Apprenticeship indentures

Why would a master craftsman spend years training the apprentice who will one day compete with him?

A joiner takes a boy of fourteen into his house, feeds him, houses him, and over several years hands him everything he knows — the tricks that took him twenty years to learn and that no book contains. At the end of it the boy is entitled to set up shop, possibly in the same street. Why would anyone do this? And notice the question has a mirror image: why would the boy's family pay for the privilege, and then trust a stranger to actually teach rather than simply use a free pair of hands for seven years?

Begin with what the master gets. Not nothing: an apprentice is labour. But look at how the value is distributed over time. In the first year a fourteen-year-old ruins stock, needs constant supervision, and consumes board and lodging — he is a net cost. By the fourth or fifth year he is doing real work, at the price of his keep. So the arrangement is an investment: the master pays early and collects late.

That tells us where the danger lies. Suppose the apprentice, once competent, simply leaves — or is poached by a rival who never paid the training costs and can therefore offer him more. The master has borne the whole cost and collected none of the return. Any master who foresees this refuses to teach anything valuable, and the craft cannot reproduce itself.

Now run it from the other side, because the symmetry matters. What stops the master pocketing the premium, working the boy at a labourer's tasks and teaching nothing? By the time he discovers the fraud, the years are gone.

So we have two parties, each of whom must commit before the other can safely perform, and neither of whom can verify the other's future conduct. What would you build to hold that together? Look at what they actually built, feature by feature, and each one answers one half of the problem.

The term is long and fixed — in England, seven years by custom and, after the Statute of Artificers of 1563, by law for most trades. That length is not mystical: it is long enough that the productive later years repay the unproductive early ones, which is exactly what makes the master willing to teach in year two.

The contract is written, indented, and enrolled with the guild or town, so a third party holds the evidence and either side can be brought before it. The parents pay a premium up front, scaled by trade, which prices the training the master is about to give away. The apprentice lives in the master's house, under his discipline — cheap surveillance against absconding, and the reason apprenticeship was governed as a domestic relation rather than merely an economic one.

And the decisive piece: the reward for completing is not a certificate but the freedom — the right to trade in that town or craft, which only the guild or corporation can grant, and grants only on a completed term. That is what makes the threat credible. The runaway does not merely break a promise; he forfeits the only legitimate route to practising the trade he has just learnt. Meanwhile a master who taught nothing risked complaint to the guild, which had reason to care, since shoddy work by its members devalued everyone's mark.

Two refinements, and then the honest caveat.

First, the premium and the labour are substitutes: where parents paid a large premium the apprentice was effectively purchasing tuition, and terms could be softer. Where no premium was paid — pauper and parish apprentices bound out by the poor law — the child's labour was the whole payment, and the training could be nominal. The same legal instrument covered both, and it is a mistake to read the second through the first.

Second, this is not a claim that guilds were benign. Everything above works just as well as a barrier to entry: a seven-year term keeps out competitors and restricts the supply of masters. The historiography is genuinely split — Stephan Epstein argued that guild apprenticeship solved exactly the commitment problem sketched here and so spread skills, while Sheilagh Ogilvie reads guilds primarily as rent-seeking bodies whose training function is overstated.

There is also awkward evidence for the tidy story. Work on London indentures, Patrick Wallis's especially, found that a very large share of apprenticeships — by my recollection of those findings, something close to half or more — ended before the full term without either side going to law, which suggests a far more flexible, negotiated arrangement than the contract's language implies. I flag that proportion as recalled rather than verified here.

That gives us a falsification test. If indentures solve the training-investment problem, terms and premiums should track how costly and how portable the skill is, and completion should be what unlocks the trade. The account would be refuted if term length turned out unrelated to skill complexity, or if those who never completed practised the craft as freely as those who did — in which case the seven years were buying entry restriction, not commitment.

• How parish apprenticeship under the poor law used the same instrument for an entirely different purpose.
• Why the system decayed where factory production made portable craft skill less valuable.

Every term the collection defines is gathered in the glossary.