The venture begins before anyone can know what it will earn. Wages must be promised, equipment ordered and a place secured; the customers arrive later, if they arrive at all. Some hazards can be named and insured against. The building may burn, a shipment may be lost, a supplier may fail to deliver. Yet a policy covering those events leaves the central judgment exposed: whether this particular arrangement of people and things will produce something the world wants at a price that meets its cost. The entrepreneur commits resources on an answer that the future has yet to give.
Frank Knight made that exposure the center of his account of profit. In Risk, Uncertainty, and Profit, published in 1921, he distinguished probabilities established by a calculation or a sufficiently dependable classification of cases from the estimates on which much business conduct rests. The distinction concerns the grounds of a judgment. A fair die supplies a defined set of alternatives; a useful insurance table groups experiences whose likeness matters to the loss being insured. The decision to build a new factory also draws on experience, but deciding which experiences belong together is part of the contested judgment. There may be no sufficiently dependable class from which to read the probability of this venture's success.
Uncertainty in this sense can receive a price. The buyer can make an offer, the founder can state a forecast and the lender can demand a margin. Knight's argument makes those numbers more interesting because it asks how much confidence their makers have in them. A price of forty may represent a careful estimate of a payoff, a concession to someone who doubts the estimate, or the most a buyer can afford to lose. Writing the number down settles the terms of an exchange; it does not settle the reliability of the belief beneath it. Even the distinction between the familiar risk and the unfamiliar venture is a working judgment, open to revision when circumstances change.
The firm gives that judgment a distribution. Its workers, landlords and lenders bargain for payments specified in advance, while the owner claims the remainder. Knight describes a specialization of responsibility and control: someone chooses what to attempt and offers others a degree of protection against the result. The remainder has no corresponding promise. It is what the venture realizes after the agreed claims have been met. An ordinary wage can enter the founder's own account for work performed; a return on supplied capital can enter it too. The pure profit Knight wants to explain is the residual beyond those ordinary payments.
Consider an imagined undertaking whose committed costs, including ordinary compensation for the founder's work and capital, total eighty. Its receipts will be either one hundred and twenty or sixty. The residual will therefore be a profit of forty or a loss of twenty. If the founder assigns equal probability to the two outcomes, the expected residual is ten. If she assigns only a one-in-four chance to the better outcome, it is minus five. These are two assessments of the same list of possibilities. Neither assessment changes what will be paid when the result arrives; each changes whether committing the eighty appears worthwhile. The promise of profit lives in the assessment, and the profit itself lives in the eventual remainder.
The example also shows why profit is an awkward kind of wage. The entrepreneur can perform the same act of commitment, with the same care and the same resources at stake, and receive forty or lose twenty. Nor is a positive average secured merely because many people undertake such acts. At a one-in-three chance of the better result, the example's expected residual is zero. Knight himself allows that entrepreneurs as a class can incur a net loss when their confidence bids the prices of productive services beyond what the ventures realize. Competition among people eager to bear uncertainty can enrich those selling inputs while disappointing those who believed they would earn the remainder.
Insurance can improve this position without taking it over. A policy may compensate the owner if the factory burns while leaving her exposed to a product nobody buys. A supplier guarantee may make a promised delivery more dependable while shifting attention to the supplier's capacity to honor it. A partner may share the financial consequences while sharing control over the decisions that produce them. Such arrangements change the exposure and its distribution. Their usefulness does not require every remaining uncertainty to disappear. Indeed, the terms of the protection may themselves depend on judgments about behavior, verification and the strength of the person promising to pay.
The same qualification reaches the supposedly certain incomes on the other side of the firm. A contractual wage is a claim; its payment still depends on the arrangements that make the claim effective. A worker may lose future employment, a supplier may be left unpaid and a lender may receive less than the face of a debt. Knight closes his chapter by recognizing partial guarantees and distributed entrepreneurship: people who trust a manager may commit their services or property without complete security, and thereby share the exposure. The residual claimant can bear a substantial danger without bearing every danger created by the enterprise.
Judgment nevertheless has a task. The absence of a dependable frequency does not make investigation useless or every forecast equally good. One founder may understand a customer, a process or an institution that another has misread. She can test a small commitment before making a larger one, negotiate room to revise a decision or abandon a plan when its premise fails. These are ways of changing what must be faced, as well as ways of choosing who faces it. Their value has to be assessed against their cost. An inexpensive trial can be sensible where an irrevocable commitment would be rash; delay can also sacrifice an opportunity. There is no single premium that decides all these choices in advance.
A successful outcome is evidence to interpret within such a comparison. Knight treats repeated performance as a possible basis for judging ability while warning that a good decision can fail and a poor one can succeed. One observed profit cannot supply all the probabilities needed to distinguish the rival explanations of that profit. More evidence may help: what the founder knew, what alternatives she rejected, how comparable decisions turned out and whether the result depended on protection from competition. The same forty in the example could follow from insight or an unexpectedly favorable event. Its amount alone cannot tell the history of the judgment that earned or happened upon it.
The question of what is earned therefore follows the path of the exposure. Committing resources to a useful venture can be a service; absorbing a loss that others could poorly withstand can be another. The size of a favorable residual does not measure either service on its own. It also reflects the prices at which resources were secured, the alternatives open to their owners and the events that occurred after the bargain. To explain why a residual exists is to identify an economic position. To justify its whole amount as a wage requires an account of what that position actually contributed and what burdens remained elsewhere.
The residual names who receives what is left; it does not, by itself, tell us who bore everything that could go wrong.
Knight gives the volume a sharper object of inquiry: the person who commits resources while accepting that judgment may fail. The reward of that commitment is contingent, its protection of others incomplete, and its relation to ability a matter for evidence. The risks that can be laid off have found another bearer; the risks left behind still have to be located. The next plate follows that trail from the person who collects the return to the people whose lives or claims absorb the loss. The argument remains open.