# The Costs and Benefits of Ownership (Grossman & Hart 1986)
## π― ν΅μ¬ μ£Όμ₯
**Integration = ownership of residual control rights.** When contracts are incomplete, the party who owns the asset has the right to decide all matters not specified in the contract. This affects ex-ante incentives to invest.
## π‘ Null Breaking
**They Said**: Williamson's TCE explains integration through verbal logic and transaction costs.
**GH Said**: We need a **formal model**. Ownership matters not because it changes day-to-day operations (contracts can specify that) but because it changes **fallback positions** when unforeseen contingencies arise.
**Surprise**: Integration is about **ex-ante incentives**, not ex-post coordination. Ownership shapes relationship-specific investments before anything goes wrong.
## π Keep / Retire
### Keep β
- **Residual control rights**: Non-contractible decisions go to asset owner
- **Hold-up problem**: Fear of expropriation reduces ex-ante investment
- **Incomplete contracts framework**: Formal modeling of Coasean insights
- **Property rights logic**: Ownership as governance mechanism
### Retire β
- **Integration as binary**: BGM (2002) show integration serves relational contracts
- **Only material assets matter**: Rajan-Zingales (1998) critiqueβwhat about human capital?
- **No role for formal contracts**: Later work shows formal-relational interaction
## π§± λ
Όμ¦ λΈλ‘
### Block 1: Setup
Two parties need to collaborate. Contracts are incomplete (can't specify all contingencies). Parties make relationship-specific investments.
### Block 2: The Hold-up Problem
If you don't own the asset, I might expropriate your investment returns β you underinvest ex-ante.
### Block 3: Ownership as Solution (Partial)
Give ownership to the party whose investment is more important β that party invests more.
### Block 4: Trade-offs
But now the OTHER party underinvests! **Optimal ownership** balances these effects.
### Block 5: Integration Decision
Integrate (one party owns both assets) when joint ownership costs exceed benefits of specialized ownership.
## π Connections
### Builds On
- **Coase (1937)**: Incomplete contracts as foundation
- **Williamson (1975)**: Hold-up and asset specificity
- **Klein-Crawford-Alchian (1978)**: Vertical integration to avoid hold-up
### Built Upon By
- **Hart-Moore (1990)**: Extended to multiple assets, debt contracts
- **Baker-Gibbons-Murphy (2002)**: Integration decision serves relational contracts (formal + informal)
- **Gibbons (2025)**: Integration is ONE form of visible-hand coordination
### Critiqued By
- **Rajan-Zingales (1998)**: What about human capital? Can't "own" people
- **Holmstrom (1999)**: Downplays incentive intensity vs. authority
- **Whinston (2003)**: When does integration change incentives?
## π Model Structure
### Timeline: G-A-S-D
```
t=0: Governance (G)
ββ Decide ownership structure
t=1: Actions (A)
ββ Parties invest (non-contractible)
t=2: State (S)
ββ Contingency realized
t=3: Decisions (D)
ββ Owner decides use; parties bargain; payoffs realized
```
### Key Innovation
**Ownership matters through THREAT POINTS in bargaining**, not through actual control in equilibrium.
### Core Trade-off
```
More control for Party A
β A invests more β
β B invests less β
Integration optimal when:
Value of A's extra investment > Loss from B's reduced investment
```
## π Teaching Notes
### The "Aha!" Moment
Most students think integration improves ex-post coordination. GH insight: **ex-ante** incentives matter more!
### Example: Apple + Supplier
If Apple owns supplier's factory:
- **Ex post**: Doesn't matter (they'll negotiate efficiently anyway)
- **Ex ante**: Supplier won't invest in specialized equipment (fear of hold-up)
- **Solution**: Sometimes better for supplier to own factory
### Three Meanings of "Integration"
1. **Legal**: Common ownership
2. **Operational**: Vertical coordination (Lafontaine-Syverson: often zero!)
3. **GH**: Residual control rights
## π¬ Research Implications
### Predictions
1. Integration when one party's investment >> other's
2. Vertical integration in asset-intensive industries
3. Non-integration when both parties need strong incentives
### Empirical Challenges
- Hard to measure "residual control rights"
- Hard to observe ex-ante investments
- Hold-up problem is **avoided** in equilibrium (so not directly observable)
### Extensions
- **Hart-Moore (1990)**: Debt as allocation of residual control
- **Aghion-Tirole (1997)**: Formal vs. real authority
- **BGM (2002)**: Integration + relational contracts
## π Impact
### Academic
- **15,000+ citations**
- **Nobel Prize 2016** (Hart, Holmstrom)
- Created "Property Rights Theory" branch of organizational economics
### Practical
- Influenced M&A decisions: "Do we need control or just relationship?"
- Venture capital contracts: Who controls pivots?
- Open source: Why give away control?
## π Personal Notes
### Why This Paper Matters
First **formal model** of boundary of the firm. Proved you could mathematically analyze Coase's insights.
### Limitations Gibbons Notes
1. **Only material assets**: What about human capital, knowledge, relationships?
2. **No relational contracts**: BGM show integration serves informal agreements
3. **Binary choice**: Real world has many hybrid forms
4. **No coalition view**: Treats firms as unitary actors
### Connection to Gibbons (2025)
GH provided the **property rights pillar** of modern organizational economics. But Gibbons argues we now need:
- Less focus on "should we integrate?"
- More focus on "how do visible hands build equilibria?"
Integration matters, but it's **one choice among many** for visible-hand coordination.
## π― One-Sentence Summary
**Ownership of assets matters because it determines who controls unspecified decisions, which shapes ex-ante investment incentives in incomplete contracts.**
---
*"Integration changes threat points, not equilibrium outcomes."* β Grossman & Hart (1986)
*"GH gave us the property rights foundation; now we need the equilibrium-building superstructure."* β Gibbons (2025)