Ronald Coase worked the question out as an undergraduate, on the travels he made through American firms in 1931 and 1932, and published it in 1937 as The Nature of the Firm. It had not really been asked before.

Economics said that markets allocate resources efficiently through prices. If that is true, why does any firm exist? Why is a company not simply a set of individuals contracting with each other for every task, with prices sorting out who does what?

Look inside a company and you find the opposite of a market. Work is directed rather than priced. Nobody negotiates a fee with a colleague for a piece of work; somebody decides and it is done.

The answer: using a market is not free

Coase's insight is that transactions have costs of their own, and those costs are invisible in the standard account because the standard account assumes them away.

To buy something on a market you have to:

Find out what things cost. Who supplies this, at what price, at what quality. Search takes time and the time is a cost.

Negotiate and write a contract. For every transaction. A separate agreement each time somebody does a piece of work for you.

Enforce it. Check the work, chase what is late, resolve what is disputed.

For a one-off purchase these costs are trivial. For work that recurs daily, that is hard to specify in advance, and where quality is hard to verify, they are enormous.

So the firm exists as the alternative. Inside a firm, one open-ended contract — employment — replaces an unbounded series of specific ones. You do not renegotiate each morning. Somebody has the authority to direct the work, and that authority is what you bought instead of a price.

Where the boundary sits

This is the part that is useful rather than merely interesting, because it sets a limit in both directions.

A firm expands as long as organizing one more activity internally costs less than buying that activity on the market. It stops when the cost of internal coordination — management attention, communication, the principal-agent problem — exceeds what the market would charge.

That is a real boundary and it explains things that are otherwise puzzling. It is why companies outsource payroll and not their sales relationships; payroll is easy to specify and verify, so the market is cheap, while sales relationships are neither. It is why the same activity is inside one company and outside another. And it is why "should we hire or contract this" is not a question about budget. It is a question about how specifiable the work is.

Three factors push work inside:

Frequency. A task done daily makes the per-transaction cost of contracting unbearable.

Specificity. Work that requires knowledge of your particular business cannot be bought off a shelf, because the supplier's investment in learning you is only valuable to you — which makes both sides nervous about being held up.

Difficulty of measurement. If you cannot verify quality without doing the work yourself, a contract cannot protect you and direction is the only mechanism left.

What it says about your business

On hiring. The case for an employee over a contractor is strongest where the work is frequent, specific to you, and hard to evaluate from outside. Where it is none of those, employing somebody is buying coordination cost for nothing.

On staying a contractor. The reverse holds for the person selling. A contractor whose work is highly specifiable competes on price forever. One whose work requires knowing a client's business deeply is expensive to replace, which is the same fact seen from the other end.

On selling a business. A buyer is acquiring a bundle of arrangements that were cheaper to organize internally than to contract for. If all of that lives in one person's head, the buyer is not acquiring a firm — they are acquiring a person who will leave, which is exactly why a business that runs without the owner is worth more.

On productizing. Productization is deliberately lowering the transaction cost of buying from you — specifying the work so precisely that a client can purchase it without a negotiation. That moves work from the firm side of the boundary to the market side, on purpose, and it is why it scales.

Coase won the Nobel for this in 1991, fifty-four years after writing it.