Property Rights Theory of the Firm

The Property Rights Theory of the Firm—widely known as the Grossman–Hart–Moore (GHM) model—is the formal incomplete-contracts framework in organizational economics and business and economics theory that explains the boundaries of the firm, vertical and lateral integration, and asset ownership. Developed by Sanford J. Grossman and Oliver D. Hart in their 1986 Journal of Political Economy article "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration" and extended to multi-asset hierarchies by Oliver Hart and John Moore in their 1990 Journal of Political Economy article "Property Rights and the Nature of the Firm", the theory resolved a central question in transaction cost economics: if contracting difficulties and opportunism plague market transactions between independent firms, why do those same hazards not persist when the transaction is brought inside a single integrated firm, and how can integration ever reduce welfare? By defining ownership of a firm as possession of residual rights of control over nonhuman physical assets when contracts are incomplete, Grossman, Hart, and Moore demonstrated that integration shifts bargaining leverage in ex-post renegotiations, strengthening the acquiring owner's ex-ante investment incentives while weakening the acquired manager's incentives.

Limitations of Earlier Theories of the Firm

Prior to Grossman, Hart, and Moore, three major traditions sought to explain why economic activity is organized within firms rather than through spot market prices:

1. Neoclassical Production Function Theory: Treated the firm as a technological production set converting inputs into outputs, offering no explanation for why two stages of production reside under common ownership rather than trading across a market interface.
2. Principal–Agent and Nexus-of-Contracts Theories: Alchian and Demsetz (1972) and Jensen and Meckling (1976) viewed the firm as a nexus of contracts governed by monitoring and incentive schemes (principal–agent problem and moral hazard), but could not distinguish a long-term incentive contract between two separate firms from an employment contract inside one unified firm.
3. Coase–Williamson Transaction Cost Economics (TCE): Ronald Coase (1937), Klein, Crawford, and Alchian (1978), and Oliver Williamson (1975, 1985) argued in transaction cost economics that asset specificity combined with incomplete contracts creates ex-post holdup and haggling costs in market exchange, which vertical integration mitigates by replacing market bargaining with hierarchical fiat. Yet TCE struggled to formalize why an integrated corporation cannot replicate market incentives internally when separate ownership is superior, or what mathematically changes when Firm 1 buys Firm 2.

Specific Rights vs. Residual Rights of Control

Grossman and Hart (1986) provided a unified microfoundation by distinguishing specific rights of control (which can be delineated and assigned in an explicit contract) from residual rights of control—the legal right to decide how a nonhuman asset (machinery, factories, patents, client lists, or brand names) is used in any contingency not spelled out in a prior contract.

When states of nature, product specifications, and relationship-specific actions are too complex or multidimensional to be verified by a third-party court, ex-ante contracts are incomplete. Because inalienable human capital always belongs to the individual worker or manager, what changes under a merger or acquisition is who holds the residual right to exclude others from using the nonhuman physical assets if ex-post negotiations break down. Consequently, asset ownership determines each party's disagreement payoff (outside option) in ex-post bargaining, which in turn governs each party's marginal return to ex-ante relationship-specific investments.

Timeline, Non-Contractible Investments, and Asset Structures

Consider a two-party production relationship between Party 1 (an upstream supplier or research unit managed by Manager 1) and Party 2 (a downstream manufacturer managed by Manager 2), utilizing two nonhuman physical assets, a₁ and a₂. An ownership structure A = (A₁, A₂) partitions the physical assets {a₁, a₂} between the two managers across four organizational governance forms:

• Non-Integration (A⁰): Manager 1 owns asset a₁ (A₁ = {a₁}) and Manager 2 owns asset a₂ (A₂ = {a₂}) as two independent firms.
• Type-1 Integration (A¹): Manager 1 owns both assets (A₁ = {a₁, a₂}, A₂ = ∅), employing Manager 2 as a subordinate division head.
• Type-2 Integration (A²): Manager 2 owns both assets (A₁ = ∅, A₂ = {a₁, a₂}), employing Manager 1 as a subordinate division head.
• Joint Veto Ownership: Both managers hold veto rights over assets, which is typically dominated when investments are in human capital because mutual veto minimizes both parties' outside options.

The strategic interaction unfolds across three sequential dates under symmetric information between the two managers, coupled with non-verifiability by outside courts:

1. Date 0 (Contracting over Asset Ownership): The parties select an ownership structure A ∈ {A⁰, A¹, A²} and make upfront lump-sum transfers allowed by the Coase theorem. Because specifics of future trade and investments cannot be verified by a court, no state-contingent production or price contract can be written at Date 0.
2. Date 1 (Relationship-Specific Human-Capital Investments): Manager 1 and Manager 2 simultaneously and non-cooperatively choose relationship-specific investments e₁ ≥ 0 and e₂ ≥ 0 at private personal costs C₁(e₁) and C₂(e₂), where C_i′ > 0 and C_i″ > 0.
3. Date 2 (Ex-Post Bargaining and Trade): Once investments (e₁, e₂) are sunk, the parties observe their values and negotiate over whether to cooperate and trade.

Ex-Post Nash Bargaining and Disagreement Payoffs

At Date 2, if the two managers collaborate (pooling their human skills and both physical assets a₁ and a₂), they generate a total cooperative ex-post surplus S(e₁, e₂) = R₁(e₁) + R₂(e₂), independent of nominal asset ownership because all assets and both managers are deployed together in equilibrium.

If Date 2 negotiations break down, each Manager i walks away with only the physical assets in their owned portfolio A_i and earns an outside option (disagreement payoff) v_i(A_i, e_i). Because cooperation is efficient at Date 2 (S(e₁, e₂) > v₁(A₁, e₁) + v₂(A₂, e₂)), the two parties reach agreement and split the net gains from trade 50:50 via the symmetric Nash bargaining solution. Manager i's ex-post gross payoff U_i(A, e₁, e₂) at Date 2 is their disagreement payoff plus half of the net cooperative bargaining surplus:

First-Best vs. Second-Best Investment Incentives

In a first-best benchmark where investments (e₁, e₂) are contractible at Date 0, the parties choose (e₁^FB, e₂^FB) to maximize net social surplus W(e₁, e₂) = S(e₁, e₂) − C₁(e₁) − C₂(e₂), satisfying:

Under incomplete contracts, each Manager i at Date 1 chooses e_i non-cooperatively to maximize U_i(A, e₁, e₂) − C_i(e_i), yielding the decentralized Nash equilibrium first-order conditions:

Two key properties follow from these conditions:

1. Supermodularity of Asset Ownership and Investment: Hart and Moore (1990) assume that owning a larger subset of physical assets raises the marginal return to a manager's own human-capital investment in the disagreement state:

1. Specificity and Underinvestment: Because investments are relationship-specific, the marginal return when trading with the partner exceeds the marginal return in the disagreement state: ∂S/∂e_i > ∂v_i(A_i, e_i)/∂e_i. Consequently, both parties underinvest relative to the first best (e₁*(A) < e₁^FB and e₂*(A) < e₂^FB) under every ownership structure A.

The Unified Trade-Off of Vertical Integration

When Firm 1 acquires Firm 2 (moving from Non-Integration A⁰ = ({a₁}, {a₂}) to Type-1 Integration A¹ = ({a₁, a₂}, ∅)), two opposing incentive effects occur simultaneously:

• The Benefit of Integration: Manager 1 gains residual control over asset a₂ in addition to a₁, raising Manager 1's marginal disagreement return from ∂v₁({a₁}, e₁)/∂e₁ to ∂v₁({a₁, a₂}, e₁)/∂e₁ and increasing Manager 1's investment (e₁(A¹) ≥ e₁(A⁰)).
• The Cost of Integration: Manager 2 loses residual control over asset a₂ (becoming an employee with A₂ = ∅), lowering Manager 2's marginal disagreement return from ∂v₂({a₂}, e₂)/∂e₂ to ∂v₂(∅, e₂)/∂e₂ and decreasing Manager 2's investment (e₂(A¹) ≤ e₂(A⁰)).

Thus, GHM proves that selective intervention is impossible: transferring asset ownership to cure holdup against Party 1 automatically worsens holdup against Party 2.

Asset Complementarity, Independence, and Indispensability

Hart and Moore (1990) derived four propositions characterizing the optimal allocation of property rights:

1. Complementary Assets Should Be Owned Together: Two assets a₁ and a₂ are strictly complementary if neither asset has marginal productivity on its own without the other: ∂v₁({a₁}, e₁)/∂e₁ = ∂v₁(∅, e₁)/∂e₁ and ∂v₂({a₂}, e₂)/∂e₂ = ∂v₂(∅, e₂)/∂e₂. Consolidating both assets under either Type-1 Integration (A¹) or Type-2 Integration (A²) raises the acquiring manager's investment without reducing the acquired manager's investment, so integration dominates non-integration.
2. Economically Independent Assets Should Be Owned Separately: Two assets a₁ and a₂ are economically independent if acquiring the second asset adds zero marginal productivity to either manager's outside option: ∂v₁({a₁, a₂}, e₁)/∂e₁ = ∂v₁({a₁}, e₁)/∂e₁ and ∂v₂({a₁, a₂}, e₂)/∂e₂ = ∂v₂({a₂}, e₂)/∂e₂. Here, merging confers no investment benefit on the acquirer while depressing the target's investment, so non-integration (A⁰) dominates integration.
3. One-Sided Investment Dominance: If Manager 1's investment e₁ accounts for nearly all variation in cooperative surplus (∂S/∂e₂ ≈ 0), Type-1 Integration (A¹) is uniquely optimal.
4. Indispensable Human Capital: Manager 1 is indispensable if neither asset generates any marginal return to Manager 2's investment in the absence of Manager 1 (∂v₂({a₁, a₂}, e₂)/∂e₂ = ∂v₂(∅, e₂)/∂e₂), making Type-1 Integration (A¹) optimal.

Classic Case Studies and Global Value Chains

Grossman and Hart (1986) applied their model to the insurance industry, explaining why independent insurance agents own client renewal lists in property-casualty lines where ongoing agent servicing effort e₂ drives policy persistence, whereas life insurance companies own the renewal lists where upfront actuarial design and marketing e₁ dominate. Similarly, in international trade, Pol Antràs (2003) embedded the GHM model into a Helpman–Krugman general-equilibrium trade framework to explain intrafirm trade vs. offshore outsourcing: multinational firms vertically integrate foreign affiliates in capital-intensive industries (protecting headquarters investment e₁) and outsource to independent suppliers in labor-intensive industries (incentivizing supplier effort e₂).

Critiques, Access Models, and Reference Points

Subsequent theoretical literature identified important extensions to the baseline GHM paradigm:

• De Meza and Lockwood (1998) Outside-Option Principle: In non-cooperative alternating-offer bargaining rather than Nash bargaining, an outside option binds only when it exceeds the player's Rubinstein share S/2, and when it binds, the player captures 100% of its marginal increase.
• Rajan and Zingales (1998) Access Model: Showed that when investments specialize human capital to a physical asset, regulated access can outperform fee-simple ownership by preventing the owner from coasting on generic fallback rents.

For his foundational contributions to contract theory and the property rights theory of the firm, Oliver Hart shared the 2016 Nobel Memorial Prize in Economic Sciences with Bengt Holmström.

How does the Grossman–Hart–Moore property rights theory differ from Williamson's transaction cost economics?

While both frameworks rely on incomplete contracts, relationship-specific investments, and ex-post holdup, transaction cost economics (Williamson, 1975, 1985) emphasizes ex-post maladaptation and haggling inefficiencies during market renegotiation that hierarchical authority inside a firm resolves. In contrast, the GHM property rights model assumes ex-post bargaining is always efficient (Coase theorem at Date 2) and traces organizational inefficiency to ex-ante distortions in relationship-specific investments (e₁, e₂), modeling both market firms and integrated firms within the same incomplete-contract game.

Why can't two merged divisions inside one company replicate the incentives of two independent firms?

Under independent ownership (A⁰), each manager holds the residual legal right to walk away from negotiations with their own physical asset a_i, giving each manager credible bargaining leverage to capture a larger share of the return on their specific investment e_i. Once Firm 1 acquires Firm 2 (A¹), top management of Firm 1 holds the legal residual right to fire Manager 2 and retain asset a₂ whenever a dispute arises; any promise by Firm 1's CEO not to exercise that residual power in uncontracted contingencies is non-binding at Date 2.

Why is joint ownership with mutual veto power rarely optimal in the GHM model?

Under joint ownership where both Manager 1 and Manager 2 must consent to any use of assets {a₁, a₂}, if Date 2 bargaining breaks down, neither manager can deploy either asset in their disagreement state (A₁^veto = ∅ and A₂^veto = ∅). When investments (e₁, e₂) are embodied in the managers' human capital, mutual veto reduces both managers' marginal disagreement returns to their lowest levels simultaneously.

1. Grossman, S. J., & Hart, O. D. (1986). "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration." Journal of Political Economy, 94(4), 691–719. https://doi.org/10.1086/261404
2. Hart, O., & Moore, J. (1990). "Property Rights and the Nature of the Firm." Journal of Political Economy, 98(6), 1119–1158. https://doi.org/10.1086/261729
3. Hart, O. (1995). Firms, Contracts, and Financial Structure. Oxford: Oxford University Press. https://doi.org/10.1093/0198288816.001.0001
4. Antràs, P. (2003). "Firms, Contracts, and Trade Structure." Quarterly Journal of Economics, 118(4), 1375–1418. https://doi.org/10.1162/003355303322552829
5. Rajan, R. G., & Zingales, L. (1998). "Power in a Theory of the Firm." Quarterly Journal of Economics, 113(2), 387–432. https://doi.org/10.1162/003355398555630