Introduction

Organisations frequently face decisions that require trading off multiple dimensions of performance, such as financial returns versus environmental and social (ES) impact. While conventional wisdom suggests appointing leaders who share the organisation’s overall objectives, this research demonstrates that appointing “non-value-maximising” managers can be optimal. Central to this result is the observation that measured performance may diverge from actual output to varying degrees. For instance, shareholder value can be proxied by stock prices, but many ES impacts are qualitative and third-party ratings are often inconsistent. Hiring managers with an intrinsic bias toward poorly measured dimensions allows a board to use high-powered incentives on better-measured tasks—like profits—without causing the manager to neglect the more difficult-to-measure goals.

Model

The study develops a parsimonious principal-agent model of resource allocation. A risk-neutral principal hires a manager who receives private information about the productivity of potential resource allocations. The manager allocates resources across two activities: “profits” (the better-measured activity) and “impact” (the worse-measured activity). While true output is non-contractible, imperfect and contractible performance measures are available for both. The principal’s objective is the total output from both activities, minus the manager’s pay and the cost of deviating from a default allocation. The manager maximises their pay minus the cost of deviating from their own preferred resource allocation, which reflects their personal preferences and bias.

Analysis

In a “first-best” scenario where productivity and resource allocation are observable, congruence—hiring a manager whose preferences match the principal’s—is optimal. However, in the second-best scenario where resource allocation is non-contractible, distortions arise. Because profits are more precisely measured, optimal incentive contracts naturally place greater weight on them. This asymmetry induces the manager to prioritise profits at the expense of impact. To mitigate this distortion, the principal optimally appoints a manager who values impact more than the firm does. This bias creates a counterweight, allowing the firm to offer stronger incentives on profits to induce responsiveness to private information without causing excessive neglect of impact. The optimal bias is higher when the manager’s private information is more valuable and lower when the firm’s costs of deviating from its preferred allocation are high.

Matching in competitive markets

The research extends the model to a continuum of firms and managers with heterogeneous measurement qualities and preferences. The equilibrium involves negative assortative matching: firms with the poorest measurement of impact hire managers with the strongest intrinsic preferences for it. This occurs because strong preferences and weak measurement function as complements. The marginal rent of a manager is higher if their matched firm has poorer measurement for impact, as their bias becomes more valuable. Managers with intermediate biases often earn the highest rents, balancing the benefit of compensating for poor measurement against the deadweight loss of resource-allocation distortions. Numerical examples demonstrate this; for instance, in Example 1, where the firm’s deviation costs are moderate (c = k = 1), the highest rent is earned by a manager with a bias (αM = 0.63) above the principal’s benchmark (αP = 0.5).

Unobservable managerial types

When managerial preferences are unobservable, they may be signalled through costly actions, such as charitable donations, volunteering, or prior managerial decisions. If the sorting effect dominates (high variation in firm measurement), signals exaggerate differences from a “pivot” type, leading to more extreme stated preferences. Conversely, if the direct effect dominates (high firm adjustment costs), signals are compressed toward the pivot as managers present themselves as more balanced. This explains both “virtue signalling” for social impact and signalling “commercial discipline” through actions like closing plants or reducing employment to improve long-run profitability despite personal unpopularity.

Implications

The model provides several governance implications. It suggests that managerial preferences should be matched to measurement technology: within a firm, managers should be biased toward the worse-measured dimension. Furthermore, more impact-motivated executives should receive weaker explicit impact incentives. Strong profit incentives in mission-focused firms do not necessarily indicate “greenwashing”; the firm may simply be pursuing its mission through a biased manager rather than through flawed impact metrics. Improved measurement of impact, such as through superior ESG reporting standards, may optimally shift corporate culture away from impact, as such output can instead be induced via financial incentives.

Applications

The framework applies to executive compensation, venture capital, universities, and healthcare. In healthcare, hospitals should hire doctors with intrinsic preferences for “prevention” (harder to measure) while using financial incentives tied to “cures” (easier to measure). For hospital CEOs, hiring a physician can proxy for stronger intrinsic preferences for patient care, allowing compensation to focus on financial dimensions. In universities, institutions often rely on hiring faculty with intrinsic motivation for teaching when student evaluations are noisy or easily gamed, while using explicit incentives for better-measured research output.

Conclusion

The paper concludes that managerial selection and incentive design are complementary tools. Appointing a biased manager enables the use of stronger incentives on well-measured tasks, amplifying asymmetries in pay-for-performance across dimensions. These findings suggest that incentive contracts cannot be interpreted independently of hiring policies, as strong incentives on one dimension may be optimal precisely because the organisation has selected a manager intrinsically committed to another.