Risk vs Uncertainty

Frank Knight's distinction between risks that can be measured and uncertainty that cannot, and why it matters for insurance, business and policy.

In everyday speech, “risk” and “uncertainty” mean much the same thing. In economics, they are often distinguished. The distinction goes back to the economist Frank Knight, who wrote about it in his 1921 book Risk, Uncertainty and Profit.

Measurable risk

Knight used risk to describe situations where we do not know what will happen, but we can measure the probabilities. The chance of rolling a six with a fair die is one in six. Insurers know from large amounts of data roughly how many houses in a region will catch fire each year, or how many 40-year-old drivers will have an accident. Because these probabilities are known, risk can be priced and insured.

True uncertainty

Uncertainty, sometimes called Knightian uncertainty, describes situations where we cannot reliably measure the probabilities at all. Will a brand-new technology succeed? Will a war break out next year? How will an unprecedented pandemic spread? There is no long history of identical events to count.

Knight argued that true uncertainty cannot be insured, because no one can calculate a fair price. He saw bearing uncertainty as the core role of the entrepreneur, and profit as the reward for taking it on.

Insurance for fire versus a new product

A homeowner can buy fire insurance at a reasonable price because insurers have centuries of data on house fires. But a company launching a completely new kind of product cannot buy insurance against customers ignoring it. No one has the data to estimate that probability. The company's owners must bear that uncertainty themselves.

Why it matters

The distinction helps explain several things:

Financial crises: models that treat the future as measurable risk can fail badly when truly new situations arise, as in 2008.
Policy decisions: for problems like climate change or new diseases, policymakers must act without reliable probabilities, which is why some favour building in safety margins.
Behaviour: experiments show many people dislike situations with unknown probabilities even more than those with known risks, a pattern called ambiguity aversion.

Treating every future event as a calculable probability

It is tempting to put precise numbers on everything. But for truly new events, the numbers can create false confidence. Recognising when we face uncertainty rather than measurable risk encourages caution, flexibility and backup plans.

Key takeaways

Frank Knight distinguished measurable risk from unmeasurable uncertainty in 1921.
Risk can be priced and insured because probabilities are known from data.
True uncertainty cannot be reliably priced, and Knight saw bearing it as the entrepreneur's role.
The distinction matters for financial crises, policy and how people behave.