Developed by Stephen A. Ross (The Economic Theory of Agency, 1973) and by Michael C. Jensen and William H. Meckling (Theory of the Firm, 1976), after Adolf A. Berle and Gardiner C. Means's split of ownership from control (The Modern Corporation and Private Property, 1932).
Analyzes any arrangement where one party acts for another, naming where their goals split, what the agent knows that the principal cannot see, and what oversight costs.
Eight things a franchisee does for the brand, sorted by how far interests split and how hard the work is to see
A sandwich brand (the principal) relies on its franchisees (the agents) for eight things. The first is placed as an example. Select each other one, then the box for how far the franchisee's interest departs from the brand's and how costly it is for the brand to check: each box names the fix the gap calls for. Show the sort to compare yours. Illustrative.
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
- Not placedPlacedTrust itMonitor itPay for the outcome
How far interests split
Cost for the brand to check
Your sort matched 0 of 8.Your sort matched 1 of 8.Your sort matched 2 of 8.Your sort matched 3 of 8.Your sort matched 4 of 8.Your sort matched 5 of 8.Your sort matched 6 of 8.Your sort matched 7 of 8.Your sort matched 8 of 8.
- Trust it
- Interests line up: a standard clause and an occasional look are enough; heavy oversight would cost more than it saves.
- Monitor it
- Interests split but the work is cheap to see: write it into the contract and check it, with audits, system data, or inspections.
- Pay for the outcome
- Interests split and the work is costly to see: tie the franchisee's reward to a result the brand can measure, so the agent's interest moves toward the principal's.
The sort
- Reporting true sales: Monitor it. Underreporting saves the franchisee royalties, and a till system that reports to the brand makes the check nearly free.
- Food-safety steps: Monitor it. Skipping steps saves labor, and inspections and sensor logs make them cheap to check.
- Opening hours: Monitor it. Hours are easy to see from the till and the website, so a contract term and a check do the job.
- Friendly service: Pay for the outcome. Courtesy costs effort and no one can watch every shift, so pay a bonus on customer scores.
- Local marketing effort: Pay for the outcome. The brand gains from every visit, the franchisee pays for the effort, and effort is hard to see, so match funds against results.
- Keeping the store clean: Trust it. The franchisee bears most of the cost of a dirty store, so interests are close, and a quarterly visit is enough.
- Hiring enough staff: Trust it. Short staffing hurts the franchisee's own sales, so interests mostly line up, even though staffing on each shift is costly to check.
- Using approved suppliers: Monitor it. The margin pulls the franchisee off the list, and purchase records make the check cheap.
An example
A sandwich brand with 140 franchised stores was rewriting its franchise agreement, and the first draft piled on oversight: more inspections, more reports, more penalties, for everything. Franchisees threatened to walk, and the brand's own field team said it could not police half of it.
The head of franchising sorted what the brand depended on franchisees for along two questions: how far does the franchisee's interest depart from the brand's on this, and how costly is it for the brand to see the work? Reporting true sales, food safety, opening hours, and approved suppliers split interests but were cheap to check, so they became contract terms checked through the till system, sensor logs, and purchase records. Friendly service and local marketing split interests and were costly to watch, so the brand moved to pay on outcomes: a royalty discount tied to customer scores and matched funds for marketing that brought in visits. Store cleanliness and staffing mostly served the franchisee's own sales, and the draft's clauses on them shrank to a line each.
Why it works
Whenever one party acts for another, two things open a gap. The agent's interests differ from the principal's, and the agent knows or does things the principal cannot see. Either alone is harmless: aligned interests need no watching, and visible work needs no trust. Together they create agency costs, and the usual reflex, more oversight everywhere, spends money where it buys nothing.
Agency analysis asks where the two gaps coincide, then matches the fix to the gap. Where the work is cheap to observe, monitor it and write it into the contract. Where it is costly to observe, pay for an outcome the principal can measure, so the agent's interest moves toward the principal's. Where interests already line up, leave it alone. The result is less oversight overall and more of it in the places that count.
The thinking behind it
Michael C. Jensen Michael C. Jensen, a financial economist at Rochester and Harvard, wrote the paper known as Theory of the Firm (1976) with William H. Meckling in the Journal of Financial Economics. The paper founded agency theory as used in corporate finance. It treats the firm as a set of contracts, and it locates agency costs in the fact that a manager who owns only part of the firm bears only part of the cost of acting in their own interest.
Stephen A. Ross Stephen A. Ross, a financial economist known for arbitrage pricing theory and the binomial option pricing model, wrote The Economic Theory of Agency: The Principal's Problem (American Economic Review, 1973). It gave one of the first formal models of the agency problem from the principal's side: how to design a payment that leads an agent to act in the principal's interest. That is the question behind the figure's top-right box.
Adolf A. Berle Adolf A. Berle, a Columbia law professor and diplomat, wrote The Modern Corporation and Private Property (1932) with the economist Gardiner C. Means. The book showed that ownership and control had separated in large American corporations: dispersed shareholders could not oversee the managers who ran the firm. It is a foundation of corporate governance and of the agency theory that followed four decades later.
How to run it
- Name the principal and the agent. Be precise: the brand and one franchisee, the board and the CEO, the company and a vendor.
- List what the principal depends on the agent for. Break the relationship into specific tasks and decisions, not one overall judgment.
- Rate how far interests split on each. Ask who gains and who pays if the agent cuts the corner.
- Rate how costly each is to observe. Count what it takes to see the work: systems data, inspections, or nothing short of standing in the room.
- Monitor where checking is cheap. Write those tasks into the contract and check them with the cheapest reliable source.
- Pay for outcomes where checking is costly. Tie the agent's reward to a result the principal can measure and the agent can influence.
- Leave the aligned tasks alone. Drop oversight that costs more than the gap it closes, and review the sort each contract cycle.
Where it goes wrong
The common failure is uniform oversight. Every task gets the same inspection and reporting, the agents resent it, and the principal spends most of its effort where interests already line up. Sort the tasks first and spend oversight only where the two gaps meet.
A second failure is an outcome measure the agent can game. A bonus on customer scores invites coaching customers to give high scores; a bonus on sales invites discounting. Pick measures the agent influences but cannot easily fake, and spot-check them. The third is ignoring the agent's side. A contract that loads all risk on the agent drives away the good ones; ask what each fix costs the agent, and price it in.
When to use something else
- Moral Hazard Analysis. When the main worry is that insurance or a guarantee makes the protected party take more risk.
- Mechanism Design. When the task is designing a whole set of rules so that every party's best move serves the goal.
- Skin in the Game. When the question is whether the people making decisions share in their downside.
Where it sits in SOLVE eX
Go to the source
- The Economic Theory of Agency: The Principal's Problem, American Economic Review, 1973
- Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics, 1976
- The Modern Corporation and Private Property, 1932