Firms: Organization, Contracts, and Governance

So far this course has treated the economy as a web of trades between separate people. But most production happens inside firms, where a boss tells people what to do, output is shared rather than individually sold, and nobody negotiates a fresh price for every task. If markets and prices coordinate strangers so well, why does so much of economic life take place inside organizations that suppress the price mechanism and replace it with command? We answer in pieces: why teams produce more than scattered individuals, who owns a firm, how people protect investments that expose them to a partner, and how the corporation lets strangers pool wealth.

A firm does not abolish markets. It chooses a boundary: some tasks are coordinated by continuing relationships and authority, while inputs, finance, workers, and finished products still meet outside prices and alternatives. The boundary shifts when contracting, monitoring, adaptation, technology, or legal costs change.

Firms Exist Because Team Output Cannot Be Divided Up

Start with the gain that makes firms worth forming: people can produce more together than apart. Suppose Jack working alone produces output worth $5 and Jill alone produces $7. Apart they make $12; together they make $15, creating $3 in team value.

How should the team split $15? You might say “pay each person what they add to the total.” Try it: without Jack the team would have only Jill’s $7, so Jack adds $15 minus $7, or $8; without Jill it would have only Jack’s $5, so Jill adds $15 minus $5, or $10. Add those and you get $18, more than the $15 there is to divide. The arithmetic is impossible.

Jack and Jill	Value produced
Jack, working alone	$5
Jill, working alone	$7
Total, working separately	$12
Jack and Jill, working together	$15
Jack’s “marginal product” in the team ($15 − $7)	$8
Jill’s “marginal product” in the team ($15 − $5)	$10
Sum of the two marginal products	$18

The two contributions sum to more than the joint output because the output is genuinely joint, or non-separable: there is no honest way to slice the $15 and say “this part is Jack’s doing, that part is Jill’s.” The gain belongs to neither alone.

So if pay cannot track each person’s slice, what sets it? Competition. Each member earns roughly what he could get on his best alternative team elsewhere, and if the whole team produces less than the sum of what its members could earn apart, it will not form or last. (How much a worker is worth to an employer connects to the demand for labor, our next topic.)

When Contributions Can’t Be Metered, People Are Tempted to Shirk

The non-separability that makes teams valuable also makes them vulnerable. If no one can measure exactly what you contributed, you can ease off and let the others carry you. Economists call this shirking, and the general problem it belongs to is moral hazard: when someone can capture the gains from cutting corners while others bear the cost, and the cutting is hard to detect, expect more of it.

Moral hazard is everywhere, not just on the shop floor: an employee paid by the hour, a manager spending other people’s money, a depositor whose account the government guarantees. Because shirking is the cost of team production, somebody must watch for it, and monitoring is itself a job that creates value. This answers a puzzle about sports: the coach never plays and the conductor never makes a sound, so what do they contribute? They monitor. A player at full tilt cannot also bench his slacking teammates, and the conductor adds nothing to the noise and everything to the music by judging each performance.

The Owner Is the Residual Claimant, and That Is Why He Monitors

If monitoring is so important, who does it? You cannot literally “own a firm”; what you own is the firm-specific resources whose value rises and falls most with how well the firm does. That holder is the owner, with a special claim: after everyone else has been paid their agreed wages, rents, and interest, whatever is left, profit or loss, is his. We call him the residual claimant.

Residual claims are one powerful response to the monitoring problem. The claimant benefits from cost savings and bears losses after fixed promises are paid, so authority over hiring, direction, and monitoring can be joined to financial consequences. But the mechanism is not frictionless. Diversified shareholders may monitor little; managers possess information owners lack; limited liability caps some downside; and contracts, regulation, or external effects can place costs elsewhere. Boards, incentive pay, audits, lenders, worker voice, reputation, and product-market competition supplement the residual claim.

A firm is a network of contracts, but that does not make every term equally informed, costless to leave, or legally enforceable. Employees can quit and employers can dismiss within the contract and law, while firm-specific skills and search costs can make either side dependent. For-profit, nonprofit, cooperative, and government organizations also use different control systems. A salable residual claim supplies a direct market-value signal; mission, boards, donors, budgets, elections, professional norms, and legal duties can supply other signals. The question is which behavior each system rewards, what it can measure, and how errors are corrected.

Nontransferable claims provide a useful mechanism rather than a universal verdict. If a worker, resident, or manager cannot sell a stake whose value reflects future maintenance, less of that future gain is capitalized into a personal asset. Repeated membership, norms, user rules, public accountability, or mission can still reward stewardship. This connects to T5b: open access, governed common property, and private ownership are different institutions, not synonyms for good and bad care.

Monitoring Devices Align Incentives Where Watching Is Costly

Watching every worker every minute is expensive, so firms lean on devices that make people want to behave even when no one is looking. The simplest is to let someone else monitor for free: a restaurant cannot easily watch whether a waiter is attentive, but the customer can, and the tip turns the customer into an unpaid monitor whose reward tracks the service he just got. A second is the premium wage: pay a worker noticeably more than he could earn elsewhere, and getting fired now costs him that extra stream for years, so the threat of losing a good thing keeps him honest more cheaply than constant supervision would. Up-or-out promotion, deferred compensation, and tenure run on the same principle: load part of the reward into the future, payable only to those not caught slacking.

Specific Investments Create Dependence, and Dependence Invites Hold-Up

The next problem arises the moment someone sinks money or effort into an asset worth far more in one relationship than anywhere else, a specific or dependent investment. Picture a refinery erected next to the one pipeline that can feed it, or a worker who spends years learning skills useful only to his current employer. Once the investment is sunk, its owner is exposed: the partner he depends on can demand better terms, threatening to walk unless he renegotiates. This is the hold-up problem, having you over a barrel because you cannot cheaply take your investment elsewhere. It runs only as far as the value you would lose by relocating; the gap between what your asset is worth here and in its next-best use is the prize you fight over.

The worker who learns firm-specific skills is exposed to hold-up from the other direction: he now earns more here than anywhere else, so the employer can threaten to cut his pay toward his outside option. The very devices that deter shirking shield against this: up-or-out promotion forces the firm to decide rather than string an underpaid junior along, and tenure and seniority guarantee a dependent employee that his firm-specific value will not be confiscated once he can no longer credibly leave. The institution that disciplines the worker also protects him.

The same dependence runs through ordinary commerce wherever two things are worth more together than apart. A ballpark and its parking lot are complements, and if they have different owners their interests clash, since each would like the other to charge less to draw more fans. Manufacturers, wholesalers, and retailers along a distribution chain disagree about pricing for the same reason; the conflict is reciprocal. (One remedy, a manufacturer setting the minimum price a retailer may charge, follows the same logic; the full analysis of such vertical restraints belongs with our treatment of market power.)

Reputation, Contracts, and Hostages Protect the Dependent Party

Because hold-up is predictable, people design safeguards. The cleanest is integration: bring both interdependent assets under one owner. This removes bargaining across an ownership boundary, but replaces it with internal monitoring, information, and bureaucracy problems. Reciprocal dependence or an exchange of hostages can make opportunism costly to both sides. Long-term or exclusive contracts can fix terms before investment, but no contract specifies every future contingency and exclusivity can also weaken outside competition. The relevant comparison is among imperfect safeguards.

A studio contract illustrates risk sharing. A studio invests in many unknown performers and may seek a long claim on the few successes to cover the failures. The performer receives training, financing, and an opportunity while surrendering some later bargaining freedom. That mechanism explains why a successful performer may earn below a later spot-market offer; it does not by itself establish that every term was informed, competitive, fair, or worth enforcing. Bargaining power, alternatives, disclosure, and contract law matter too.

Reputation is another safeguard: others’ belief about future behavior can lower inspection and contracting costs. A brand makes some of that reputation portable, and expected future sales give the seller something to lose from disappointing customers. The discipline is stronger when quality is observable, information travels, future business matters, and customers can switch; it is weaker when defects appear slowly or exit is costly. Franchising adds mutual monitoring: the franchiser protects a shared name while a franchisee invests in a local outlet. It can align incentives, but also creates conflict over standards, fees, territory, and control.

Reputation is therefore neither instant nor sufficient. Competition and exit strengthen it, while law, audits, accreditation, boards, elections, and complaint systems may matter where exit is limited. The right question is what future value the decision-maker can lose and who can credibly impose that loss.

The Corporation Pools Vast Wealth Through Limited Liability and Tradable Shares

Now look at the legal form that lets firms grow enormous. Of the three ways to own a business, a proprietorship has a single owner and a partnership has several, but a corporation is a separate legal entity, distinct from the people who own it, carrying two features that change everything: limited liability and freely transferable shares.

Limited liability means an owner can lose only what he put in, never more. That is the standard rule for a shareholder’s liability for corporate debt, not immunity for the shareholder’s own misconduct, personal guarantees, or every statutory obligation. By capping passive investors’ exposure, the corporate form makes diversification and participation in large risky ventures easier.

The second feature is transferable shares. Shares of a public corporation can ordinarily change hands without renegotiating the firm’s contracts, although securities law, charters, private-company agreements, trading windows, and market liquidity can limit transfer in practice. Continuity despite an owner’s death or sale gives the corporation a life separate from any shareholder. Transferability and limited liability together let many investors pool capital while retaining an exit option.

Firm count and consumer choice are not the same statistic. Wider transport and communication can reduce the number of firms while expanding the sellers each buyer can reach; mergers can also reduce meaningful alternatives or entrench power. The answer depends on market boundaries, entry, and substitution. More fundamentally, firms do not exist only to pool wealth. They can lower the cost of repeatedly finding compatible teammates, negotiating interdependent tasks, adapting to change, and monitoring joint production.

The Market for Corporate Control Disciplines Managers

The corporation separates investment from day-to-day management. This is useful specialization because most savers do not want to run each company they finance. It also creates an agency problem because managers possess information and control resources owned for others. Governance must capture the benefit of specialization while limiting self-dealing, slack, and empire building.

Several controls address that agency problem: boards, fiduciary duties, audits, disclosure, incentive compensation, lender covenants, shareholder voting, product and labor markets, and takeover bids. Transferable shares make control contestable. Run a company poorly and its share price sags, because the price already reflects the firm’s long-run prospects. An outsider who expects to improve performance can bid for shares and replace managers. The threat can discipline management, but financing constraints, defensive tactics, regulation, information problems, and private benefits can also produce failed or value-destroying takeovers.

This reframes the “corporate raider.” Because today’s share price already capitalizes the firm’s expected future earnings, a buyer who pays a premium is betting that control, synergy, or information is worth more than the market currently expects. The premium is a forecast, not proof: the bidder may improve operations, transfer value among claimants, enjoy private benefits, or simply overestimate the gains. Prior overspending remains a sunk cost, but future integration and financing costs are not.

That premium is a bet on durable ability, and here a trap waits. A single good year reflects skill, conditions, and luck. Regression to the mean means an extreme result partly produced by temporary luck is likely to move closer to average. Treating one lucky year as proof of lasting ability is the regression fallacy. Competitive bidding penalizes buyers who repeatedly overpay for noise and rewards better forecasts, but market prices can still be mistaken until evidence arrives. The mechanism creates pressure toward correction; it does not make markets infallible.

Control contests are governed institutions, not a frictionless auction. Proxy rules, voting rights, staggered boards, disclosure, fiduciary duties, takeover defenses, and securities law affect who can challenge management and at what cost. Some defenses protect bargaining value or long-term investment; others can entrench managers. Their effects must be evaluated rather than inferred from whether incumbents or bidders favor them. A firm can also report positive accounting income while disappointing prior expectations, causing its market value to fall.

Corporate Claims Arise from Contracts, Ownership, Law, and Harm

A shareholder acquired his claim by a prior agreement: he bought in, on terms, from the owners. Shareholders are residual claimants: after contractual and legal claims are met, the remaining value rises or falls with the firm. Other stakeholders can hold different claims. Employees, suppliers, lenders, and customers may have contracts; workers and communities may hold statutory rights; people harmed by the firm’s conduct may have tort or regulatory claims even without a prior bargain.

Corporate purpose and fiduciary duty also depend on legal form and jurisdiction. Ordinary Delaware corporate law centers directors’ fiduciary duties on the corporation and stockholders as residual claimants, while a Delaware public-benefit corporation must balance stockholder interests, the interests of people materially affected, and a stated public benefit. Neither form erases contracts, labor and environmental law, consumer protection, or liability for harm. The analytical task is to identify the source, priority, and enforceability of each claim rather than divide the world into owners and people to whom nothing was promised. See Delaware’s ordinary limited-liability rule and public-benefit corporation duties.

One last lesson from the market for control is worth stating plainly: a cost is not the same as a waste. Eliminated positions impose real losses on workers, while a genuinely more productive combination can create benefits elsewhere. Whether a particular merger does so is empirical, and distribution remains relevant even when total value rises.

In the consumer-sovereignty benchmark, the firm’s ultimate controller is none of them; it is the consumer. Demand and exit discipline a firm from below, while investors, workers, lenders, law, and governance discipline it through other channels. Consumer control is weaker with market power, switching costs, subsidies, incomplete information, or harms imposed on nonbuyers. No single control is universal; the firm is governed by their interaction.

For Further Reading

Want to explore the source material? This lecture draws on the following chapters from two books by Armen A. Alchian and William R. Allen:

• Universal Economics (Liberty Fund, 2018): Ch. 22, “Teamwork and Firms”; Ch. 23, “The Firm’s Control and Reward Structure”; Ch. 24, “Protecting Your Dependencies”; Ch. 25, “Dependency Assurance by Reputation and Predictable Price”; Ch. 27, “The Corporate Firm”; Ch. 28, “Competition for Control of the Corporation”.
• Exchange and Production, 3rd ed. (Wadsworth, 1983): Ch. 9, “Business Firms: Ownership, Control, and Profits”.

• Full pinned UE 22–25, 27–28 and E&P 9 sections read; current Delaware ordinary and public-benefit corporation law checked.
• Residual claims remain central but are compared with boards, contracts, pay, audits, lenders, worker voice, reputation, mission, law, and product/labor markets.
• Integration, long contracts, reputation, takeovers, and regression correction are taught as safeguards with costs and fallibility, not self-validating solutions.
• Corporate claims now distinguish residual, contractual, statutory, tort, regulatory, and public-benefit sources.
• Three objectives/takeaways, eleven Core anchors, fifteen coverage anchors, the Jack-and-Jill table, and migration routing preserved.
• Canonical words 3,657 -> 3,008; budget 3,665 -> 3,016, retaining 8 headroom.
• Local closeout: default/Core/coverage lint PASS; t09 extension lint PASS; migration 320/0; fixtures 17/17; policy-claim YAML and diff check PASS. —>