Incomplete Contracts and the Internal Organization of Firms
GAICIL Insights
Every company runs on contracts that cannot say everything.1 That gap decides who gets to decide. Incomplete Contracts and the Internal Organization of Firms, by P. Aghion, N. Bloom and J. Van Reenen, surveys theory and evidence on decentralization within firms: how firms allocate decision-making authority, which shapes their ability to adapt, innovate and perform. The authors discuss how incomplete contracts shape views about the organization of decision-making and review evidence on decentralization's determinants and effects.
A hard question for boards adopting AI agents: write down what authority each system holds, as carefully as a delegation memo for a senior manager. The theory of incomplete contracts suggests that the unwritten parts are where the trouble will come from.
The theory is elegant. Because contracting over every possible state of the world is impossible, principals delegate control to agents as a way to commit in advance to letting agents capture some returns from costly activities such as collecting information on the best actions. Delegation motivates agents to invest in information benefiting the firm. Authority is a promise that effort will count.
The evidence shows wide variation. Like productivity, decentralization varies widely across firms and countries, with Scandinavian and Anglo-Saxon firms more decentralized than those from Asia and Southern Europe. Factors highlighted by incomplete contract theory help account for differences, especially heterogeneity and congruence of preferences, proxied by trust. Product market competition, human capital and firm size are robustly positively correlated with decentralization. Information technology also appears to foster decentralization. Trust, it turns out, is an organizational technology.
The authors flag substantial gaps between theory and empirics, call for matching empirical work more closely to theory and suggest ways to bridge the gap. The survey maps what is known and what remains to be learned.
Artificial intelligence is reshaping exactly these trade-offs. AI transforms the information available inside firms and the cost of processing it. Information technology appears to foster decentralization, and AI could extend the effect by giving local managers and employees better tools for analysis and decision-making. It could also enable centralization by giving headquarters better data on local conditions. How AI tilts the allocation of authority will shape firm performance. The same software can empower the branch or the boardroom.
Trust takes on a new dimension. Delegation depends on congruence of preferences between principals and agents, proxied by trust. As AI systems take on decision-making roles, questions of trust extend to machines. Firms must decide how much authority to delegate to AI systems, depending on alignment between the systems' objectives and the firm's goals. Incomplete contract theory offers a framework for such delegation. Delegating to an algorithm is an incomplete contract with an unusual agent.
Competition pushes authority outward. Competitive pressure may push firms to delegate to those closest to customers and operations. In intensely competitive AI-intensive markets, decentralized structures allowing rapid adaptation may pay.
Human capital matters too. Skilled workers are more likely to be given authority. As AI changes the skills firms need, decision rights may shift toward workers able to combine AI tools with judgment. Investing in human capital may support more decentralized, adaptive organizations. The employee who knows when to overrule the model becomes the one worth trusting.
For policy, organizational practices matter for productivity. Cross-country differences in decentralization may contribute to performance differences. Policies fostering competition, human capital and trust may encourage structures supporting innovation and adaptation.
The theory-evidence bridge is timely. AI provides new data on decision-making within firms, letting researchers test theories of organization more directly. This could accelerate progress in understanding how firms should be organized.
Seen whole, the paper's message is plain: internal organization is shaped by incomplete contracts, trust, competition, human capital and technology. Decentralization lets firms harness local information, and its extent varies widely. As AI transforms information within organizations, these insights frame how authority should be allocated among managers, employees and intelligent systems.
Introduction
We survey the theoretical and empirical literature on decentralization within firms. We first discuss how the concept of incomplete contracts shapes our views about the organization of decision-making. We then overview the empirical evidence on the determinants of decentralization and on the effects of decentralization on firm performance. A number of factors highlighted in the theory are shown to be important in accounting for delegation, such as heterogeneity and congruence of preferences as proxied by trust. Empirically, competition, human capital, and IT also appear to foster decentralization. There are substantial gaps between theoretical and empirical work and we suggest avenues for future research in bridging this gap (JEL O31, O32, O33, F23).
Takeaway
We began by surveying the theoretical literature on the organization of firms and the optimal decentralization of decision rights within firms. We discussed how the concept of incomplete contracts shapes this organization of decision making within firms. In particular, the inability to contract over all possible states of the world leads principals to delegate control to agents as a way to ex ante commit to letting agents expropriate some of the returns from costly activities, like collecting information on the best actions to take. We then overviewed some of the empirical evidence on the organization and management of firms, focusing on decentralization. We looked at within firm organization, especially decentralization following the theoretical survey, and we also focus on econometric studies of large scale firm databases rather than case studies. There has recently been a number of papers measuring management and organizational practices across firms and countries. Like productivity, decentralization varies a lot across firms and countries (e.g. Scandinavian and Anglo-Saxon firms are more decentralized than those from Asia and Southern Europe). A number of factors highlighted in the incomplete contract theory are shown to be important for accounting for differences in firm organization - in particular heterogeneity and the congruence of preferences (as proxied by trust). Several other factors appear robustly positively correlated with decentralization, such as product market competition, human capital and firm size.1
In terms of future work, we see two areas of opportunity. First, there is a need to match up empirical work more closely with the theory. Until recently, comprehensive datasets on measures of organization across many firms and countries were unavailable. Now that this gap in the core data infrastructure is being covered, there is a great opportunity for testing some of the theories of organizational economics. This is challenging, first because many of the important aspects in the environment emphasized in the theoretical literature are hard to match into empirical counterparts, and second because identifying the causal relationship between organizational changes and firm outcomes like productivity and growth is difficult. Even with strong measurement and tight links to theory, it is essential to identify the direction of causality in some of stylized empirical results we have identified above - for example, do more skilled managers enable decentralized decision making, or are skilled managers attracted to decentralized firms? We see the theory and empirics of the organization of the firm as one of the key growth areas in economics over the next twenty-five years.
Notes
- P. Aghion, N. Bloom & J. Van Reenen, Incomplete Contracts and the Internal Organization of Firms, 30 J.L. Econ. & Org. i37 (2014), https://doi.org/10.1093/jleo/ewt003. Takeaway source: 5. Conclusions, reproduced from a free author or repository copy (working-paper version). ↩
Cite as: P. Aghion, N. Bloom & J. Van Reenen, Incomplete Contracts and the Internal Organization of Firms, 30 J.L. Econ. & Org. i37 (2014), https://doi.org/10.1093/jleo/ewt003
Header image: Office interior 2022 queue.jpg by Rakoon, CC0, via Wikimedia Commons.
