The Rise And Fall Of The Medieval Guild

AThe verdict passed on the craft guild by nineteenth-century political economy was severe and, for a long while, unopposed: an association of masters who fixed prices, restricted entry and postponed every useful invention until it could no longer be resisted. That verdict was itself a period piece, framed by writers whose real quarrel lay with the regulated economies of their own day, yet it entered the textbooks and stayed there, and a survey of European economic history published in 1911 still treated the guild chiefly as an obstruction to be cleared away. Its appeal is easy to understand. Guild ordinances do indeed lay down who might trade, at what price and with how many hands, and a body that writes such rules invites the description of a cartel. What the ordinances cannot explain is why anyone outside the circle of masters, whether the councils that granted the charters, the parents who paid to place a son, or the customers who bought the goods, should have tolerated the arrangement for four hundred years.

BThe most sustained challenge to that reading comes from Ingrid Correia of the Bruges Institute for Urban History, who contends that the guild is better understood as a remedy for a problem of trust in training. Skill in most crafts was tacit, picked up at the bench and impossible to certify on paper, so a boy could not tell whether a master would really teach him, and a master could not tell whether the boy, once taught, would stay long enough to repay the trouble. Between 2004 and 2012 Correia transcribed 3,142 apprenticeship contracts drawn up in Leiden and Norwich, recording the term agreed, the fee paid and whether either party later returned to court. Where a guild stood behind the contract, 61 per cent of apprentices completed the full term; where no guild held jurisdiction, the figure was 38. She traced wages as well: journeymen who had finished a supervised term earned about 14 per cent more than men trained privately in the same craft.

CCorreia is careful about where her argument applies. The earnings advantage shows up only in crafts whose quality an untrained buyer could not judge, among them dyeing, goldsmithing and the fitting of locks; in trades whose output could be inspected at a glance, nails and coarse cloth being the obvious examples, membership made no measurable difference to what a man was paid. She concedes a second limit. Her figures come from two towns whose guild courts kept unusually complete registers, and a court that recorded little may also have enforced little, which would exaggerate the contrast she reports. Even so, the direction of the evidence is consistent. On her reading, the ordinances that look so much like restraint of trade were the price of a training system that no master could have financed on his own.

DPiero Casadei of the Aldgate Centre for Economic History rejects the inference rather than the figures. The boys who entered guild trades, he argues, were never a random sample: the fines charged on admission were substantial, and a father able to pay one was generally a man of some standing whose son would have prospered in any trade whatever. Casadei re-examined 900 of the Leiden contracts in 2019, adding the occupation of the father wherever the parish books recorded it, and found the wage advantage falling from 14 per cent to roughly 4 once that variable was taken into account. He does not claim to have disposed of the training story altogether. He insists only that most of what has been read as the effect of instruction is in truth the effect of who was let in. The weakness of his own exercise is plain enough: paternal occupation could be established for barely a third of the sample, and the men he failed to identify are precisely those least likely to have been prosperous.

EBoth accounts must in the end explain how the guilds lost their hold, and here the archives point away from the drama of abolition. Long before any government dissolved them, merchants were carrying wool out to villages where no ordinance ran, paying rural households by the piece and collecting the finished cloth by cart; in the country around Norwich the share of cloth completed beyond the town walls rose from roughly a tenth in 1560 to more than half by 1640. The guilds saw what was happening. Entry fines in the Leiden weavers' guild were cut by a third between 1591 and 1614, which is not the conduct of a body confident of its monopoly, and several guilds began accepting men who had served no formal term at all. The institution was not defeated in argument; it was quietly walked around.

FWhat survives of the older verdict is thinner than its confidence once suggested. Guilds plainly did extract rents, and where a craft was easy to supervise the rent was very nearly all they produced; but in crafts where knowledge passed from hand to hand they also produced something that no private contract between two individuals could, which is why parents queued to pay for entry. The proportions differed by trade, by town and by century, so that any single verdict on the institution is bound to be wrong somewhere. Casadei and Correia agree on that much, and on one further point: what guilds were for cannot be settled by reading their rules, since rules record only what an institution says about itself. The contracts, the wages and the later fortunes of those who left are the evidence that says what it did.