Compensation consultants face inherent conflicts.
The conflict structure:
- Consultants hired by the company (influenced by CEO)
- Consultant revenue depends on client satisfaction
- Recommendations of higher pay are easier to sell
- Low-pay recommendations risk client loss
The effect:
Even well-intentioned consultants may unconsciously skew toward recommendations that please management. The market rewards consultants who find reasons to pay more.
Reason 4: Narrative Power
The pay-for-performance story is compelling regardless of evidence.
The narrative appeal:
- It sounds fair: perform well, get paid well
- It sounds logical: incentives motivate performance
- It sounds American: meritocracy at work
- It provides cover for high pay levels
The effect:
Boards, shareholders, and the public want to believe pay reflects performance. The narrative persists because people prefer it to the alternative explanation—that high CEO pay reflects power, not performance.
What Actually Drives CEO Pay
Driver 1: Market Dynamics
Supply and demand for CEO talent affects pay levels.
Market effects:
- Perceived scarcity of qualified CEOs enables price increases
- Competition for CEO talent bids up compensation
- Portable skills allow CEOs to seek alternative opportunities
- Private equity and other alternatives create outside options
The market reality:
Market dynamics explain some pay variance—but markets can be inefficient, especially when information asymmetries and relationship factors affect hiring.
Driver 2: Company Size
Larger companies pay more, controlling for other factors.
The size effect:
- Larger companies pay higher CEO compensation
- The relationship is robust across time and methodology
- Size may proxy for complexity, scope, or visibility
The size puzzle:
If pay were purely performance-based, company size shouldn't matter beyond its effect on performance. The strong size-pay relationship suggests pay reflects job size more than performance.
Driver 3: CEO Bargaining Power
CEOs with more power extract more compensation.
Power sources:
- Founder status or large ownership stake
- Long tenure creating information advantage
- Celebrity status creating outside options
- Board relationships reducing governance intensity
The power effect:
CEOs with more bargaining power negotiate better compensation packages. This isn't pay-for-performance; it's pay-for-power.
Driver 4: Rent Extraction
Some CEO pay represents value transfer from shareholders to executives.
Rent extraction mechanisms:
- Excessive severance unrelated to performance
- Hidden compensation through perks and benefits
- Favorable equity terms and timing
- Guaranteed pay regardless of results
The rent reality:
Not all above-market CEO pay reflects above-market performance. Some represents successful extraction of value by powerful CEOs from inadequately vigilant boards.
What Boards Should Do
Recommendation 1: Honest Assessment
Boards should honestly assess their pay-performance relationships.
Assessment questions:
- Over the past decade, has CEO pay tracked company performance?
- Would the CEO have earned less during poor performance periods?
- Can we justify our CEO's pay relative to industry performance?
- Are we paying for luck or skill?
The assessment discipline:
Most boards would find weaker pay-performance linkage than they assume. Honest assessment is the first step toward improvement.
Recommendation 2: Luck Filtering
Boards should filter out luck from performance-based pay.
Filtering mechanisms:
- Relative performance metrics that compare to peers experiencing similar conditions
- Industry-adjusted returns that remove sector effects
- Discretion to adjust for windfall gains or losses
- Explicit consideration of what the CEO actually controlled
The filtering discipline:
If the CEO didn't cause the performance—positive or negative—the CEO shouldn't be rewarded or penalized for it. This simple principle is rarely implemented rigorously.
Recommendation 3: Reduce Complexity
Simpler compensation structures create clearer accountability.
Simplification approaches:
- Fewer performance metrics (focus on 2-3 that matter most)
- Simpler equity vehicles (straight stock rather than exotic instruments)
- Clearer vesting conditions
- More transparent disclosure
The simplification benefit:
When everyone—CEO, board, shareholders—can understand how pay relates to performance, accountability improves.
Recommendation 4: Independent Governance
Stronger governance improves pay-performance alignment.
Governance improvements:
- Truly independent compensation committees
- Committee-retained compensation consultants
- Regular executive sessions without CEO
- Robust shareholder engagement
The governance imperative:
Pay-for-performance is impossible without governance structures that enable genuine board independence and accountability.
Recommendation 5: Long-Term Orientation
Extend the time horizon for performance evaluation.
Long-term mechanisms:
- Longer performance periods (5+ years rather than 3)
- Extended vesting and holding requirements
- Retrospective performance evaluation
- Clawback provisions for value destruction
The long-term benefit:
Short performance periods enable manipulation and luck. Longer horizons reveal sustainable value creation versus temporary gains.
The Bottom Line
The pay-for-performance promise underlying CEO compensation is largely myth. Research consistently shows:
- Weak correlation between CEO pay and company performance
- Significant pay-for-luck contaminating performance-based pay
- Incentive effects smaller than commonly assumed
- Governance weakness predicting higher CEO pay
The uncomfortable truth:
Much of what boards call pay-for-performance is actually pay-for-luck, pay-for-power, pay-for-size, or pay-for-market-matching. The performance narrative provides cover for compensation driven by other factors.
What this means for boards:
Honest reckoning: Stop assuming pay-for-performance works. Evaluate whether it actually does in your company.
Structural improvement: Implement mechanisms that filter luck, extend time horizons, and create genuine performance accountability.
Governance strengthening: Ensure board independence and shareholder accountability. Weak governance enables the gap between narrative and reality.
Expectation resetting: Accept that incentive effects are modest. Compensation affects retention and recruitment more reliably than it motivates performance.
The pay-for-performance story is appealing. The evidence suggests it's largely fictional.
Boards that want CEO pay to actually reflect performance must design for that outcome intentionally—because the default is pay that rises regardless of results.
That's what 40 years of research shows.
That's what boards should finally accept.