Agency Theory (Jensen-Meckling)

In the AI Story

Jensen and Meckling's paper became one of the most cited works in financial economics, shaping business school curricula, corporate governance practices, and regulatory approaches for decades. Its influence derived partly from theoretical elegance—the reduction of complex organizational realities to a clean principal-agent framework—and partly from alignment with political currents favoring markets over institutions. The paper's assumptions—that shareholders are principals who deserve priority, that managers are self-interested agents who must be controlled, that stock prices reflect productive value—became embedded in governance structures so thoroughly that questioning them seemed naive or ideological.

Lazonick's critique of agency theory operates at the level of empirical consequences rather than theoretical coherence. The theory predicted that aligning executive incentives with shareholder interests through stock-based compensation would improve corporate performance. The evidence from four decades of governance evolution shows the opposite: firms that adopted heavy stock-based compensation exhibited declining R&D investment, increased workforce churning, massive buyback spending, and stagnant innovation performance. The alignment occurred—executives acted in shareholder interests. But shareholder interests, once made dominant, were extraction interests. The theoretical framework assumed shareholders wanted long-term productive value. The institutional reality revealed that shareholders, operating in liquid markets with low switching costs, wanted quarterly returns and stock price appreciation—regardless of long-term productive consequences.

In the AI economy, agency theory's assumptions break down with particular visibility. The theory assumed that shareholder value maximization and productive innovation would converge—that firms maximizing stock prices would naturally invest in the capabilities producing long-term competitive advantage. AI reveals the divergence: the decision that maximizes shareholder value (replace expensive human workers with cheap AI systems, distribute the savings through buybacks) is precisely the decision that destroys organizational capabilities (tacit knowledge, collaborative relationships, institutional memory) on which sustained innovation depends. The theory cannot accommodate this divergence because it treats shareholders as undifferentiated principals and ignores the temporal structure of shareholder interests in liquid markets. Shareholders who can sell tomorrow do not care about productive capabilities that mature over years.

Origin

Jensen and Meckling's framework built on earlier agency-theoretic work in economics and extended it to corporate governance. Their innovation was not the principal-agent model itself but its application to the corporation and the specific claim that stock-based compensation was the solution to agency costs. The paper's timing—published during the period when hostile takeovers and leveraged buyouts were beginning to discipline managerial behavior—gave it practical influence beyond its theoretical contributions. Corporate raiders in the 1980s used agency theory's vocabulary to justify takeovers as necessary to overcome managerial entrenchment and 'unlock shareholder value.' Business schools taught the framework to future executives and board members. Compensation consultants designed equity packages implementing its recommendations. The theory became institutional practice through this diffusion process, transforming American corporate governance.

Key Ideas

Corporation as nexus of contracts. The firm is not a productive institution but a legal convenience organizing contracting relationships—a reconception that dissolves obligations to workers, communities, and long-term capabilities.

Managers as self-interested agents. Executives are assumed to pursue personal interests at shareholder expense unless constrained by incentive alignment—a framework that justifies treating managerial autonomy as a problem requiring control rather than as a source of productive expertise.

Stock-based pay as alignment mechanism. Tying compensation to stock prices supposedly makes managers act like owners—in practice, it makes them act like short-term speculators optimizing for price movements rather than productive capability.

Shareholder primacy as theoretical conclusion. Agency theory provides the intellectual foundation for treating maximization of shareholder value as the sole corporate purpose—a normative claim presented as positive theory.

Empirical failure of predictions. Lazonick's four decades of data demonstrate that stock-based compensation produced extraction, underinvestment, and innovation decline—the opposite of the productive alignment Jensen and Meckling predicted.