CHAPTER 7
Human Capital Investment
All the master dyers of Paris promise by their loyal oaths that in future they will not take on any apprentice for less than 5 years, and they have made this agreement because they are burdened with such a great plenitude of journeymen that often half of them cannot find positions or earn a living.
—Paris dyers’ guild, 1287
Any master who wishes to take on an apprentice must do it for at least 4 years . . . and this is being done so that it will not be so easy to get into the guild.
—Iglau (Jihlava) woollen-weavers’ guild, 1510
When apprentices sign up with their masters they do not attend to learning and when they finish their term they enter the guild as masters themselves without knowing how to make one decent thing.
—Wood-carvers’ guild, Venice, 1625
Masters hold their apprentices only to wood- and water-carrying like common servant-wenches and do not instruct them in the craft . . . on account of which tyranny, many apprentices either run away from sheer fright or go to illegal masters.
—Report to Governor of Upper Austria, 1658
In this workshop he could not learn anything, because there was no journeyman there and the master does not work at all, so that during this entire time he has not even seen a key being made.
—Apprentice’s complaint, locksmiths’ guild, Nuremberg 1789
Guilds are often seen as synonymous with human capital investment. Apprenticeship is a good way of learning skills, and guilds often administered apprenticeship systems. From an economic perspective, any institution that fosters skills invites attention, since modern theories of economic growth assign a key role to human capital. Investing in our human capital should make us work more productively, invent better techniques, have fewer children and invest more in them—and thus make the whole economy grow faster.
Guilds attract even greater interest because school education, the standard human capital indicator, does not show a clear causal link with economic performance before the late nineteenth century. England’s economy did well in the early modern period and industrialized before any other society, yet its schooling and literacy were so mediocre that scholars who differ on other issues concur that education played little role in its economic success.1 Conversely, many European economies with superb educational indicators—the Northern Netherlands, Sweden, German Protestant territories—grew slowly or industrialized late.2 Even at a lower level of aggregation than cross-country comparisons, it has proved difficult to establish causal links between formal education and economic outcomes. This is not surprising, since schools imparted propositional skills and knowledge, often about subjects such as religion and ancient languages with little direct economic utility.3 Such findings have motivated scholars to shift focus to types of human capital that might have been more economically relevant. Some scholars highlight “upper-tail” education that increased the human capital of an elite.4 Others continue to emphasize human capital investment for ordinary people, but in the form of apprenticeships rather than schooling.5
Apprenticeships, it is argued, had several characteristics that made them likely to fuel economic growth. For one thing, they could convey practical skills that were applicable to the real world, indeed to a specific occupation. Second, they could impart “tacit knowledge”: informal know-how picked up through imitation and learning-by-doing. Finally, they could bring a trainee into a wider group of practitioners, which might foster continuous learning even outside formal training sessions. Apprenticeship, consequently, might be the key to economically relevant human capital.6
Guilds are linked to apprenticeships in the popular mind as well as in the arguments of scholars. But in fact, the two were distinct. Surprisingly, as this chapter will show, many European guilds had no involvement with apprenticeship: they did not require it for guild admission, they did not regulate its length, and they did not administer its terms; if guild members undertook apprenticeships, it was an individual choice. Conversely, many apprenticeships, particularly in successful economies such as those of England and the Northern Netherlands, took place outside the guild framework. Such non-guild apprenticeships trained females, paupers, peasant boys, Jews, gypsies, migrants, bastards, members of different religions and ethnicities, and many other potentially productive young people whom guilds excluded. Guilds were thus neither necessary nor sufficient for apprenticeship.
On the other hand, there were many places, times, and occupations in which guilds mandated apprenticeships and closely regulated their terms. This raises the question of what specific contribution guilds made to pre-modern apprenticeship. Did guild rules making apprenticeship an entry condition increase the quantity of human capital in the economy? Or did guild rules excluding girls (and many boys) decrease the number of young people getting vocational training? Did guilds correct market failures in ways that encouraged more people to undertake apprenticeship or increased the quality of training once young people were admitted? Did industries where guild apprenticeship existed out-perform those where it did not?
By investigating these questions, this chapter seeks the answer to a basic question. Guilds, as we have seen, caused economic damage through their entry barriers, market manipulations, and oppression of women. Did they make up for this harm by increasing human capital in ways that benefited the whole economy?
THEORIES ABOUT GUILDS AND APPRENTICESHIP
Guilds and apprenticeship are sometimes regarded as synonymous.7 But a closer look at the evidence gives a different picture, as we shall see. Many guilds did not require apprenticeships, and many apprenticeships took place outside guilds. Guilds were thus neither sufficient nor necessary for people to undertake apprenticeships.
TABLE 7.1: Theoretical Effects of Craft Guilds on Human Capital Investment |
|
Problem Guilds Might Solve |
How Guilds Might Solve It |
People are not willing to invest in appropriate level of human capital because of limited time-horizons or inability to internalize all benefits |
Guilds create monopoly rents which attract individuals into training |
People want to invest in human capital but cannot afford training fees |
Guilds enable apprentices to pay “in kind” by signing up for artificially long period at artificially low wage |
No private contract can protect trainer from trainee opportunism |
Guild mandates length of training, outlaws competition among masters, forbids changing masters |
No private contract can protect trainee from trainer opportunism |
Guild mandates length of training, penalizes bad masters, provides completion certificates |
Problem Guilds Might Create |
How Guilds Might Create It |
Guilds do not know what is the appropriate level of human capital for individual or economy and may not have incentives to find out |
Guild mandates inappropriate level or type of human capital investment |
Guilds, as monopoly providers of occupation-specific training, can increase costs to trainees of human capital investment |
Guild imposes costs (fees, time, regulatory compliance), increasing up-front costs of human capital investment |
Guilds, as monopoly providers of occupation-specific training, can sell exemptions from training |
Guild sells admission to applicants who have not undergone training |
Guilds lack incentive to control opportunism by members |
Guild apprenticeships control trainer opportunism less well than private contracts |
Guilds suffer from internal agency problems |
Guild officers admit untrained persons for personal or monetary benefits |
Guilds, as monopoly providers of occupation-specific training, have incentive to limit entry by limiting access to training |
Guild increases human capital of insiders but denies it to outsiders |
It might nonetheless be argued that guilds were particularly good institutions for investing in economically relevant human capital.8 The theoretical arguments for this idea are laid out in Table 7.1. For one thing, guilds might have overcome market failures that caused people to invest too little in their own skills. If some of the benefits of training were not reaped by trainees but spilled over onto the economy as a whole, individuals would under-invest in training. Guilds might solve this market failure by creating cartel rents that attracted into training people who would otherwise have avoided it. As Ulrich Pfister formulates this conjecture, guild restrictions “specifying the length of tramping and of waiting periods for journeymen, and the exclusion of women, foreigners, and countrymen from apprenticeship, secured a rent on human capital and provided incentives for young males to invest in acquiring skills”.9
A second possibility is that people did want to invest in acquiring skills but could not afford to pay in advance for training, even though it would increase their earnings later. Guilds might have solved this problem—essentially one of missing credit markets—by enabling apprentices to pay for training in kind. A guild could compel an apprentice to sign up for a training period longer than needed to learn the skills, at a wage below the market rate. This way, the apprentice could pay for training in the form of cheap labour.10
A third possibility is that markets could not protect masters from opportunistic behaviour by apprentices, and that deterred trainers from imparting training. An apprentice had an incentive to obtain costly training from a master while still unproductive, and then to take a better job as soon as training made him productive. Guilds could solve this market failure by mandating apprenticeships longer than were needed for learning the skills, prohibiting competition among masters for apprentices, and forbidding apprentices from changing masters in mid-training.11
Finally, markets might not have been able to protect apprentices from opportunistic behaviour by masters, and this could have deterred apprentices from undertaking training at all. Masters had an incentive to exploit apprentices as cheap workers instead of teaching them skills. Guilds might have solved this market failure by imposing rules about length of apprenticeship, penalizing masters who provided poor training, and issuing completion certificates that helped trainees get jobs.
But there are also ways guilds might have harmed human capital. These are shown in the second panel of Table 7.1. First, how does society determine the “right” level of human capital and thus identify when a market failure is preventing people from investing in it? A privileged association of producers, such as a guild, might have incentives to set training levels to benefit its own members, not trainees or the wider economy. This could waste years of young workers’ lives in dependent service when they could be contributing fully to the economy.
A second problem relates to the costs of training. A guild had privileges making its masters the sole legitimate providers of apprenticeships in that occupation. So guild masters could charge a high price for training, since those wanting it had nowhere else to go. Many guilds went further, and fixed a minimum “price” for apprenticeship by forbidding any master from charging a premium (apprenticeship fee) lower than a certain level. As we saw in Chapter 3, the apprenticeship premiums fixed by many guilds were high in terms of contemporary earning power, and applicants viewed them as a serious entry barrier. Guild entry barriers and minimum premium rules increased the cost of investing in training, deterring potential trainees.
Third, a guild could exploit its members’ position as the sole legitimate providers of apprenticeships in an occupation by creating mandatory apprenticeship requirements and then selling exemptions. The guild might profit by certifying untrained entrants in exchange for monetary or other benefits. By falsely certifying untrained people, the guild could reduce the signaling value of training, deterring people from undertaking it (and deceiving consumers).
Fourth, guilds might have the capacity to prevent opportunistic behaviour by trainers, but lack the incentive to do so. A guild master might fail to provide training and exploit his apprentice for cheap labour, but still be assured that the apprentice would not leave because his only exit option was to abscond, losing his completion certificate and thus his legal entitlement to become a guild master. As an association of employers, a guild might find it hard to prevent its own members from treating apprentices in this way. Private, legally enforceable agreements between masters and apprentices might provide better protection for both sides. This would benefit human capital both by ensuring that more training took place inside apprenticeships and by attracting more youths into apprenticeships, because they would know that they need not fear opportunistic abuse from masters.
Fifth, guilds might face internal agency problems. Even if it was in the interest of the guild as a whole that apprentices be well trained, individual guild masters might not bother to provide good training, or guild officers might certify apprentices without testing skills because they were masters’ relatives or offered tempting bribes. Again, as an association of employers, a guild might find it difficult to prevent its members from doing these things. Private, legally enforceable agreements might provide better protection, improve incentives to undertake training, and enhance training outcomes.
Finally, a guild enjoyed exclusive privileges for its members not just to provide apprenticeships in an occupation, but to practise that occupation at all. Existing guild members benefited not just by fixing a high price for training but by limiting access to it. As we saw in Chapters 3 and 5, guilds did limit access to training. They imposed fees that increased its cost, deterring the marginal entrant to training, especially the less well-off. They also denied training to those of the wrong gender, legitimacy status, religion, ethnicity, nationality, or family background. Training slots were further rationed by guild rules limiting apprentice numbers. This did harm not only to those excluded from training, but also to the level of human capital in the economy as a whole, curtailing any growth benefits training might have generated. If a guild excluded more people from training than it encouraged to undertake training, the net effect on human capital would be negative.
In theory, therefore, guilds could have had countervailing effects on markets for training. Their impact on the quantity of economically relevant human capital in pre-modern Europe was thus theoretically ambiguous. What do the empirical findings show?
HOW MUCH SKILL DID PRE-MODERN
CRAFTS AND SERVICES NEED?
Let’s start with a basic question. How much skill did pre-modern crafts and services need? The idea that guild apprenticeships played a major role in economic growth is based on the assumption that learning occupational skills required years of formal training from a certified guild master. But how accurate is this assumption?
Many guilds justified their training requirements, and indeed their entire panoply of privileges, by describing their occupation as highly skilled. The Toledo silk-twisters, for instance, declared in their seventeenth-century ordinance, “The art of silk is useful and has many secrets . . . and because of the quality and secrets of the said art, no matter how deft a man may be, he cannot master them except through long practice and the passage of time”.12 Some modern scholars have taken such statements at face value, arguing that compulsory guild training played a major role in pre-modern economic growth because “pre-industrial products made huge demands on the skills of their producers”.13
On the other hand, guilds themselves often openly acknowledged that they imposed long apprenticeships as entry barriers. In Paris in 1287, for instance, the dyers’ guild imposed a mandatory minimum apprenticeship term so as to deter entry and delay progression to journeymanship and mastership on the grounds that the existing journeymen could not “find positions or earn a living”.14 In the Bohemian town of Iglau (modern Jihlava), the woollen-weavers imposed a minimum apprenticeship term in 1510, “so that it will not be so easy to get into the guild”.15 In Speyer in 1578, the shoemakers advocated a mastership examination to prevent the guild from being “so over-filled”.16
Such evidence leads many scholars to take a more sceptical view of claims that guild training was needed because of objective skill requirements. John Rule, for instance, observes that guilds’ argument that those not trained by the guild lacked skill “was often a pretext for justifying the monopoly of the male guild masters”.17 Merry Wiesner points out how within the guild framework, “women’s work was always judged less ‘skilled’ than men’s, though the actual dexterity and facility required might be more than those in men’s work”.18 Daryl Hafter, too, observes that in early modern Europe,
[w]orkers did not require drastically specialized training to use the tools of that era proficiently. Most techniques required practice rather than abstract schooling to be performed well. Therefore, to give skill the cachet of a special quality that could be monopolized, guilds clung to the idea that craftwork involved “secrets” that could be learned only from guildmasters. In fact, skill was often an artificial label that shed more light on the sex and status of the worker than what was produced.19
As Kirsti Vainio-Korhonen points out for eighteenth-century Finland,
[t]he weaving of linen cloth required the same basic skill . . . whether it was done by a master in his weaver’s shop, by a seaman’s wife in her cottage or on the loom of a burgher family’s housemaid. But in the first case only did it constitute professional handicraft work . . . . The handicraft professions and their related skills were male virtues cultivated by the guilds, and were never attainable by women, shut outside as they were [by] the only legitimated training system.20
According to Michael Sonenscher, too, “the range of materials used in many trades, and the levels of competence required to manipulate them, did not vary greatly”. The world of the guilds, he finds, was one “in which the work that people did was often very similar. The point, however, was to make it appear to be different.”21
Explicit statements by guilds at the time, therefore, as well as scholarly assessments based on such evidence, alert us to the importance of critically examining claims about the amount of formal training needed to practise guilded occupations.22 We can do so by exploring the prevalence of guilds and guild apprenticeships in low-skilled occupations. We can compare how long guilds made apprentices train with independent evidence on how long crafts took to learn. And we can look at what happened when people illegally practised crafts without guild training.
Low-Skilled Yet Guilded
There were certainly occupations that would ordinarily be seen as unskilled but which had guilds in some parts of Europe, though not in others. The qualitative database contains 241 observations of such guilds, shown in Table 7.2. The observations span the seven centuries from 1151 to 1859 and come from twelve different societies. Although guilds are slightly more commonly observed in unskilled trades before c. 1500 than after, they continue to be found in such occupations into the second half of the nineteenth century.23 England, France, and Switzerland are significantly under-represented relative to their share of the overall guilds database, while Spain and the Southern Netherlands are significantly over-represented.24 One interpretation for exploration by future research is that Spanish rulers (by whom the Southern Netherlands were ruled throughout most of the period under analysis) were more willing to grant guild privileges to unskilled trades.
TABLE 7.2: Existence of Guilds in Low-Skilled Occupations, 1151–1859 |
||||||||||||||
|
Labourers |
Farmers |
Vine-growers |
Herdsmen & stableworkers |
Gardeners |
Fishermen |
Land transporters |
Water transporters |
Established retailers |
Petty retailers |
Second-hand dealers |
Misc |
Total |
|
no. |
% |
|||||||||||||
Country: |
||||||||||||||
Austria |
– |
– |
– |
– |
– |
– |
– |
– |
1 |
2 |
– |
2 |
5 |
2.1s |
Bulgaria |
– |
– |
– |
– |
– |
– |
– |
– |
– |
– |
– |
3 |
3 |
1.2s |
England |
– |
– |
– |
– |
1 |
1 |
6 |
– |
2 |
– |
– |
1 |
11 |
4.6ll |
France |
– |
– |
– |
– |
1 |
– |
– |
– |
5 |
7 |
2 |
2 |
17 |
7.1ll |
Germany |
1 |
4 |
9 |
2 |
6 |
7 |
15 |
8 |
4 |
2 |
– |
4 |
62 |
25.7s |
Hungary |
– |
– |
– |
– |
– |
– |
1 |
– |
– |
– |
– |
– |
1 |
0.4s |
Italy |
– |
4 |
1 |
4 |
1 |
1 |
5 |
2 |
3 |
12 |
5 |
3 |
41 |
17.0s |
N. Netherlands |
1 |
– |
– |
– |
1 |
– |
7 |
– |
2 |
– |
6 |
1 |
18 |
7.5s |
Poland |
– |
6 |
– |
– |
– |
– |
– |
– |
– |
– |
– |
– |
6 |
2.5s |
S. Netherlands |
2 |
1 |
– |
– |
1 |
2 |
2 |
– |
4 |
5 |
15 |
2 |
34 |
14.1hh |
Spain |
10 |
1 |
– |
1 |
4 |
4 |
3 |
5 |
7 |
3 |
3 |
1 |
42 |
17.4hh |
Sweden |
– |
– |
– |
– |
– |
– |
1 |
– |
– |
– |
– |
– |
1 |
0.4s |
Period: |
||||||||||||||
Medieval |
6 |
7 |
3 |
1 |
3 |
3 |
12 |
6 |
14 |
14 |
8 |
1 |
78 |
32.4h |
Early modern |
8 |
9 |
7 |
6 |
12 |
12 |
28 |
9 |
14 |
17 |
23 |
18 |
163 |
67.6s |
Total no. |
13 |
11 |
7 |
6 |
14 |
13 |
32 |
11 |
25 |
30 |
30 |
19 |
241 |
100.0 |
Total % |
5.4 |
4.6 |
2.9 |
2.5 |
5.8 |
5.4 |
13.3 |
4.6 |
10.4 |
12.4 |
12.4 |
7.9 |
100.0 |
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Switzerland has zero observations, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database. Bohemia has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 241 observations of guilds existing in unskilled occupations. |
About 27 per cent of the observations in Table 7.2 involve primary-sector occupations such as general labouring (mostly in agriculture), arable farming, animal care, vine-growing, gardening, and fishing. Another 18 per cent comprise guilds in transportation services, both by land (porters, sack-carriers, grain-deliverers, water-bearers, boat-unloaders, cellarmen, draymen, coachmen) and by water (raftsmen, bargemen, boatmen, river-shippers, sailors, longshoremen). About 35 per cent involve guilds of retailers: established traders such as shopkeepers and grocers (10 per cent), petty retailers such as peddlers, hucksters, and fruiterers (12 per cent), and a wide array of second-hand sellers (12 per cent). A final 8 per cent of cases in Table 7.2 involve guilds in miscellaneous low-skilled occupations, including chimney-sweeping, brickmaking, hairdressing, coal-picking, and street-sweeping.
All occupations in Table 7.2 were sufficiently unskilled that they were not guilded in other localities or time-periods. Brickmaking, for instance, required very few skills, as shown by the fact that in colonial Virginia bricks were made by unskilled labourers, children, and slaves.25 In early modern Italy, the Northern Netherlands, and parts of Germany, however, brickmakers formed guilds which survived for long periods.26 In the German territory of Württemberg in the eighteenth century, the brickmakers’ guild required a minimum apprenticeship of two years.27 In other European societies, brickmakers tried to set up guilds but failed: the only documented brickmakers’ guild in England existed for just three years in Westminster (between 1636 and 1639), while the brickmakers of York sought to establish a guild in 1595 but did not succeed.28 In still other European societies, brickmakers did not form guilds but rather, as in Wallonia and Lippe-Detmold, operated as sub-contracting work units, either family groups or all-male work gangs.29 Even in those Dutch and Italian brickmaking industries that were guilded, the guilds made almost no contribution to skills, operating instead as employers’ organizations which “tended to function primarily as production-cartels”.30
The sheer existence of guilds in so many low-skilled activities refutes the rosy view that guilds were formed because they provided efficient mechanisms to solve failures in markets for skilled training. What these guilds actually did casts doubt on yet another optimistic notion, the idea that mandatory guild apprenticeship terms reflected the objective skill requirements of the job. The apprenticeship requirements of all 211 low-skilled guilds in Table 7.2 cannot be discovered from the literature, but Table 7.3 shows the 50 for which this is possible. The sample is a chronologically representative one, in that both the medieval and the early modern period are represented in proportion to their share of observations in the overall guilds database. It is also geographically representative, with almost all European societies represented in proportion to their share of the guilds database. The exception, as in Table 7.2, is the Southern Netherlands, which accounts for 18 per cent of cases, significantly and substantially more than its 6.4 per cent share of the overall guilds database.
TABLE 7.3: Minimum Apprenticeship Terms Imposed by Guilds in Non-Skilled Occupations, 1268–1859 |
|||||||||||||
|
Labourers |
Vine-growers |
Gardeners |
Fishermen |
Land transporters |
Water transporters |
Established retailers |
Petty retailers |
Second-hand dealers |
Misc. |
Total |
Total |
Total |
|
mean |
mean |
mean |
mean |
mean |
mean |
mean |
mean |
mean |
mean |
mean |
no. |
% |
Country: |
|||||||||||||
Austria |
– |
– |
– |
– |
– |
– |
– |
– |
– |
4.5 |
4.5 |
2 |
4.0s |
England |
– |
– |
7.0 |
– |
7.0 |
– |
8.5 |
– |
– |
7.0 |
7.5 |
6 |
12.0s |
France |
– |
– |
4.0 |
– |
– |
– |
3.8 |
5.3 |
3.0 |
3.0 |
4.2 |
12 |
24.0s |
Germany |
– |
2.0 |
n/a |
2.5 |
– |
2.0 |
5.0 |
– |
– |
3.0 |
3.1 |
11 |
22.0s |
Italy |
– |
– |
– |
4.0 |
– |
– |
10.0 |
5.0 |
4.0 |
– |
5.1 |
7 |
14.0s |
N. Netherlands |
– |
– |
2.0 |
– |
– |
– |
– |
– |
– |
– |
2.0 |
1 |
2.0s |
S. Netherlands |
– |
– |
– |
– |
– |
– |
2.0 |
1.0 |
2.0 |
4.0 |
2.2 |
9 |
18.0hh |
Spain |
n/a |
– |
– |
– |
– |
– |
n/a |
– |
– |
– |
n/a |
2 |
4.0s |
Period: |
|||||||||||||
Medieval |
– |
– |
– |
– |
7.0 |
– |
4.0 |
1.0 |
4.0 |
– |
4.0 |
12 |
24.0s |
Early modern |
n/a |
2.0 |
4.3 |
3.0 |
7.0 |
2.0 |
5.4 |
5.2 |
2.5 |
4.1 |
4.4 |
38 |
76.0s |
Total mean |
n/a |
2.0 |
4.3 |
3.0 |
7.0 |
2.0 |
4.8 |
4.6 |
3.4 |
4.1 |
4.3 |
|
|
Total no. |
1 |
1 |
4 |
3 |
2 |
1 |
14 |
7 |
8 |
9 |
|
50 |
|
Total % |
2.0 |
2.0 |
8.0 |
6.0 |
4.0 |
2.0 |
28.0 |
14.0 |
16.0 |
18.4 |
|
|
100.0 |
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. n/a = minimum mandatory apprenticeship term of unknown length (n=9). For all countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 50 observations of the minimum apprenticeship terms imposed by guilds in non-skilled occupations. |
Across the whole sample, these guilds in unskilled occupations imposed mandatory minimum apprenticeships of 4.3 years on average. The average is slightly lower in the Middle Ages (4.0 years) than the early modern period (4.4 years), but the difference is not statistically significant. Across the spectrum of occupations, we observe a mean minimum apprenticeship term of 2 years for vine-growers and water-transporters, 3 to 4 years for fishermen and second-hand dealers, and 4 to 5 years for gardeners, petty retailers, established retailers, and miscellaneous other unskilled occupations; the two observations for land-transporters are both very high (7 years), but both are for medieval England, where municipal legislation or town “custom” often mandated a minimum apprenticeship term of 7 years, a regulation adopted in national legislation in 1563 but thereafter widely violated in practice, regardless of how skilled or unskilled the occupation was.31
That an occupation was unskilled did not deter guilds from adducing skill requirements as a justification for imposing mandatory minimum apprenticeship terms and for enjoying guild privileges at all. In Burgos in 1609, for instance, the labourers’ guild secured its privileges from the authorities on the grounds that it would ensure skilled work.32 In Württemberg in 1717, the chimney-sweeps justified the establishment of their guild and the imposition of a four-year minimum apprenticeship on the grounds that the public weal had been suffering from “the lack of proper standards” and “improper cleaning and observance of the public’s chimneys and hearths”, to remedy which the guild ordered that “from the time of this newly established ordinance onwards, no-one who has not been properly apprenticed to chimney-sweeping, according to this ordinance, shall be allowed to take it up and practise it”.33
Even contemporaries questioned these claims. In early modern Vienna, for instance, it was pointed out that all chimney-sweeps did was clean, which did not involve a high level of skill; nonetheless the Viennese chimney-sweeps’ guild continued to demand five years’ apprenticeship (three for masters’ sons) and maintained its cartel privileges into the nineteenth century.34
A similar query arose in London in 1605 when the Worshipful Company of Gardeners set itself up to defend “the mystery and craft of gardening”, claiming exclusive rights for its members to engage in “gardening, grafting, setting, sowing, cutting, arboring, rocking, mounting, covering, fencing and removal of plants, herbs, seeds, fruits, trees, stocks, sets and contriving the conveyances to the same” within a six-mile radius of the city.35 In 1608 this guild sought to demonstrate its credentials by penalizing one of its members for confusing colewort with cabbage seeds.36 Yet the work carried out by members of this guild did not consist primarily of elaborate horticultural feats, or even in retailing garden products, but rather of manual labour. In 1617, the guild described its members’ lives as being “altogether in the fields and gardens and so desirous of liberty and ayer that they will not be tied to a shopp nor are they capable of any other trade”; it characterized their work as consisting primarily of seeding, weeding, stone-gathering, and collecting urban night-soil.37 In 1635, a municipal commission cast further doubt on the guild’s claim to craft skills, objecting that its members did not “make knotts arbours walks or other like works which properly belong under gardeners”, but rather grew vegetables, ploughed fields, and cultivated grain like ordinary “husbandmen” (i.e., farmers); the commission concluded that the guild’s members “in our opinion are not nor ought to be accompted gardiners for wee find them to be husbandmen by profession”.38 These were hardly skills that would require demanding training of the 400 or more aspiring gardeners who nevertheless had to undergo the guild’s mandatory apprenticeships as a precondition for mastership.
The occupations in Tables 7.2 and 7.3 were not the only low-skilled activities in which guilds existed and imposed mandatory apprenticeship terms. In 1778, for instance, the French government sought to abolish guilds in a large number of occupations it defined as low-skilled. These included candle-making, cord-making, brush-making, whisk-making, pin-making, and container-making—none of them included in Tables 7.2 and 7.3. Despite their occupations being identified by knowledgeable contemporaries as involving little skill, many of these guilds nonetheless submitted petitions insisting that they needed to retain their compulsory apprenticeships, along with exclusive privileges over their occupations.39
LENGTH OF TRAINING—THE CONTROVERSY
But what about mainstream crafts? Surely they were highly skilled and therefore needed formal guild training? Many medieval and early modern guilds declared uncompromisingly that the work their members did involved extremely complex expertise which could only be transmitted through formal apprenticeships in which guilds themselves must specify the minimum term of training, the conditions for admission to training, the fees to be paid, and the examination to be passed. Some modern scholars echo this view, claiming that research in cognitive psychology has demonstrated that learning any kind of skill-based task requires the same minimum amount of time (roughly ten years) and that, because of the importance of “tacit knowledge,” the best institutional framework for learning this skill set is a “community of practice,” such as a guild.40 The widespread existence of mandatory guild apprenticeships, according to this view, was an objective reflection of “the cognitive foundations of human learning”.41
What cognitive foundations are these? Upon examination, they boil down to two propositions. First, learning to do something usually requires some sort of training, although this can be either formal or informal, and it will tend to consist of a combination of propositional and tacit (experiential) knowledge. Second, becoming a “top-level” expert can take quite a long time (from 5,000 to 10,000 hours), although “most professionals reach a stable, average level of performance within a relatively short time frame”.42
Few would dispute that skill requires training and training requires time. But this does not take us further in assessing the best institutional mechanisms for providing it. How much training was needed to practise a pre-modern occupation well enough to satisfy customers, stay in business, and contribute to economic growth? What type of training? What institutions were needed to administer it? Cognitive psychology provides no answers to these questions, even in theory. To address them, we need evidence.
NO STANDARD LENGTH
Many guilds (though by no means all) required any candidate for mastership to undergo a minimum length of apprenticeship with a guild master. Contemporaries sometimes criticized this requirement, claiming that guilds set minimum apprenticeship terms as entry barriers, imposing longer terms than were technically necessary to learn the skills. As we have seen, guilds themselves sometimes acknowledged that limiting entry was the main motive in imposing minimum apprenticeship terms. On the other hand, guilds also often claimed that their “art” consisted of an objective set of “secrets” that could only be mastered “through long practice and the passage of time”.43
One way to explore these opposed views is to look at what guilds did in aggregate. If there was an objective body of skills or “secrets” that needed to be learned for each occupation, and the apprenticeship term imposed by guilds reflected that objective requirement, one would expect guilds in a specific occupation to impose the same minimum apprenticeship term. The quantitative guilds database contains 1,174 observations of minimum apprenticeship terms laid down by guilds in over 270 different occupations, making it possible to find out whether there was in fact an approximate standard length for a given occupation.
How representative are these observations of minimum apprenticeship terms? As Table 7.4 shows, they span the period from 1233 to 1829, and thus cover almost six centuries of guild history, with the medieval and early modern periods both represented in proportion to their share of the wider guilds database. The observations come from 11 different European societies, but some are disproportionately represented. France, Germany, and the Southern Netherlands are over-represented relative to their share of the wider guilds database, while Italy, the Northern Netherlands, Poland, Spain, and a number of smaller societies are under-represented. England is also under-represented, but this is because none of its observations date from after 1563, the year in which English guilds ceased to be responsible for setting apprenticeship terms because the Statute of Artificers and Apprentices imposed a national minimum term of 7 years for both guild and non-guild apprenticeships. The sample is thus strongly representative of guilds in central and northwest Europe apart from England and the Northern Netherlands, the two top-performing economies. It is not strongly representative of the Mediterranean societies, although the sheer size of the sample means that there are quite large absolute numbers of cases for Italy (108 observations) and Spain (37). Despite such geographical clustering, therefore, this sample of well over one thousand observations sheds light on guild training requirements across a wide span of occupations, time-periods, and European societies.
So, did each occupation have a standard length of apprenticeship that might have reflected objective skill requirements? The evidence shows definitively that the answer is “no”. Table 7.5 shows the mandatory apprenticeship terms imposed by guilds in the 25 most commonly observed occupations, those with more than 10 observations in the database. Together, these occupations account for 525 observations of apprenticeship terms, comprising nearly 45 per cent of the total sample. The mean minimum apprenticeship term among these observations is 3.4 years, which is only slightly lower than the corresponding mean of 3.8 years across the sample as a whole.
TABLE 7.4: Guilds Imposing a Minimum Apprenticeship Term, 1233–1829 |
|||||
Country |
Medieval |
Early modern |
Total mean |
Total no. |
Total % |
Austria |
– |
36 |
4.1 |
36 |
3.1s |
Bohemia |
– |
10 |
3.5 |
10 |
0.9l |
England |
14 |
6 |
7.2 |
20 |
1.7ll |
France |
153 |
221 |
4.9 |
374 |
31.9hh |
Germany |
75 |
324 |
3.0 |
399 |
34.0hh |
Italy |
37 |
71 |
4.6 |
108 |
9.2ll |
N. Netherlands |
– |
19 |
2.4 |
19 |
1.6ll |
Poland |
1 |
1 |
2.5 |
2 |
0.2ll |
S. Netherlands |
38 |
104 |
2.5 |
142 |
12.1hh |
Spain |
10 |
27 |
3.7 |
37 |
3.2ll |
Switzerland |
1 |
26 |
3.3 |
27 |
2.3s |
Total mean |
4.6 |
3.5 |
3.8 |
|
|
Total no. |
329 |
845 |
|
1,174 |
100.0 |
Total % |
28.0 |
72.0 |
|
100.0 |
|
Significance |
s |
s |
|
|
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Bulgaria, Denmark, Estonia, Hungary, and Sweden have zero observations in this table, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database overall. For all other countries not represented in the table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. “Mean” = mean length of minimum apprenticeship term. There are no observations for England after 1563, the date the Statute of Artificers and Apprentices imposed a national minimum apprenticeship term of 7 years for both guild- and non-guild apprenticeships. Source: Quantitative guilds database: 1,174 observations of guilds imposing a minimum apprenticeship term. |
As Table 7.5 shows, each occupation displays a wide range of minimum terms. For ordinary crafts it varied widely: for bakers from 0 to 10 years; for butchers from 0 to 6; carpenters from 2 to 6; for tailors and tanners from 0 to 8. Even for more highly skilled occupations, there was no standard length: for apothecaries, it was between 3 and 10 years; for barber-surgeons between 0 and 10; for goldsmiths between 3 and 10; for cloth-shearers between 1 and 7. The same was true of crafts that were potentially export-oriented: for linen-weavers, guilds demanded between 0 and 8 years’ apprenticeship; for silk-weavers between 3 and 10; for wool-weavers between 0 and 7. Of the 25 occupations in Table 7.5, only two—the carpenters and the worsted-weavers—show less than a 5-year gap in apprenticeship terms between the most and least demanding guilds.
TABLE 7.5: Minimum Apprenticeship Terms Imposed by Different Guilds in the Same Occupation, 1233–1829 |
|||||||
Occupation |
Number of observations |
Mean |
Standard deviation |
Coefficient of variation |
Lowest |
Highest |
Gap between lowest and highest |
Apothecary-pharmacists |
12 |
5.17 |
2.04 |
0.39 |
3 |
10 |
7 |
Bakers |
29 |
2.93 |
1.94 |
0.66 |
0 |
10 |
10 |
Barbers / Barber-surgeons |
21 |
2.86 |
2.20 |
0.77 |
0 |
10 |
10 |
Butchers |
17 |
3.12 |
1.65 |
0.53 |
0 |
6 |
6 |
Buttonmakers |
11 |
6.00 |
2.32 |
0.39 |
3 |
10 |
7 |
Cabinetmakers / Joiners |
18 |
3.39 |
2.03 |
0.60 |
2 |
10 |
8 |
Carpenters |
11 |
3.64 |
1.12 |
0.31 |
2 |
6 |
4 |
Coopers |
32 |
2.34 |
1.52 |
0.65 |
0 |
6 |
6 |
Dyers |
20 |
3.30 |
1.17 |
0.36 |
1 |
6 |
5 |
Goldsmiths |
34 |
6.00 |
1.95 |
0.33 |
3 |
10 |
7 |
Hatters |
22 |
4.09 |
1.95 |
0.48 |
0 |
8 |
8 |
Linen-weavers |
21 |
3.24 |
1.48 |
0.46 |
0 |
8 |
8 |
Locksmiths |
11 |
4.55 |
2.11 |
0.47 |
2 |
8 |
6 |
Masons |
28 |
4.04 |
1.90 |
0.47 |
0 |
7 |
7 |
Retailers (general) |
19 |
3.11 |
2.81 |
0.90 |
0 |
10 |
10 |
Retailers (specific items) |
14 |
4.21 |
2.49 |
0.59 |
0 |
6 |
6 |
Saddlers |
13 |
2.69 |
2.69 |
1.00 |
0 |
8 |
8 |
Second-hand dealers |
26 |
1.00 |
1.57 |
1.57 |
0 |
5 |
5 |
Shearers, cloth-finishers |
20 |
3.15 |
1.69 |
0.54 |
1 |
7 |
6 |
Shoemakers |
19 |
2.53 |
1.81 |
0.72 |
0 |
7 |
7 |
Silk-weavers, silk-makers |
26 |
4.88 |
1.77 |
0.36 |
3 |
10 |
7 |
Tailors |
50 |
2.46 |
1.78 |
0.72 |
0 |
8 |
8 |
Tanners |
21 |
3.00 |
2.02 |
0.67 |
0 |
8 |
8 |
Wool-workers, wool-weavers |
17 |
3.24 |
1.44 |
0.44 |
0 |
7 |
7 |
Worsted-weavers |
13 |
3.00 |
0.41 |
0.14 |
2 |
4 |
2 |
Total (these occupations) |
525 |
3.41 |
2.17 |
0.64 |
0 |
10 |
10 |
Total (all occupations) |
1,174 |
3.82 |
2.31 |
0.60 |
0 |
14 |
14 |
Notes: The observations for these 25 occupations come from 12 countries (Austria, Bohemia, England, France, Germany, Italy, the N. Netherlands, Poland, the S. Netherlands, Spain, and Switzerland). Source: Quantitative guilds database: 1,174 observations of minimum apprenticeship terms required by guilds for various occupations. Table reports descriptive statistics for all occupations (n=25) with more than 10 observations, a total of 525 observations (44.7% of the total sample). |
But could it be that although the length of apprenticeship varied, the pre-mastership period of apprenticeship plus journeymanship was in fact standardized?44 Chapter 8 provides a detailed examination of journeymen’s ability to obtain and convey technical knowledge. Here we will focus just on how the journeymanship period contributed to the length of pre-mastership training. For 555 of the observations in Table 7.4 (47 per cent of the total) and 191 of those in Table 7.5 (36 per cent of the total), the mandatory minimum journeymanship term imposed by the guild is also known. This makes it possible to calculate the length of the entire pre-mastership period required by each guild.
TABLE 7.6: Minimum Length of Pre-Mastership Period (Apprenticeship Plus Journeymanship) Imposed by Guilds in the Same Occupation, 1268–1829 |
|||||||
Occupation |
Number of observations |
Mean |
Standard deviation |
Coefficient of variation |
Lowest |
Highest |
Gap between lowest and highest |
Apothecary-pharmacists |
6 |
9.50 |
2.07 |
0.22 |
6 |
12 |
6 |
Bakers |
11 |
6.09 |
2.59 |
0.42 |
2 |
10 |
8 |
Barbers / Barber-surgeons |
8 |
7.38 |
1.77 |
0.24 |
5 |
10 |
5 |
Butchers |
8 |
7.50 |
2.88 |
0.38 |
3 |
11 |
8 |
Buttonmakers |
7 |
8.00 |
1.53 |
0.19 |
6 |
10 |
4 |
Cabinetmakers / Joiners |
4 |
8.50 |
0.58 |
0.07 |
8 |
9 |
1 |
Carpenters |
3 |
6.83 |
3.01 |
0.44 |
4 |
10 |
6 |
Coopers |
6 |
4.83 |
2.93 |
0.61 |
0 |
8 |
8 |
Dyers |
8 |
5.75 |
1.67 |
0.29 |
2 |
7 |
5 |
Goldsmiths |
13 |
9.85 |
3.29 |
0.33 |
6 |
18 |
12 |
Hatters |
8 |
6.75 |
1.98 |
0.29 |
4 |
10 |
6 |
Linen-weavers |
6 |
5.00 |
1.26 |
0.25 |
3 |
6 |
3 |
Locksmiths |
5 |
6.20 |
2.86 |
0.46 |
2 |
10 |
8 |
Masons |
10 |
4.30 |
1.34 |
0.31 |
2 |
6 |
4 |
Retailers (general) |
11 |
4.00 |
3.69 |
0.92 |
0 |
10 |
10 |
Retailers (specific items) |
5 |
4.00 |
3.74 |
0.94 |
0 |
8 |
8 |
Saddlers |
4 |
9.25 |
0.96 |
0.10 |
8 |
10 |
2 |
Second-hand dealers |
8 |
0.75 |
2.12 |
2.83 |
0 |
6 |
6 |
Shearers, cloth-finishers |
9 |
5.44 |
2.99 |
0.55 |
2 |
12 |
10 |
Shoemakers |
8 |
4.75 |
4.27 |
0.90 |
0 |
11 |
11 |
Silk-weavers, silk-makers |
11 |
7.64 |
3.83 |
0.50 |
3 |
15 |
12 |
Tailors |
11 |
4.09 |
2.74 |
0.67 |
0 |
8 |
8 |
Tanners |
4 |
4.50 |
3.11 |
0.69 |
0 |
7 |
7 |
Wool-workers, wool-weavers |
9 |
5.44 |
2.46 |
0.45 |
0 |
8 |
8 |
Worsted-weavers |
8 |
6.25 |
1.81 |
0.29 |
3 |
9 |
6 |
Total (these occupations) |
191 |
6.04 |
3.29 |
0.54 |
0 |
18 |
18 |
Total (all occupations) |
555 |
6.35 |
3.23 |
0.51 |
0 |
18 |
18 |
Note: Table reports descriptive statistics for the 25 occupations in Table 7.4 (n=191) and for the whole sample of 555. Source: Quantitative guilds database: 555 observations of occupations in which mandatory apprenticeship term and journeymanship term are both known. |
Table 7.6 shows this pre-mastership period for the 25 occupations from Table 7.5. It is apparent that the compulsory pre-mastership period was also not of a standard length within a given occupation. Instead, it varied greatly. Again, this was the case for everyday occupations such as the bakers (2-10 years), butchers (3-11 years), shoemakers (0-11 years), tailors (0-8 years), and tanners (0-7 years). It was true of the elite occupations such as the apothecaries (6-12 years), barber-surgeons (5-10), goldsmiths (6-18), and shearers (2-12). And it was the case also for export-oriented textile trades such as the linen-weavers (3-6 years), silk-weavers (3-15), wool-weavers (0-8), and worsted-weavers (3-9). For only five of the 25 occupations in Table 7.6 (button-makers, cabinetmakers, linen-weavers, masons, and saddlers) was the gap in pre-mastership period less than 5 years between the most and least demanding guild. This does not support the view that guild training requirements were determined by the objective technical demands of that occupation.
It might be argued that technological progress was the reason guilds in the same occupation demanded different lengths of apprenticeship. Crafts required longer apprenticeships as the centuries passed, it is claimed, because technological advances increased the complexity of the skills to be mastered and the degree of specialization involved.45
However, the data do not support this idea. If anything, mandatory apprenticeship terms got shorter over the centuries. Table 7.7 compares the minimum apprenticeship terms imposed by guilds in the Middle Ages with the terms required in the early modern period. Across the entire sample of 1,174 observations, the average apprenticeship term was 4.6 years in the medieval period compared to only 3.5 in the early modern period, a 1.1 year difference, which is statistically significant at the 0.05 level. Differences across time within the same occupation show a similar pattern. In 16 of the 25 most frequently observed occupations, the mean mandatory apprenticeship was shorter in the early modern period than the Middle Ages, and in 5 of those the difference is statistically significant. Only 7 occupations show a longer mean mandatory apprenticeship in the early modern period, and in only 2 of those is the difference statistically significant.46
TABLE 7.7: Comparison of Apprenticeship Term in Medieval and Early Modern Period, Imposed by Guilds in the Same Occupation, 1268–1829 |
|||||||
Occupation |
Medieval |
Early modern |
Early modern vs. medieval |
||||
No. obs. |
Mean |
Std. Dev. |
No. obs. |
Mean |
Std. Dev. |
||
Apothecary-pharmacists |
0 |
n/a |
n/a |
12 |
5.17 |
2.04 |
n/a |
Bakers |
7 |
2.71 |
2.06 |
22 |
3.00 |
1.95 |
higher |
Barbers / Barber-surgeons |
6 |
3.50 |
3.39 |
15 |
2.60 |
1.59 |
lower |
Butchers |
2 |
3.00 |
4.24 |
15 |
3.13 |
1.36 |
higher |
Buttonmakers |
3 |
8.00 |
2.00 |
8 |
5.25 |
2.05 |
lower* |
Cabinetmakers / Joiners |
5 |
4.40 |
3.36 |
13 |
3.00 |
1.22 |
lower |
Carpenters |
6 |
3.83 |
0.41 |
5 |
3.40 |
1.67 |
lower |
Coopers |
17 |
2.41 |
1.73 |
15 |
2.27 |
1.28 |
lower |
Dyers |
5 |
3.40 |
1.67 |
15 |
3.27 |
1.03 |
lower |
Goldsmiths |
10 |
7.40 |
1.90 |
24 |
5.42 |
1.69 |
lower** |
Hatters |
12 |
4.83 |
2.12 |
10 |
3.20 |
1.32 |
lower** |
Linen-weavers |
7 |
3.86 |
2.41 |
14 |
2.93 |
0.62 |
lower |
Locksmiths |
3 |
7.00 |
1.00 |
8 |
3.63 |
1.60 |
lower** |
Masons |
14 |
4.57 |
2.21 |
14 |
3.50 |
1.40 |
lower |
Retailers (general) |
6 |
3.33 |
3.01 |
13 |
3.00 |
2.83 |
lower |
Retailers (specific items) |
1 |
6.00 |
n/a |
13 |
4.08 |
2.53 |
lower |
Saddlers |
5 |
2.20 |
3.49 |
8 |
3.00 |
2.27 |
higher |
Second-hand dealers |
5 |
2.80 |
2.17 |
21 |
0.57 |
1.08 |
lower** |
Shearers, cloth-finishers |
5 |
3.00 |
2.35 |
15 |
3.20 |
1.51 |
higher |
Shoemakers |
4 |
0.50 |
1.00 |
15 |
3.07 |
1.58 |
higher** |
Silk-weavers, silk-makers |
8 |
5.63 |
2.45 |
18 |
4.56 |
1.34 |
lower |
Tailors |
13 |
1.08 |
1.26 |
37 |
2.95 |
1.68 |
higher** |
Tanners |
11 |
3.45 |
2.38 |
10 |
2.50 |
1.51 |
lower |
Wool-workers, wool-weavers |
4 |
2.75 |
1.26 |
13 |
3.38 |
1.50 |
higher |
Worsted-weavers |
0 |
n/a |
n/a |
13 |
3.00 |
0.41 |
n/a |
Total (these occupations) |
159 |
3.73 |
2.68 |
366 |
3.27 |
1.90 |
lower** |
Total (all occupations) |
329 |
4.56 |
3.17 |
845 |
3.53 |
1.79 |
lower** |
Notes: ** = difference between medieval and early modern period is significant at 0.05 level. * = difference between medieval and early modern period is significant at 0.10 level. Source: Quantitative guilds database: 1,174 observations of guilds imposing a minimum apprenticeship term. Table reports descriptive statistics for the 25 occupations in Table 7.4 (those with over 10 observations). |
The only two occupations in which the early modern period shows an apprenticeship term that is significantly longer are the shoemakers and tailors, hardly in the vanguard of technological progress. Of the other five occupations in which early modern apprenticeship terms were longer than medieval ones (without being statistically significant), four—bakers, butchers, saddlers, and cloth-shearers—definitely did not see technological advances before the nineteenth century, and it is unclear that the very small and statistically insignificant increase among the wool-weavers can be ascribed to technological progress. By contrast, the occupations which saw a decline in mandatory minimum apprenticeship between the medieval and the early modern period include some in which one might expect technological innovations to have played a greater role, such as the dyers, goldsmiths, hatters, locksmiths, retailers, and silk-weavers. The data thus provide no support for the sanguine view that the wide differences in minimum apprenticeship terms within the same occupation, and the lengthening duration across the centuries, reflected technological advances requiring more complex skills.
The factors that did lead to changing mandatory apprenticeship terms within the same occupation are not always reported in the documentary sources. However, one frequent cause was institutional: pressure from other guilds. In 1573, for instance, the Frankfurt cloth-shearers’ guild found itself constrained to double its minimum apprenticeship period because its apprentices, although “well able to go on the tramp as journeymen”, were being rejected by less liberal shearers’ guilds in other towns.47 In 1591, similar pressures from potters’ guilds in other German towns compelled the Archbishopric of Mainz to increase the minimum apprenticeship for potters from two years to three.48 In 1606, a Worms shearer who had fulfilled the mandated one-year apprenticeship in his home town found himself refused employment and shunned as “dishonourable” in other Rhineland towns because he had not fulfilled the two-year apprenticeship mandated by their shearers’ guilds.49 Less liberal guilds ostracized journeymen and rejected mastership applicants from more liberal guilds, creating an incentive for the latter to tighten up their entry barriers.
On the other end of the spectrum, some guilds did not impose minimum apprenticeship terms at all. Table 7.8 shows observations of 13 guilds which did not impose minimum terms, for which data are available on the length of terms privately agreed between masters and apprentices. The observations come from France, Italy, the Southern Netherlands, and Spain, and are mostly medieval, since the practice of imposing minimum apprenticeship terms proliferated among guilds in the early modern period.
The figures in Table 7.8 suggest that where masters and apprentices were free to negotiate, training periods varied considerably, adjusting to suit individual cases. These variations can be observed among practitioners of everyday crafts such as bakers (where the range lay between a few months and two years), but also in high-quality luxury crafts, such as those of the goldsmiths (between four and seven years) and woollen-weavers (between one and five years). Where a guild did not mandate a minimum length of training, the shortest apprenticeships could be as much as nine years shorter than the longest—and across the entire sample the average difference between the shortest and longest apprenticeship was five years. Thus, in the same place, occupation, time-period, and guild, trainers and trainees did not deem a fixed length of training to be objectively necessary to learn the skills.
TABLE 7.8: Variation in Length of Apprenticeship in Same Guild When Not Legally Fixed |
||||||
Locality |
Country |
Period |
Occupation |
Min |
Max |
Gap |
Paris |
France |
1260–1351 |
strapmakers |
2 |
11 |
9 |
Padua |
Italy |
16th century |
wool-weavers |
1 |
5 |
4 |
Genoa |
Italy |
medieval |
smiths |
6 |
10 |
4 |
Genoa |
Italy |
medieval |
wool-weavers and wool-workers |
2 |
8 |
6 |
Genoa |
Italy |
medieval |
shoemakers |
4 |
5 |
1 |
Tournai |
S. Netherlands |
pre-1424 |
goldsmiths |
4 |
7 |
3 |
Tournai |
S. Netherlands |
pre-1424 |
weavers |
2 |
5 |
3 |
Tournai |
S. Netherlands |
pre-1424 |
cutlers |
2.5 |
6 |
4 |
Tournai |
S. Netherlands |
pre-1424 |
bakers |
0.4 |
2 |
1.6 |
Barcelona |
Spain |
end 15C |
shoemakers |
2 |
9 |
7 |
Majorca |
Spain |
end 15C |
barbers |
3 |
8 |
5 |
Seville |
Spain |
1450–90 |
carpenters and masons |
< 1 |
7 |
> 6 |
Zaragoza |
Spain |
end 15C |
shoemakers |
1 |
12 |
11 |
Total |
|
|
|
< 2.4 |
7.3 |
> 4.9 |
Source: Quantitative guilds database: 13 observations of guilds which did not fix an apprenticeship term. |
This is borne out by the statements and actions of contemporaries. As we shall see later in this chapter, a substantial percentage of apprentices in most pre-modern guilds failed to complete the term of training mandated by the guild. Most apprentices who failed to complete training were still alive and thus left for other reasons. In some cases, this was because training had failed, due mainly to various types of opportunistic behaviour (to be discussed later). In other cases, it was because training had succeeded: the apprentice had been taught the desired skills in less than the term of apprenticeship mandated by the guild.
In medieval and early modern London, for instance, two out of every three tailors’ apprentices failed to complete their terms, many leaving once they had acquired sufficient skill to set up a tailoring business in a setting where a guild mastership was not required, either in a non-guilded village, unincorporated town, or suburb, or illicitly in the “informal sector” of London itself.50 One shoemaker’s apprentice in early modern London stated explicitly that he had left training because he “understood [his] business pretty well”.51 A large proportion of apprentices in early modern Bristol also failed to complete the required period of training. One weaver’s apprentice explained he had left because he could practise his “profession” in the countryside where he did not need to attain the guild mastership status he could only get by completing the guild-mandated term.52 In seventeenth-century Norwich, masters in the successful and dynamic worsted industry were happy to employ illegal “Creepers”, young men who had not completed the compulsory seven years of formal apprenticeship, often preferring such untrained workers to legitimate guild apprentices.53 In eighteenth-century Finland, many of those prosecuted as illegal non-guilded “interlopers” were men who had dropped out at the apprenticeship or journeymanship stage to set up as illegal producers from whom customers were happy to buy.54 In nineteenth-century Vienna, likewise, an official Craft Survey found that many young artisans failed to finish their terms because they were able to earn a living after a much shorter training than the guild required.55
Such testimony from individual apprentices is consistent with contemporary reports on entire occupations. In Freiburg im Breisgau in 1544, the gem-borers’ guild itself acknowledged that the craft could be learned in one year and did not need the mandatory four.56 In Silesia in 1572, rural linen-weavers were competing successfully with urban guild masters after “teaching linen-weaving in a short period of time to their male and female servants”.57 In the Württemberg Black Forest in 1582, peasants and women were competing successfully with guilded worsted-weavers “after learning combing and weaving for only a few weeks or months”, much less time than the three-year apprenticeship demanded by the guild.58 In sixteenth- and seventeenth-century Padua, the woollen industry produced fine items for international markets with producers learning weaving, nap-raising, and dyeing in as little as one year.59 In eighteenth-century Brittany, the building crafts were non-guilded in rural areas, where apprentices and masters drew up contracts in front of notaries rather than guild officials and agreed on a training period of just a year, the period needed to learn the skills rather than the term the urban guilds demanded.60 In the eighteenth-century English worsted industry, according to one expert, “a boy of common capacity would learn weaving, including dressing the warp and fixing it in the loom, in six months”.61 Even the highly skilled activity of cloth-dressing “could be learned in a little over twelve months, [so] there was not the least occasion for seven years’ training”.62
For activities as various as gem-boring, fine-woollen-weaving, dyeing, worsted-weaving, linen-weaving, cloth-dressing, tailoring, and shoemaking, therefore, guilds imposed apprenticeship terms that were far longer than required to learn the skills. Contemporary experts stated this explicitly and—as we shall see—entire industries flourished on the basis of training that was shorter than required by guilds. A non-trivial proportion of apprentices themselves voted with their feet by leaving training before completing their guild-mandated terms. These findings cast doubt on any optimistic view that the mandatory duration of guild apprenticeships primarily reflected the objective difficulty and complexity of the skills to be learned.
MINIMUM APPRENTICESHIP TERMS—A BOND ON PERFORMANCE?
People at the time knew guilds imposed unnecessarily long apprenticeship terms. Adam Smith’s unsparing assessment was that guilds imposed minimum apprenticeship terms “to restrain the competition to a much smaller number than might otherwise be disposed to enter into the trade . . . by increasing the expense of education”.63 Some modern scholars, on the other hand, believe that guilds fixed long training periods as an efficient solution to market failures arising from apprentice opportunism. An apprentice consumed more than he produced for some initial segment of his training period, before he learned anything. This created the risk that once he learned something he would depart, leaving the master out of pocket. This deterred masters from taking on apprentices. Guilds therefore imposed a longer apprenticeship period than strictly needed to learn the skills as a bond to prevent apprentices from departing as soon as they became useful employees. This removed the risk that would otherwise have deterred masters from taking on apprentices, with the result that training markets worked better and human capital investment increased.64 The reasoning behind this argument makes sense. But how much factual content does it have?
One relevant fact is how old apprentices were when they started training. A very young apprentice was more likely to spend a substantial period generating more costs than benefits for his master. An older one was more likely to work productively from the beginning, if only at unskilled or non-vocational tasks, and less likely to generate net costs that his master needed to recoup by imposing a longer apprenticeship term than needed to learn the skills of the occupation.65 So how old were guild apprentices on average?
In many guilds, it turns out, apprentices were above the age at which their productivity covered their consumption costs. In pre-modern Europe, a young person began to cover his consumption costs between the ages of 11 and 14.66 As we saw in Chapter 3, the qualitative database contains a number of observations of guilds imposing minimum age conditions for entering apprentices. In more than two-thirds of cases, apprentices were required to be at least 14 years old, and in about one-tenth of cases they had to be at least 18.
What happened in practice? The quantitative database contains 27 observations of the average age at which a sample of apprentices entered training. As Table 7.9 shows, these observations are drawn from nine different European societies and span the six centuries from 1250 to 1857. Most derive from samples of mixed occupations, but there are also samples consisting solely of weavers, silk-workers, grocers, carpenters, construction craftsmen, and maritime occupations. Across the 22 samples of exclusively male apprentices, the lowest average age of starting apprenticeship was 13 years and the highest was 20, with an unweighted average of 16.4.67 The observations include only two samples of exclusively female apprentices, unsurprisingly since as we saw in Chapter 5 guild apprenticeship for females was rare. Both show an average age of starting apprenticeship of 13 to 14 years. In the three samples combining male and female apprentices, there is a wide range of average ages at entering apprenticeship (from 14 to 24 years), with the unweighted average of the three samples lying at 18.3 years. The unweighted average across all 23 observations in Table 7.9 is 16.4 years. These data suggest that the typical age at which young people started guild apprenticeship was well over the “break-even” age of between 11 and 14 years, by which time they were productive enough to cover their consumption costs.
TABLE 7.9: Observed Average Age at Starting Apprenticeship, 1250–1857 |
||||||
Locality |
Country |
Period |
Occupation |
Boys |
Girls |
Mixed |
Austrian towns |
Austria |
early modern |
construction trades |
18.5 |
|
|
Austrian towns |
Austria |
19th c. |
silk-weavers |
14.0 |
|
|
Vienna |
Austria |
18th c. |
various |
13.0 |
|
|
Bristol |
England |
early modern |
various |
17.0 |
|
|
London |
England |
c. 1556 |
carpenters |
18.5 |
|
|
English towns |
England |
16th c. |
maritime trades |
14.5 |
|
|
London |
England |
1556 |
all |
18.5 |
|
|
London |
England |
early 14th c. |
various |
14.0 |
|
|
London |
England |
1400 |
various |
15.5 |
|
|
London |
England |
1500 |
various |
18.0 |
|
|
London |
England |
1572–94 |
carpenters |
19.5 |
|
|
Lyon |
France |
18th c. |
silk-workers (born in Lyon) |
15.0 |
|
|
Lyon |
France |
18th c. |
silk-workers (born outside Lyon) |
17.0 |
|
|
Paris |
France |
16th c. |
grocers |
16.0 |
|
|
Toulouse |
France |
1770–90 |
goldsmiths |
14.5 |
|
|
Toulouse |
France |
1770–90 |
shoemakers & cobblers |
16.5 |
|
|
German towns |
Germany |
early modern |
construction trades |
18.5 |
|
|
German towns |
Germany |
early modern |
various |
14.0 |
|
|
Venice |
Italy |
late 16th c. |
various |
14.0 |
|
|
Venice |
Italy |
early 17th c. |
various |
16.5 |
|
|
Spanish towns |
Spain |
1250–1600 |
various |
20.0 |
|
|
Malmö |
Sweden |
1811–25 |
various |
17.0 |
|
|
Austrian towns |
Austria |
19th c. |
silk-weavers |
|
14.0 |
|
Venice |
Italy |
late 16th c. |
various |
|
13.0 |
|
Vienna |
Austria |
1789–1857 |
silk-weavers |
|
|
14.0 |
Bedfordshire |
England |
1500–1800 |
various |
|
|
17.0 |
Ghent |
S. Netherlands |
15th c. |
various |
|
|
24.0 |
Total |
|
|
|
16.4 |
13.5 |
18.3 |
Note: A number of sources only report an age-range, which is averaged in the table. Thus expressions of the form “14 to 16” are standardized to “15”. Source: Quantitative guilds database: 27 observations of the average (or “typical”) age at starting apprenticeship. |
Even for guilds where the precise ages of apprentices beginning their training are not known, indirect evidence suggests they were often mature, productive, or both. The qualitative database contains 35 observations of such evidence. These come from occupations as various as those of barber-surgeons, bargemen, carpenters, drapers, dyers, masons, mirror-makers, printers, roofers, scythe-smiths, silk-spinners, silk-weavers, spear-burnishers, tailors, tilers, and turners, as well as from samples of mixed occupations. As Table 7.10 shows, they are drawn from studies of guilds in seven different European societies and span more than five and a half centuries, from 1180 to c. 1750.
One type of evidence, comprising 9 per cent of the cases in Table 7.10, is based on accounts in which apprentices were described as being of advanced age. In Venice in the 1620s, for instance, masters in a number of guilds were said to be taking on apprentices “who are full-grown men, older, in fact than the masters”.68 Another category of case (11 per cent of the total) involves apprentices who had arranged their own contracts, a reliable sign that they had reached the age of legal majority: in Arles between 1350 and 1450, for instance, 46 per cent of apprentices arranged their contracts personally.69 Then there are cases (14 per cent of the total) where apprentices had already held jobs, as in early modern Hildesheim where the roofers’ guild only accepted into apprenticeship boys who had already worked as limestone-mixers or helpers.70 Sometimes (in 3 per cent of cases), training contracts acknowledged the productivity of apprentices by including cash penalties for absences, as in 1304 when a Tournai draper required an apprentice to pay 60s. or provide “a journeyman as adequate” if he went on holiday for more than a month without permission.71 In 9 per cent of observations, apprentices were productive enough to be entrusted with important jobs, as in 1444 when a London tailor paid the guild the equivalent of 50 days’ master’s earnings for a license permitting his apprentice to “keep shop for his master’s profit within his term [of training]”.72 In some cases, guilds themselves treated apprentices as productive workers by including them in workforce ceilings, as in 1533 when the Landau tailors’ guild allowed a master to employ either three journeymen, or two journeymen and one apprentice, or one journeyman and two apprentices.73
TABLE 7.10: Evidence that Apprentices are Mature, Productive, or Both, 1180–1750 |
|||||||||
|
Is described as being of advanced age |
Arranges contract himself, hence over age of majority |
Has prior work experience |
Is paid a wage from beginning |
Is missed for his work when he is absent |
Is entrusted with important work |
Is counted by guild towards maximum number of employees |
Total |
|
no. |
% |
||||||||
Country: |
|||||||||
Austria |
– |
– |
1 |
3 |
– |
– |
– |
4 |
11.4hh |
England |
– |
1 |
2 |
1 |
– |
1 |
– |
5 |
14.3s |
France |
– |
2 |
– |
2 |
– |
– |
– |
4 |
11.4s |
Germany |
– |
– |
2 |
4 |
– |
1 |
1 |
8 |
22.9s |
Italy |
3 |
1 |
– |
6 |
– |
– |
– |
10 |
28.6hh |
N. Netherlands |
– |
– |
– |
1 |
– |
– |
– |
1 |
2.9s |
S. Netherlands |
– |
– |
– |
1 |
1 |
1 |
– |
3 |
8.6s |
Period: |
|||||||||
Medieval |
– |
4 |
– |
8 |
1 |
2 |
– |
15 |
42.9hh |
Early modern |
3 |
– |
5 |
10 |
– |
1 |
1 |
20 |
57.1ll |
Total no. |
3 |
4 |
5 |
18 |
1 |
3 |
1 |
35 |
100.0 |
Total % |
8.6 |
11.4 |
14.3 |
51.4 |
2.9 |
8.6 |
2.9 |
100.0 |
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. For all countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 35 observations of evidence that apprentices are mature, productive, or both. |
Arguably the most important evidence that apprentices were productive, comprising 51 per cent of the cases in Table 7.10, is the fact that they were often paid wages. In one sample of 169 medieval Genoese apprenticeship contracts, 29 per cent stipulated that the master should pay a wage.74 In medieval Orléans, wages were promised in 53 per cent of apprenticeship contracts between 1380 and 1440, and in 93 per cent between 1440 and 1490.75 In early modern London, tilers’ apprentices earned a daily wage far in excess of what it would have cost to feed them.76 In sixteenth- and seventeenth-century Venice, a study of 3,687 apprenticeship contracts found that 80 per cent involved paying wages, from which the authors conclude that the principal purpose of taking on apprentices was to get cheap labour, not to provide training.77 In Germany and Austria, according to Reith, payment of wages to apprentices was already common in the late Middle Ages, and by the early modern period had become pervasive in the textile, metal-working, and construction sectors, which together employed a majority of the industrial workforce.78 A study of more than 800 apprentices in eighteenth-century Leiden and Utrecht found that wages increased annually, implying that working and training occurred in tandem and neither masters nor apprentices lost much if an apprenticeship ended prematurely.79 Taken together, these quantitative and qualitative findings cast doubt on the idea that unnecessarily long apprenticeships were an efficient guild device for protecting masters against early departures by apprentices.
The notion that guilds had to mandate excessively long training as a bond on apprentice performance relies also on a second assumption: that other institutions for guaranteeing apprenticeship contracts were lacking. But this is an inaccurate picture of pre-modern European economies. Both money and the legal system existed in medieval and early modern Europe, and people used both to guarantee fulfilment of contracts—including for guild apprenticeships.
The qualitative database contains 33 observations of guild apprenticeships in pre-modern Europe in which masters and apprentices used cash penalties and legal remedies as bonds against opportunism. As Table 7.11 shows, these observations come from 7 European societies and span some 600 years from the twelfth to the eighteenth century. It might be thought that alternative institutional solutions would be lacking in the medieval period, but Table 7.11 does not support this view: the medieval period accounts for fully one third of observations, not significantly different from its share of the overall guilds database. Even in their medieval beginnings, therefore, alternative institutional solutions meant guilds did not need to impose excessively long apprenticeships.
TABLE 7.11: Non-Guild Institutional Mechanisms for Enforcing Apprenticeship Contracts, Twelfth to Eighteenth Centuries |
|||||||||
|
Master pays higher wages as training progresses |
Apprentice pays deposit to master, if misbehaves |
Apprentice makes loan to master, forfeited if misbehaves |
Master promises payment to apprentice, conditional on finishing training |
Older/more productive apprentice pays lower premium |
Apprentice provides personal guarantors |
Contract signed before notary or in municipal office, promising good behaviour |
Total |
|
no. |
% |
||||||||
Country: |
|||||||||
Austria |
– |
1 |
– |
– |
– |
1 |
2 |
4 |
12.1hh |
Bohemia |
– |
1 |
– |
– |
– |
– |
– |
1 |
3.0s |
Germany |
2 |
3 |
– |
– |
1 |
2 |
2 |
10 |
30.3s |
Italy |
8 |
1 |
– |
1 |
– |
1 |
2 |
13 |
39.4hh |
N. Netherlands |
1 |
– |
– |
– |
– |
– |
– |
1 |
3.0s |
S. Netherlands |
1 |
1 |
1 |
– |
– |
– |
– |
3 |
9.1s |
Spain |
– |
1 |
– |
– |
– |
– |
– |
1 |
3.0s |
Period: |
|||||||||
Medieval |
3 |
2 |
1 |
1 |
– |
1 |
3 |
11 |
33.3s |
Early modern |
9 |
6 |
– |
– |
1 |
3 |
3 |
22 |
66.7s |
Total no. |
12 |
8 |
1 |
1 |
1 |
4 |
6 |
33 |
100.0 |
Total % |
36.4 |
24.2 |
3.0 |
3.0 |
3.0 |
12.1 |
18.2 |
100.0 |
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. France has zero observations in this table, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database overall. England has zero observations in this table, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database overall. For all other countries not represented in the table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 33 observations of non-guild institutional mechanisms used as a bond on performance in guild apprenticeships. |
One contractual solution, comprising 36 per cent of observations in Table 7.11, was for the master to promise gradually rising wages across the term of training. In twelfth- and thirteenth-century Genoa, as we have seen, about 30 per cent of apprenticeship contracts involved payment, and in many cases a careful gradation of wages as the apprentice’s skills increased.80 A second arrangement (24 per cent of cases) required the apprentice to deposit a fee with the master, to be repaid at the end of training if the apprentice refrained from opportunism; again, this can be observed in some of the earliest surviving medieval apprenticeship contracts.81 Another variant (3 per cent of cases) had the apprentice’s family make a loan to the master, which did not have to be repaid if the apprentice misbehaved; this device was also used from a very early date.82 A different solution, also comprising 3 per cent of cases, saw the master promising the apprentice a cash payment, to be given at the end of the agreed training period conditional on good behaviour.83 Yet another arrangement involved varying the premium, with younger apprentices paying a higher premium to cover any period during which their costs outweighed their utility.84
In another 12 per cent of cases, personal guarantors stood surety for the apprentice’s behaviour. In medieval German towns, for instance, apprentices often had to nominate two persons who pledged that the apprentice would complete the agreed term, not abscond before the end of training, pay all fees to master and guild, conduct himself well, and observe guild regulations.85 In early modern Lyon, masters sued guarantors for compensation and damages when apprentices departed without finishing their term.86
A final institutional solution (18 per cent of cases) was for trainer and trainee to go to a notary or municipal office and sign a detailed contract promising to refrain from opportunistic behaviour on pain of legal penalty. As early as 1271, a notarized contract for a tailor’s apprentice in the Venetian colony on Crete contained detailed penalty clauses to deal with opportunism by either party.87 In medieval and early modern Germany and Austria, as Reith points out, even when there was no notarized agreement, “authorities beyond the guilds had a strong interest in the quality of training and disputes were handled by the business courts in the large manufacturing towns”.88
These institutional mechanisms existed from the twelfth or thirteenth century onwards, and they were used in the context of both guild and non-guild apprenticeships. This casts doubt on the idea that excessively long guild apprenticeships were the efficient, or even the preferred, solution to apprentice opportunism. Money bonds, personal pledges, and contractual devices against opportunism can all be found in the earliest apprenticeship agreements, so there was never a period when alternative institutional mechanisms were lacking.
Craft Practice by Females Who Were Denied Guild Training
A third way of assessing lengthy guild apprenticeships is to look at what happened when people tried to practise trades without guild training. As we saw in Chapter 5, females were largely excluded from guild apprenticeship, except in the tiny minority of guilds—less than half of one per cent—that were all-female or mixed-sex. Females’ lack of formal guild training did not, however, stop them from practising skilled trades, both legally and—more often—illegally.
Masters’ daughters became skilled at crafts even though almost all guilds denied them apprenticeship. In one Augsburg tailor’s workshop in 1724, for instance, “the journeymen just did the repairs, while the daughters carried out the fine work”.89 That same year, the orphaned daughter of a master of the Augsburg painters’ guild recounted how she had learned the occupation from her father, and produced “several proofs of her work consisting of painted coffee-cups”.90 The daughter of an eighteenth-century Augsburg fountain-master described how she was “daughter, journeyman, apprentice, and handyman to her father in his art in everything that came to hand”, including helping him “make and lay all the artifice- and water-pipes”.91 In the 1750s, a conflict arose in the Lyon silk guild over whether local girls should be allowed to undergo formal apprenticeship: the employers were enthusiastic, arguing that females “have a recognized delicacy above that of the men for all the light stuffs . . . and they are no less clean at making the richer stuffs”.92
Maidservants, female relatives, and independent women were also regarded as capable of learning craft skills, to such an extent that guild-trained males barred them as dangerous competitors. In 1720, for instance, the Augsburg bookbinders’ guild justified prohibiting the work of maidservants and female relatives for fear that they would leave service, set up in business on their own accounts, “and in so doing cause no little loss and damage to our profession”.93 In England between 1726 and 1756, the middle-aged spinster Anna Maria Garthwaite became the pre-eminent silk designer in the country, without having pursued any apprenticeship to the trade, from which she would in any case have been excluded by the London Weavers’ Company since she was not related to a male guild master.94
Masters’ wives were also regarded as having craft skills equivalent to those obtained by males with apprentice training. One Augsburg bookbinder petitioned in 1750 to be allowed to employ an extra journeyman because his wife had broken her arm and would not be able to give him “her assistance which, not to exaggerate, I value as absolutely equal to that of a journeyman”.95 Another master described how “my wife has rendered me a very helpful hand in many matters, and especially has spared me much loss of time by taking care of the large amount of correspondence involved in my business trading into foreign places”.96 In 1778, an Augsburg baker explained that the reason his workshop had produced an underweight loaf of bread was that “my wife was ill and I had to leave the baking to the journeyman, who did not pay sufficiently accurate attention and therefore baked below weight”.97
Masters’ widows were also tacitly viewed as possessing complete mastery of the skills of their husbands’ occupations for, as we saw in Chapter 5, the vast majority of European guilds granted them rights to continue operating the workshop. Except in those cases where the guild required her to keep a journeyman, the widow conducted the business by herself. Although almost all guilds forbade a master’s widow to take on new apprentices, parents were willing to send their boys to train with a widow illegally, as in Paris in 1751 when the parents of a 16-year-old boy apprenticed him for five years to a sculptor’s widow in return for the promise that she would “show him the art of sculpture without hiding anything from him”.98 Even in strictly guilded places such as eighteenth-century Augsburg, guild courts accepted the certificates of skill which widows issued for their journeymen.99
Skilled work in many guilded trades was thus successfully carried out by females who were excluded from guild apprenticeship. In the case of wives and widows, it was not because women had been informally “apprenticed” to their husbands. We saw in Chapter 5 that guilds imposed many conditions on widows’ rights, but length of marriage was not one of them. In eighteenth-century Paris, out of 121 guilds all but two permitted a widow automatically to continue the workshop, and none required her to have been married (and thus “apprenticed” to her husband) for a minimum period.100 In the Württemberg district of Wildberg between 1598 and 1760, 10 to 15 per cent of weavers’ workshops were operated by masters’ widows; 18.5 per cent of these widows had been married for less than the six years of apprenticeship plus journeymanship required by the guild, and 9.4 per cent for less than the three years of mandatory apprenticeship. The widows must have done the work themselves, since none kept apprentices or journeymen.101 Such women, their customers, and the guild all clearly believed that craft skills could be learned in less than the minimum apprenticeship term.
DID APPRENTICESHIP NEED GUILDS?
There were, then, many unskilled occupations that did not need any of the training guilds claimed for them, and many somewhat skilled occupations that did not need the length of training guilds claimed for them. But some occupations needed at least some formal training. In theory, apprenticeship has many advantages: it is oriented to the job at hand; it is taught through demonstration more than exposition; and it brings the trainee into a wider group of practitioners which fosters continuous, tacit learning even outside formal training sessions.
So what institutions are needed for apprenticeship? Some hold the view that a guild was the only possible institution for making apprenticeship work well, because only it could address the market failures listed in Table 7.1: externalities, lack of credit, and opportunism by trainers and trainees.102
The historical facts, however, tell a different tale. Apprenticeships existed widely in pre-modern Europe without guilds.103 They existed in guilded occupations without guild involvement, and they existed in occupations that did not have guilds. In some European societies, apprenticeships were more common outside guilds than within them. This was true in England, for instance, where by the seventeenth century, guilds were inactive in enforcing contracts even in guild apprenticeships, yet apprenticeship flourished, and continued to do so long after the guilds began to decline and even after they had disappeared.104 The widespread existence of apprenticeships that were not mandated, regulated, or enforced by guilds casts doubt on the idea that guilds were necessary to solve failures in training markets.
Societies where guilds were rare nonetheless had apprenticeships. As David Nicholas points out, in many parts of medieval Europe the earliest apprenticeship contracts “precede the formal grant of corporate privileges to the guilds and could exist outside them in trades that had no formal guild structure”.105 In many towns in medieval Provence, guilds were either absent or weak, yet apprenticeships were widespread, as shown by the numerous notarized apprenticeships recorded for Manosque between 1369 and 1480.106 Guilds were rare in early modern Russia, but this did not prevent people from entering into apprenticeship contracts.107 From the seventeenth century onwards, the Yaroslavl village of Pavlovo, which specialized in high-quality small iron wares, did not have guilds but instead supported a system of apprenticeship based on private contracts.108
Societies with fully developed guild systems also had apprenticeships that were organized without guild involvement, and this occurred in both guilded and non-guilded occupations. The qualitative guilds database contains 28 observations of guilds, towns, or entire European societies that had guilds, but also had non-guild apprenticeships. These are shown in Table 7.12. The observations come from occupations as various as those of building craftsmen, dyers, millwrights, potters, seamstresses, retailers, second-hand dealers, shipbuilders, silk-weavers, woollen-weavers, and worsted-weavers, as well as from 11 studies analyzing multi-occupation samples. Non-guild apprenticeships existed across the entire period of European guild development, as shown by the fact that the observations in Table 7.12 span the 600 years from the thirteenth to the nineteenth century, with the medieval and early modern periods represented in proportion to their share of the overall database. The observations come from eight European societies, but it should be noted that England and the Northern Netherlands are significantly over-represented, while Germany is significantly under-represented.
TABLE 7.12: Non-Guild Apprenticeships in Places That Had Guilds, Thirteenth to Nineteenth Century |
||||
Country |
Medieval |
Early modern |
Whole period |
|
no. |
% |
|||
England |
2 |
5 |
7 |
25.0hh |
Finland |
– |
1 |
1 |
3.6hh |
France |
– |
5 |
5 |
17.9s |
Germany |
– |
2 |
2 |
7.1ll |
Italy |
1 |
4 |
5 |
17.9s |
N. Netherlands |
– |
5 |
5 |
17.9hh |
S. Netherlands |
1 |
1 |
2 |
7.1s |
Spain |
– |
1 |
1 |
3.6s |
Total no. |
4 |
24 |
28 |
100.0 |
Total % |
14.3 |
85.7 |
100.0 |
|
Significance |
s |
s |
|
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. For all countries not represented in the table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Quantitative guilds database: 28 observations of guilds, towns, or entire countries in which there were both guilds and non-guild apprenticeships. |
Apprenticeships thus did not need guilds. In practice, there were multiple institutional mechanisms through which apprenticeships could be and were negotiated, enforced, and certified. England provides a vivid example of these variegated mechanisms at work. Guilds were largely limited to London and the old borough towns, but apprenticeships were found almost everywhere: in suburbs, guild-free “new towns”, villages, and the countryside.109 In 1563, national legislation—the Statute of Artificers—created a legal framework for private contracts of apprenticeship in any occupation. This made apprenticeships unusually widespread in England. Individuals, families, and parish welfare authorities arranged apprenticeships, and these contracts were monitored and enforced by the ordinary court system and Justices of the Peace (public magistrates) rather than guilds.110 As a consequence, apprenticeships pervaded the entire spectrum of English occupations, included many non-guilded activities, and were open to females to an uncommon degree.
In the early modern Northern Netherlands, non-guild apprenticeships also existed on a large scale. Urban and provincial governments, welfare institutions such as orphanages, and municipal regulatory systems such as the Leiden “neringen” provided a “generalized” institutional framework for the registering and enforcing of apprenticeship contracts: “If one of the parties defaulted, the aggrieved person could always have recourse to the courts”.111 This encouraged the proliferation of apprenticeships outside the guild system. In a sample of Amsterdam apprenticeship contracts for the years between 1595 and 1670, for instance, only 39 per cent were in occupations that had guilds.112
England and the Northern Netherlands may have provided a particularly effective array of institutions for organizing apprenticeships outside the guild system, but similar mechanisms existed in many parts of Europe. In thirteenth- and fourteenth-century Flanders, privately organized apprenticeships were widespread, supported by contractual forms and notarial offices; it was only around 1400 that apprenticeships began to move in greater numbers into the guild system, mainly because guilds started to reinforce their entry barriers by refusing to recognize private apprenticeships that were not also registered with the guild.113 In early modern France, whether or not an occupation was guilded, the term “apprenticeship” was used, and the contract was formalized in front of a notary.114 In the eighteenth century, the French crown also sometimes played a direct role by inviting innovators to instruct apprentices and then granting the trainees mastership, in circumvention of the guilds.115 In medieval and early modern Italy, apprenticeships were increasingly agreed upon through private contracts between masters and trainees, drawn up privately or in front of notaries, and then enforced in civil courts: “Written contracts regulated the apprenticeship and were the prevalent tool for reducing transaction costs in the relationship between the master and the apprentice.”116
Even guild apprenticeships made use of municipal, legal, and state institutions. In Ypres in 1281, for instance, the city government required all guild apprenticeship agreements to be drawn up in writing in front of two city aldermen.117 From the thirteenth century onwards, Venice required apprenticeship contracts to be registered in the Giustizia Vecchia, an office of the state, and both masters and trainees were expected to use the public law-courts to resolve any dispute.118 In Bologna in 1288, the statutes of the cotton guild explicitly instructed masters to record apprenticeship agreements in front of notaries.119 In medieval and early modern London, apprenticeship contracts were monitored and enforced by the mayor and aldermen, with apprentices recorded and registered before officials at a public meeting; in the seventeenth and early eighteenth centuries, the Lord Mayor’s Court heard cases involving an estimated 3 to 8 per cent of all guild apprentices.120
Apprenticeships thus existed widely without guilds. Contracts were agreed privately between trainees and trainers, and were recorded, monitored, and enforced by a wide array of institutions, including notaries, municipal offices, charities, public poor relief systems, public magistrates, public law-courts, and even government regulatory bodies. Apprenticeship agreements did not need guild compulsion, monitoring, or enforcement, and guilds themselves often demanded that contracts be registered by outside institutions.
DID ALL GUILDS REQUIRE APPRENTICESHIP?
Guilds and apprenticeships are popularly conflated because of the idea that all guilds required them. Some scholars refer to “the typical stipulations made by guild statutes, particularly with respect to the provision of training”.121 Others confidently declare that “with virtually all guilds, education of new tradesmen was one of the principal functions of the organisations”.122 But we now know this is not correct.
Guilds certainly had strong incentives—apart from concern for the common weal—to make rules about training. One was practical: training requirements made it easier to monitor and limit entry. The other was rhetorical: skill was indisputably a good thing, and could be used to justify other, less defensible things such as entry barriers and gender discrimination.123
This makes it the more striking that not all guilds required apprenticeship. In medieval Flanders, for instance, guilds did not universally mandate apprenticeship and only gradually introduced it “as a means of limiting the access of outsiders”.124 The same was true in medieval Spain. When the Murcia carpenters’ guild held its annual assembly in 1394, 6 of the 12 men present had only ever worked as labourers, and only 3 had worked as master carpenters in 1392.125 In the early modern Dutch Republic, many guilds, including those of bargemen, boatmen, brokers, carters, fishmongers, merchants, peddlers, and porters, did not even distinguish a separate category of apprentices.126 The same was true in medieval and early modern England, where in 1348 the London pewterers’ guild admitted anyone who was “otherwise true workman known and tried among them”, and in 1561 the Bristol tailors’ guild accepted “for a reasonable fine any honest person being a good workman, although he hath not been an apprentice to the same craft”.127 By c. 1500, as we saw in Chapter 3, the custom of London allowed any citizen to enter a guild through purchase or patrimony rather than apprenticeship, and to practise any occupation, not just the one to which he had been apprenticed, without being a member of the relevant guild.128 There were also major Italian manufacturing centres where guilds did not mandate apprenticeships. In early modern Padua, for instance, important crafts, including those in the export-oriented wool and silk industries—weavers, dyers, knitters, and ribbon-makers—had no regulations requiring apprenticeships, or only introduced them at a relatively late date.129
These guilds were not exceptional. The quantitative database contains 23 observations of the prevalence of apprenticeship prescriptions in samples of European guilds. As Table 7.13 shows, the samples come from five societies, including ones representing northern, central, and southern Europe. They date from a variety of time-periods between 1220 and 1800, and thus span nearly six centuries. In aggregate, they cover a total of 2,291 guilds. They measure prevalence in multiple ways: the share of guilds with apprenticeship prescriptions, the share with minimum training terms, the share requiring proof of apprenticeship, and the share of guild members in guilds that mandated compulsory apprenticeship.
Not all guilds had any apprenticeship prescriptions at all, as shown by 12 samples in Table 7.13, which together comprise a total of 1,766 guilds. The sample spanned a wide range, from zero guilds with apprenticeship prescriptions in medieval Murcia to 100 per cent of guilds in seventeenth-century Middle Rhine towns. Some of the largest samples show strikingly low percentages: among over eleven hundred medieval and early modern Italian guilds, just 40 per cent had apprenticeship prescriptions; among over three hundred medieval and early modern Spanish guilds, just 23 per cent did so. Across the total of 1,766 guilds for which this measure is available, the weighted average is less than 44 per cent.
TABLE 7.13: Prevalence of Apprenticeship Prescriptions in European Guilds, 1220–1800 |
|||||||
Locality |
Country |
Period |
No. guilds |
% with apprenticeship prescriptions |
% with minimum apprenticeship terms |
% requiring proof of apprenticeship |
% members in guilds with apprenticeship prescriptions |
Paris |
France |
1268 |
102 |
81.4 |
– |
– |
– |
Toulouse |
France |
1340s |
19 |
10.0 |
– |
– |
– |
Paris |
France |
1766 |
113 |
92.0 |
– |
– |
– |
Middle Rhine towns |
Germany |
1346–99 |
9 |
> 0.0 |
0.0 |
– |
– |
Middle Rhine towns |
Germany |
1400–49 |
10 |
> 70.0 |
70.0 |
– |
– |
Middle Rhine towns |
Germany |
1450–99 |
17 |
> 76.5 |
76.5 |
– |
– |
Middle Rhine towns |
Germany |
1500–49 |
13 |
> 76.9 |
76.9 |
– |
– |
Middle Rhine towns |
Germany |
1550–99 |
17 |
> 76.5 |
76.5 |
– |
– |
Middle Rhine towns |
Germany |
1600–29 |
9 |
100.0 |
100.0 |
– |
– |
Middle Rhine towns |
Germany |
1346–99 |
2 |
– |
– |
0.0 |
– |
Middle Rhine towns |
Germany |
1400–49 |
2 |
– |
– |
0.0 |
– |
Middle Rhine towns |
Germany |
1450–99 |
12 |
– |
– |
8.3 |
– |
Middle Rhine towns |
Germany |
1500–49 |
7 |
– |
– |
14.3 |
– |
Middle Rhine towns |
Germany |
1550–99 |
21 |
– |
– |
47.6 |
– |
Middle Rhine towns |
Germany |
1600–29 |
13 |
– |
– |
69.2 |
– |
50 Italian cities |
Italy |
1220–1800 |
1,132 |
40.2 |
– |
– |
– |
Amsterdam |
N. Netherlands |
1688 |
53 |
– |
– |
– |
< 60.0 |
Castilian towns |
Spain |
1251–1600 |
253 |
– |
10.0 |
– |
– |
Spanish towns |
Spain |
1250–1600 |
309 |
22.7 |
9.4 |
– |
– |
Cordova |
Spain |
1250–1600 |
45 |
– |
4.4 |
– |
– |
Grenada |
Spain |
1250–1600 |
41 |
– |
12.2 |
– |
– |
Murcia |
Spain |
medieval |
16 |
0.0 |
0.0 |
0.0 |
0.0 |
Seville |
Spain |
1250–1600 |
76 |
– |
10.5 |
– |
– |
Total no. guilds |
|
|
2,291 |
1,691 |
815 |
73 |
69 |
Weighted average |
|
|
|
43.4 |
14.9 |
28.8 |
46.1 |
Notes: Weighted average is average % across samples, weighted by the number of guilds in each sample. In calculating weighted average, numbers of the form “> 76.5” are standardized to “76.5”. Source: Quantitative guilds database: results of 18 studies of prevalence of apprenticeship prescriptions in samples of European guilds. |
A slightly more stringent measure is whether a guild imposed a minimum training term. German guilds were among the strongest in Europe, with the strictest regulations, but even they only gradually moved towards imposing minimum apprenticeship terms. Before 1400, no Middle Rhine guild had imposed minimum terms; between 1400 and 1600, about three-quarters did so; and the requirement became universal only after 1600. Spanish guilds were also among the strongest in Europe, but here, at least before 1600, minimum apprenticeship terms were imposed by only about 10 per cent. Table 7.13 has 12 observations of whether guilds imposed a minimum apprenticeship term, which together comprise a total of 815 guilds: across these, the weighted average was just 15 per cent.
The German sample also makes it possible to trace the emergence of guild demands for proof of apprenticeship. No guilds in the Middle Rhine sample imposed this requirement before 1450. The share rose to between 8 and 14 per cent in the period between 1450 and 1550, reaching 48 per cent in the later sixteenth century, and 69 per cent in the early seventeenth. Thus, even in some of the largest towns in Germany with the longest guild traditions, it was not until around 1600 that a majority of guilds required mastership applicants to prove they had undergone apprenticeship.
Table 7.13 also presents two studies of guild samples from the Netherlands and Spain, showing the percentage of total masters in guilds with mandatory apprenticeship. In medieval Murcia it was zero. This might be thought to reflect an economy that was in some way primitive, were it not for the fact that in seventeenth-century Amsterdam, arguably the most economically advanced city in early modern Europe, less than 60 per cent of guild masters were subjected to mandatory apprenticeship requirements.
The numbers speak clearly. Guilds were not synonymous with apprenticeship. In many European societies, just as apprenticeship existed without guilds, so guilds existed without apprenticeship. In some societies a majority of apprenticeships were organized and enforced outside the guild system. In others a majority of guilds existed without requiring apprenticeship. These were not backward and peripheral parts of Europe, but included some of the most advanced economies of their time: medieval Italy, early modern England, and early modern Holland. In highly successful economies at the forefront of pre-modern development, therefore, guilds were neither necessary nor sufficient for providing skilled training in the secondary and tertiary sectors.
ENFORCEMENT: PRINCIPLE OR PRIVILEGE?
When guilds did impose apprenticeship rules, what was their attitude towards enforcing them? Did they apply the rules equally to everyone, signaling a belief that training was objectively needed? Or did they apply them arbitrarily, as another privilege to use for extracting cartel profits? It varied. Some guilds enforced apprenticeship rules stringently; others hardly at all; still others granted lavish exemptions.
On the severe end of the spectrum lay Germany, where by c. 1600 most guilds imposed minimum apprenticeship terms and enforced them quite strictly. An illustrative example is provided by the Mainz cabinetmakers’ guild, which combined the four occupations of cabinetmakers, barber-surgeons, turners, and bath-masters. The guild imposed a minimum apprenticeship term for each occupation, which Kurt Wesoly checked against the actual length of training of 193 apprentices admitted from 1575 to 1618. All 80 barber-surgeon apprentices fulfilled the mandated minimum term. Among the 16 turner apprentices, 15 fulfilled the mandated term and one extended it by a year to avoid paying the minimum premium. Among the 97 cabinetmaker apprentices, 88 trained for the mandated period and 9 trained for longer to avoid paying the minimum premium. Not one apprentice trained for less than the minimum mandated period, and nearly 6 per cent trained for a longer period.130 The rural-urban worsted-weavers’ guild of the district of Wildberg in the Württemberg Black Forest shows similarly strict enforcement: of 1,258 apprentices admitted in 104 years of surviving records between 1598 and 1760, fewer than 3 per cent were granted dispensations from normal entry requirements; of 1,035 masters admitted in 143 years of surviving admission records between 1598 and 1760, fewer than 1 per cent received dispensations; most dispensations merely involved a reduction in fees, and none involved a reduction in the mandatory minimum apprenticeship term.131
On the relaxed end of the enforcement spectrum, by contrast, were societies such as the Northern Netherlands and Italy. Here, as we have seen, many guilds did not require apprenticeship at all and even fewer imposed minimum terms. As a general rule, young people in these societies entered apprenticeships when they or their families regarded it as useful, for the length of training that they deemed necessary, and by means of contracts with masters that were neither required nor registered by the guild. Lax enforcement was also the norm in England. Despite the widespread custom of many English towns—enshrined in national law after 1563—that required guild and non-guild apprentices alike to serve a seven-year term, in practice this rule was not strictly enforced. Many English apprentices signed up for seven years but then began late, left early, or interrupted their training in midstream for lengthy periods. In early modern Shrewsbury, for instance, less than 70 per cent of apprentices were actually living in their masters’ households (and thus receiving training) in the first year of apprenticeship, and the proportion resident with their masters never exceeded 71 per cent during any subsequent year of the legally mandated seven; thus even in a city such as Shrewsbury, which had notoriously strong guild regulations by English standards, a large share of apprentices shortened their seven-year terms by starting late, departing early, or both.132 In many English towns, apprentices took work as journeymen or set up in business as independent masters without completing apprenticeship, and guild scholars regard the few prosecutions of such non-apprenticed practitioners as only the tip of a much larger iceberg of tolerated non-compliance.133 By 1619, violations of guild training requirements were so widespread that the English crown “established a commission to sell pardons to those who had evaded the obsolete apprenticeship laws”.134
Most societies lay somewhere in the middle of this spectrum: more flexible than Germany, but less so than England, the Northern Netherlands, or Italy. In such societies, guilds typically imposed mandatory apprenticeships but then issued exemptions. Many guilds waived lengthy periods of apprenticeship in return for fees paid to the individual master as a premium, to the guild as a dispensation fee, or to both. Other guilds granted exemptions to masters’ relatives, as a perk for existing guild members. Still other guilds exempted applicants who bought privileges from the crown, nobles, the church, or other powerful bodies.
The qualitative database contains 134 observations of guilds that created such privileged pathways, shown in Table 7.14. The observations of specific guilds cover more than 80 different occupations, including such skilled crafts as bookbinders, booksellers, cabinetmakers, enamellers, furriers, gilders, glaziers, gold- and silver-smiths, jewellers, locksmiths, makers of strings for musical instruments, rosary-makers, saltpeter-makers, shearers and finishers of fine broadcloth, silk-weavers, sword-makers, tapestry-makers, wigmakers, and wire-drawers. The table also includes a number of samples that cover a mix of occupations.
TABLE 7.14: Exemptions from Apprenticeship Granted by Guilds, 1268–1848 |
||||||||||||
|
|
Masters’ sons |
|
Masters’ sons-in-law |
Masters’ widows’ husbands |
Guild officers’ sons-in-law |
Citizens’ sons |
Those who pay fee |
Total |
|||
Group exempted |
solely |
full |
partial |
full |
partial |
full |
partial |
full |
full |
partial |
no. |
% |
Country: |
||||||||||||
Austria |
1 |
– |
1 |
– |
– |
– |
– |
– |
– |
2 |
4 |
3.0s |
Bohemia |
– |
2 |
– |
1 |
– |
– |
– |
– |
– |
– |
3 |
2.2s |
France |
4 |
2 |
– |
– |
– |
– |
1 |
1 |
6 |
42 |
56 |
41.8hh |
Germany |
2 |
2 |
5 |
1 |
1 |
1 |
– |
– |
– |
15 |
27 |
20.1l |
Italy |
1 |
4 |
– |
2 |
– |
– |
– |
– |
1 |
2 |
10 |
7.5ll |
N. Netherlands |
– |
– |
– |
– |
– |
3 |
– |
– |
– |
– |
3 |
2.2l |
S. Netherlands |
4 |
12 |
1 |
– |
– |
– |
– |
– |
2 |
2 |
21 |
15.7hh |
Spain |
– |
4 |
– |
– |
1 |
– |
– |
– |
– |
– |
5 |
3.7s |
Switzerland |
– |
4 |
1 |
– |
– |
– |
– |
– |
– |
– |
5 |
3.7s |
Period: |
||||||||||||
Medieval |
2 |
7 |
2 |
– |
1 |
– |
– |
– |
– |
38 |
50 |
37.3hh |
Early modern |
10 |
23 |
6 |
4 |
1 |
4 |
1 |
1 |
9 |
25 |
84 |
62.7ll |
Total no. |
12 |
30 |
8 |
4 |
2 |
4 |
1 |
1 |
9 |
63 |
134 |
100.0 |
Total % |
9.0 |
22.4 |
6.0 |
3.0 |
1.5 |
3.0 |
0.7 |
0.7 |
6.7 |
47.0 |
100.0 |
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. England has zero observations in this table, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database overall. For all other countries not represented in the table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. “Solely” = guild admits only masters’ sons, without apprenticeship. Source: Qualitative guilds database: 136 observations of guilds granting exemptions from apprenticeship. |
As Table 7.14 shows, the observations come from nine societies in northwest, central, and southern Europe, but France and the Southern Netherlands are significantly and substantially over-represented compared to their share of the overall guilds database. French guilds habitually sold apprenticeship exemptions to raise revenues, while the guilds of the Southern Netherlands habitually granted exemptions to masters’ sons. Conversely, England and Italy are significantly under-represented in Table 7.14. The absence of exemptions in England may arise partly from the fact that English apprenticeships were enforced through state rather than guild rules after the 1563 Statute of Artificers was in place, and partly from the notorious laxity of English guilds in enforcing their rules about anything. The under-representation of Italy seems likely to arise from the fact that, as we saw in Table 7.13, 60 per cent of Italian guilds included no apprenticeship prescriptions in their ordinances. Although by 1600 German guilds usually had apprenticeship requirements, imposed minimum terms, and enforced them systematically, it is nonetheless represented in Table 7.14 in proportion to its share of the overall database, indicating that even one of the strictest guild systems in Europe sometimes treated apprenticeship as a matter of privilege rather than principle.
The observations in Table 7.14 span a 580-year period, from the guilds of Paris in the thirteenth century to those of Zürich, Bavaria, and Berlin in the nineteenth. The medieval period accounts for 38 per cent of observations, significantly higher than its share of the guild database as a whole. Nonetheless, in each century of the early modern period there were observations of guilds granting apprenticeship exemptions, and the practice of selling exemptions rather than merely granting them to masters’ relatives spread as the early modern period progressed. Exemptions from apprenticeship requirements were thus commonplace in European guilds across a wide range of occupations, societies, and time-periods.
Nearly 40 per cent of observations in Table 7.14 involve guilds exempting sons of guild masters—another indication, if one were needed, that European guilds did not render human capital investment independent of family ties.135 In 10 per cent of cases, guilds admitted only masters’ sons and waived any apprenticeship, as did the Paris guilds of the tanners, lemonade-makers, gold-beaters, and string-makers in 1766.136 In 22 per cent of cases, guilds admitted applicants who were not sons of masters but required them to have been apprenticed, while exempting masters’ sons. In a further 6 per cent of cases, guilds exempted masters’ sons from apprenticeship partially, as in medieval Ghent where the tapestry-weavers’ guild admitted masters’ sons after a single year of apprenticeship, while demanding five years for outsiders.137
It might be argued that masters’ sons did not need to undergo apprenticeships since they had been informally “apprenticed” to their fathers.138 But this is too sanguine. Apprenticeship waivers for masters’ sons were based on their privileged identity, not on their skills. Some guilds admitted masters’ sons at ridiculously young ages. In early modern Steyr the knife-smiths’ guild admitted them in their cradles;139 in one eighteenth-century Spanish town the weavers’ guild accepted a master’s son so young that “he did not know what a loom was”;140 and in nineteenth-century Zürich the dyers’ guild admitted masters’ sons the day they were born.141
Contemporaries also doubted that kinship with masters guaranteed proficiency. In eighteenth-century Bologna, the shoemakers objected strongly when their guild officers eased admission of masters’ relatives who lacked the requisite technical skills.142 In 1755, the officers admitted a master’s son who had never practised the craft “in any shop whatsoever” or even taken “the test for cutting and forming a shoe”.143 In all honesty, guilds themselves doubted that being a master’s son automatically conveyed skills that substituted for apprenticeship. In 1763, the Oulu coppersmiths rejected two masters’ sons in their late teens whose mother claimed they had learned the craft by helping their deceased father.144 In 1788, all the Montbéliard guilds collectively ordained that “in order to render the manufactures of our little country recommendable in foreign parts, no-one is to be dispensed from the masterpiece, and in particular the son of a master is required to do one, as is someone who marries the widow or daughter of a master”.145
In fact, as emerges from 9 per cent of cases in Table 7.14, guilds granted exemptions from training to men who married masters’ daughters, men who married masters’ widows, and even men who were simply sons of local citizens. Such men did not even have the advantage of informal family “apprenticeships”. It was their identity, not their skills, that got them their guild masterships. As the former Bavarian royal official Joseph von Destouches remarked in 1809, the Freystadt wire-drawers’ guild exempted from training anyone who married a master’s widow or daughter, so that “persons are taken into manufacturing businesses who are lacking in skill”.146
The same was true of the many cases, comprising over half of the observations in Table 7.14, in which guilds allowed men to pay a higher fee in return for a full or partial exemption from apprenticeship. In these cases, no-one pretended that mandatory apprenticeship had anything to do with learning skills. The guild imposed the apprenticeship requirement in order to create a license that could be sold. To give just one of many examples, the soap- and candle-makers of early modern Linz allowed apprentices to reduce their training period from 3 years to 2 as long as they paid 20 Reichstaler to the master, 4 Reichstaler to the guild, and 4 Reichstaler to the local authorities, a sum equivalent to two and a half years’ wages for a Viennese journeyman.147
A non-trivial reduction in training time could be achieved by paying such fees. The quantitative guilds database contains 56 observations of this practice, drawn from five European societies across a period of nearly six centuries, from 1268 to 1848. As Table 7.15 shows, the mean duration of apprenticeship without paying for a dispensation was 7.2 years; the mean duration if an apprentice paid was 5.3. The percentage of the training period that could be bought off ranged from 14 per cent (a guild in eighteenth-century Austria) to 100 per cent (a guild in the eighteenth-century Southern Netherlands). Between the medieval and early modern period, the average apprenticeship term in the sample declined from 8.3 to 4.4 years, while the average number of years that one could buy off remained stable at about 2. The average percentage reduction that could be achieved by paying a fee thus increased from 24 per cent in the medieval period to 43 per cent in th early modern period. Leaving aside the special case of the Southern Netherlands, where you could buy a 100 per cent reduction, the percentage reduction was 22 to 24 per cent in Austria and France and 31 to 36 per cent in Germany and Italy.
TABLE 7.15: Reduction in Apprenticeship Term Obtained by Paying Premium or Fee, 1268–1848 |
||||||
|
No. obs. |
Years without fee (mean) |
Years with fee (mean) |
Years of reduction (mean) |
Reduction as % of total years without feea |
Fee as days of journeyman’s wagesb |
Country |
||||||
Austria |
2 |
5.0 |
4.0 |
1.0 |
23.8 |
672.0 |
France |
33 |
9.0 |
7.0 |
2.0 |
22.2 |
67.0 |
Germany |
17 |
4.3 |
2.8 |
1.6 |
36.1 |
213.0 |
Italy |
2 |
6.5 |
4.5 |
2.0 |
30.9 |
– |
S. Netherlands |
2 |
3.0 |
0.0 |
3.0 |
100.0 |
14.3 |
Period |
||||||
Medieval |
39 |
8.3 |
6.4 |
1.9 |
23.9 |
67.6 |
Early modern |
17 |
4.4 |
2.6 |
1.8 |
42.6 |
304.3 |
Total |
56 |
7.2 |
5.3 |
1.9 |
29.6 |
114.9 |
Notes: a Calculated by averaging the reduction as % of total years without fee for each individual guild, not the average for the country or period category in the table. b In 11 cases, the size of the fee charged by the guild for a reduction in apprenticeship is not recorded. Sources: Quantitative guilds database: 56 observations of reductions in apprenticeship term obtainable by paying a fee. |
The fees people paid to avoid apprenticeship were also non-trivial. For 45 of the 56 observations, we know the precise exemption fee the guild charged. The final column of Table 7.15 shows the equivalent in terms of journeymen’s wages. The lowest was 14 days in the eighteenth-century Southern Netherlands, and the highest 672 days in eighteenth-century Austria. The cost of buying off periods of apprenticeship rose over time, from 68 days’ wages in the Middle Ages to over 300 in the early modern period. Across all 45 observations, the average cost of avoiding apprenticeship was 115 days, approximately 43 per cent of a fully employed working year of 270 days. The bargain might still have seemed a good one, given the average two-year reduction in apprenticeship. However, except in the Southern Netherlands, the sums were high enough that only young men with private means could afford the privilege.
Guilds also sold masterships to completely non-apprenticed persons. In Paris in the 1730s, the goldsmiths accused their guild officers of admitting “persons without quality”— i.e., who had no training—in return for fees of 1,000 livres, a gigantic sum in an era when a journeyman in a non-luxury craft might earn 60 livres annually.148 In Naples in the 1760s, the silk guild was accused of accepting large fees in exchange for admitting men who had undergone no training of any sort.149
There were strong financial incentives for guilds, their officers, and the public authorities to impose mandatory apprenticeship requirements and then sell exemptions. The sale of masterships to non-apprenticed persons often brought in a large share of the guild’s revenues, which could then be used to satisfy fiscal demands; this motivated state and guild officials to approve the practice.150 Guild officers also kept the fees from selling masterships as personal perquisites to recoup the high prices they had paid for their own offices.151 Sometimes guild officers or state authorities would create a block of “extraordinary” masterships which were offered for sale to applicants who had no training. The Paris goldsmiths’ guild created hundreds of these in the 1730s to sell at high prices to young men who had not undergone apprenticeship; and in 1765–66, 150 new masterships were offered to fee-paying, untrained entrants in a collusive arrangement between the Paris wigmakers’ guild officers and the crown.152
The scale of guild admissions of untrained persons could be quite large. The quantitative guilds database contains 10 observations of French guilds which recorded the number of masters admitted under each possible rubric. These are shown in Table 7.16. The observations come from five different cities, cover various periods between 1720 and 1789, and are based on a total of over 9,000 entering masters. They include all-female as well as all-male guilds, and the occupations from which they are drawn span a wide spectrum, from seamstresses to goldsmiths.
Across all observations in Table 7.16, nearly two-thirds of new masters got in without serving apprenticeships. In only two cases were more than half the new masters admitted via apprenticeship, and these two were guilds that were unusual in being all-female (the seamstresses of Paris and Caen). This does not mean that all-female guilds were invariably meritocratic, since one of the lowest proportions of admissions via apprenticeship (10 per cent) was that of the all-female Rouen ribbon-makers’ guild. In the all-male tailors’ guilds, a maximum of about one-third of masters was admitted via apprenticeship (in Caen); in Aix-en-Provence it was less than 9 per cent. But even in high-quality, luxury crafts such as that of the Paris goldsmiths, less than half of new masters were admitted via apprenticeship in the early eighteenth century, and the number fell to about one-third after 1750. Across the whole sample of 9,180 new masters in Table 7.16, the weighted averages (adjusted for sample size) show that 53 per cent had undergone apprenticeships, compared to 27 per cent admitted as masters’ relatives, 19 per cent through purchase or special privileges, and 1 per cent as charity cases. In both everyday and luxury crafts, therefore, guilds in one of the most advanced economies in western Europe admitted around one-half their masters on the basis of privilege rather than skilled training—even though, on paper, these guilds imposed an apprenticeship requirement.
TABLE 7.16: Percentage of Masters Admitted without Apprenticeship, Various French Guilds, 1720–1789 |
|||||||||||
Place |
Period |
Occupation |
Guild type |
Apprenticeship or masterpiece (%) |
Master’s offspring (%) |
Master’s son-in-law (%) |
Married master’s widow (%) (%) |
Guild fee, privilege, or government order in lieu of apprenticeship (%) |
Other, mostly charity (%) |
Total admitted without apprenticeship (%) |
Total apprentices no. |
Aix-en-Provencea |
1745–75 |
tailors |
male |
8.3 |
52.8 |
22.2 |
0.0 |
13.9 |
2.8 |
91.7 |
72 |
Bordeaux |
1776–89 |
goldsmiths |
male |
0.0 |
– |
– |
– |
– |
– |
100.0 |
25 |
Caenb |
1724–75 |
tailors |
male |
34.1 |
47.7 |
0.0 |
0.0 |
18.2 |
0.0 |
65.9 |
44 |
Caenc |
1724–75 |
seamstresses |
all-female |
76.2 |
23.8 |
0.0 |
0.0 |
1.6 |
0.0 |
23.8 |
63 |
Paris |
1720–75 |
goldsmiths |
male |
45.8 |
31.7 |
0.0 |
0.0 |
22.6 |
0.0 |
54.2 |
616 |
Parisd |
1735–76 |
tailors |
male |
13.5 |
22.0 |
31.3 |
8.5 |
22.0 |
2.7 |
86.5 |
2,681 |
Paris |
1735–76 |
seamstresses |
all-female |
75.0 |
8.3 |
0.0 |
0.0 |
16.6 |
0.1 |
25.0 |
5,509 |
Paris |
1735–8 |
goldsmiths |
male |
45.1 |
49.0 |
0.0 |
0.0 |
5.9 |
0.0 |
54.9 |
51 |
Paris |
1757–9 |
goldsmiths |
male |
33.7 |
32.6 |
0.0 |
0.0 |
33.7 |
0.0 |
66.3 |
89 |
Rouene |
1768 |
ribbonmakers |
all-female |
10.0 |
56.7 |
0.0 |
0.0 |
16.7 |
16.7 |
90.0 |
30 |
Unweighted average |
|
|
|
34.2 |
36.1 |
5.9 |
0.9 |
16.8 |
2.5 |
65.8 |
9,180 |
Weighted average |
|
|
|
53.4 |
15.2 |
9.3 |
2.5 |
18.6 |
0.9 |
46.6 |
|
Notes: a other = Hôpital de la Charité permit. b excludes 18 masters admitted by non-specified paths. c excludes 9 masters admitted by non-specified paths. d other = had served at Hôpital de la Trinité; excludes 60 masters admitted by non-specified paths. e other = charity permit. Sources: Quantitative guilds database: 10 observations of guilds for which % of masters admitted under different rubrics is known. |
DID GUILDS MAKE SURE MASTERS TRAINED THEIR APPRENTICES?
Guilds, we have seen, were neither necessary nor sufficient for apprenticeship: there were many guilds without apprenticeships and many apprenticeships without guilds. But perhaps apprenticeships inside the guild framework were in some way better than those that took place outside it? One of the great challenges in studying human capital investment, after all, is that education is a black box. What actually happens between teacher and pupil? Clearly some training relationships work well and transform untrained neophytes into skilled practitioners. But others work poorly: resources go in, but nothing much comes out. What institutions make the process of human capital investment work better?
Some argue that guilds did. A major problem in the relationship between trainer and trainee is opportunism. A trainee may abscond between getting training and becoming a productive worker. A trainer may fail to provide training and instead exploit the apprentice for menial tasks. Fear of such opportunism may deter both parties from entering a training relationship. A guild might provide mechanisms to control opportunism, inducing teachers and pupils to form training relationships they would otherwise have avoided, thus benefiting the wider economy.153
These ideas sound persuasive, but do the facts support them? A test proposed by some scholars is whether guild apprentices actually completed their periods of training. These scholars interpret high completion rates as an indicator that guilds were indeed dealing well with opportunism and low ones as an indicator that something was going wrong. Even if the guild was not causing the problem, it was failing to provide mechanisms to prevent it.
The quantitative database contains 92 observations of European guilds for which some measure of the apprentice completion rate is known. The observations refer to more than 40 different occupations and span almost five centuries from 1370 to 1858. They are drawn from seven different European societies, although England and the Southern Netherlands dominate.
TABLE 7.17: Measures of Apprenticeship Non-Completion, 1370–1859 |
||||||
|
Not completing apprenticeship (%) |
Not obtaining town citizenship locally (%) |
Not becoming guild master locally (%) |
All three measure of apprentice non-completion (%) |
||
|
mean % |
mean % |
mean % |
mean % |
N |
% of obs. |
Country: |
||||||
Austria |
33.0 |
– |
– |
33.0 |
10 |
10.9 |
England |
– |
60.8 |
87.1 |
62.0 |
46 |
50.0 |
Finland |
33.3 |
– |
66.7 |
50.0 |
2 |
2.2 |
France |
30.0 |
– |
– |
30.0 |
1 |
1.1 |
Germany |
23.8 |
– |
81.8 |
36.7 |
9 |
9.8 |
S. Netherlands |
41.7 |
– |
78.7 |
73.9 |
23 |
25.0 |
Sweden |
26.0 |
– |
– |
26.0 |
1 |
1.1 |
Period: |
||||||
Medieval |
– |
61.3 |
78.4 |
73.9 |
23 |
25.0 |
Medieval & early modern |
– |
58.6 |
85.5 |
66.3 |
7 |
7.6 |
Early modern |
30.9 |
61.0 |
79.2 |
51.6 |
62 |
67.4 |
Unweighted mean |
30.9 |
60.8 |
79.1 |
58.3 |
|
|
N |
23 |
44 |
25 |
|
92 |
|
Samples with 2 measures: |
||||||
London tailors, 1425–45 |
– |
65.0 |
86.6 |
|
|
|
London tailors, 1453–8 |
– |
65.0 |
87.6 |
|
|
|
London mercers, 1391–1464 |
– |
49.5 |
75.4 |
|
|
|
Turku mixed, 18th c. |
33.3 |
– |
66.7 |
|
|
|
Wildberg weavers, 17th c. |
57.0 |
– |
89.7 |
|
|
|
Notes: Mean is unweighted, as information on sample size is not available for all samples. Expressions such as “53 to 61” have been averaged to 57. Source: Quantitative guilds database: 92 observations of different measures of apprenticeship non-completion. |
Table 7.17 reports three different measures of apprentice non-completion used by different studies. The first is the percentage of apprentices not completing their term of training; this is the most accurate figure, since it takes a particular group of youths who started training and records whether they finished it. The second is the percentage of apprentices not obtaining town citizenship; this measure is specific to England, where a young man who completed apprenticeship in a town had a virtually automatic entitlement (though no obligation) to take up town citizenship. This measure will over-estimate the non-completion rate by including people who completed apprenticeship but left the town without taking citizenship because they intended to work elsewhere. The third measure in Table 7.17 is the percentage of apprentices not becoming masters of the guild in that town; this will over-estimate non-completion even more, by including people who completed apprenticeship but either left the town or stayed but failed to get guild mastership.
The gap between these three measures was wide. The final panel of Table 7.17 displays findings from five studies which provide multiple measures for the same sample of apprentices. The third measure (failing to become a local master) is about 22 percentage points higher than the second (failing to become a local citizen) and about 33 percentage points higher than the first (failing to complete training).
Nonetheless, the figures in Table 7.17 show that apprentice non-completion rates were high. The 79 per cent of apprentices not becoming local guild masters and the 61 per cent not becoming local town citizens undoubtedly greatly over-estimate the extent to which apprentices did not finish training. But even the most conservative measure shows that an average of about 31 per cent of guild apprentices did not complete their agreed training term. These high rates are consistent with newly reported figures for the seventeenth and eighteenth centuries (published too late for inclusion in the quantitative database), showing at least 34 per cent of youths not completing apprenticeships in Lyon (whether because of death, cancellation, or transfer to another master), 46 per cent in Shrewsbury, and 53 to 64 per cent in Leiden and Amsterdam.154
The question is how to interpret these findings. There were multiple reasons guild apprentices might fail to complete training. One was death. But this can only account for a minority of non-completions. In sixteenth-century London, for instance, only about 10 per cent of apprentices died during training.155 In smaller cities, mortality was lower, as shown by the fact that in Lyon in the 1680s, 1740s, and 1760s, only 1.2 per cent of apprentices left training because they died.156
A second reason for non-completion was apprentice opportunism. As we have seen, guilds imposed minimum training terms much longer than technically required to learn the skills. So, many apprentices quit when they had learned everything they thought they would need, intending to set up as independent masters outside guild jurisdictions where they would not be penalized for practising without a guild completion certificate.157 One reason for the high non-completion rates in England and the Netherlands may have been the comparative weakness of their guild systems, which offered more interstices for earning a living without a guild completion certificate.158 Chapter 9 discusses this guild weakness more comprehensively.
A third reason for apprentice non-completion was master opportunism, as shown by the many observations in the qualitative database in which guild masters failed to train apprentices, impelling them to quit. In Genoa in 1462, for instance, an apprentice silk-weaver quit because his master refused to teach him to weave, so he was unable to “learn the art”.159 In Upper Austria in 1658, an official reported masters failing to train apprentices and making them do menial work so that they “either run away from sheer fright or go to illegal masters”.160 In Nuremberg in 1789, a tailor’s apprentice complained that his master gave him “little instruction in craft matters,” while a locksmith’s apprentice lamented that “during this entire time he has not even seen a key being made”.161 Masters were less likely to behave opportunistically towards apprentices who were their own sons, and this is reflected in evidence from early modern Shrewsbury, Amsterdam, and Leiden, showing that apprentices being trained by their fathers were substantially and significantly less likely to quit training than those bound to strangers—yet another indication that guilds did not make human capital investment independent of family ties.162
Evidence on the timing of apprentices leaving their training provides indirect indications of their reasons for quitting. If departures are clustered at the beginning of the training period, it suggests master opportunism, or at best a poor match between trainer and trainee. If departures are clustered towards the end of training it suggests apprentice opportunism, or at least that apprentices thought they had already learned sufficient skills. Evidence from early modern London, Bristol, Shrewsbury, and Lyon shows no clear clustering, with apprentices quitting steadily throughout their terms, implying a mixture of motivations among both apprentices and masters. However, guilds in these cities show a substantial increase in the numbers of apprentices quitting in the second half of the prescribed training period, suggesting that many left because they believed they had obtained sufficient skills for their planned future.163
However we interpret the high apprentice non-completion rates in Table 7.17—as apprentice opportunism, master opportunism, or both—they reveal the extent to which guilds failed to correct failures in markets for training. Some guilds did penalize opportunism on the part of masters or apprentices, at least on paper. But others did not. An analysis of 309 Spanish guild ordinances between 1250 and 1600, for instance, found that hardly any contained provisions to deal with opportunistic behaviour by either party, in contrast to private apprenticeship contracts where such provisions were universal.164 Even where rules existed to deal with opportunism in training relationships, guilds often failed to enforce them, especially when the fault lay with the master. Many guilds, as associations of employers, tended to judge conflicts in favour of those employers.165 Others honestly sought to ensure fair play but did not have the power to coerce their more powerful members.
In either situation, the parties felt they could not rely on guild enforcement and instead often resorted to more impartial tribunals. In the fifteenth century, London apprentices repeatedly resorted to public law-courts when the tailors’ guild failed to enforce judgements against abusive but powerful masters,166 while in the seventeenth and eighteenth centuries, large numbers of London apprentices turned to the Lord Mayor’s Court seeking dissolution of their contracts on the grounds that their masters had failed to provide training.167 Apprentices whose masters failed to provide training and whose guilds failed to remedy the abuse resorted to public courts of law in many other European towns, including Aachen,168 Antwerp,169 Augsburg,170 Berlin,171 Bristol,172 Ghent,173 Nuremberg,174 Paris,175 Valencia,176 Würzburg,177 and various towns in Austria,178 Württemberg,179 and Spain.180 In eighteenth-century Augsburg, a special municipal tribunal recorded a ceaseless stream of apprentice complaints to the effect that “he could certainly learn more if he were used for more than just taking care of the children”, “he had not learned anything”, or “he has so far produced only one sort of work”. Frequently, according to apprentices’ accusations, the master “does not teach the essentials”, “never shows him anything from the trade”, “never gives him new tasks to complete”, or “never let him near a chair or showed him how to set one up”.181
The high non-completion rates for apprentices in many guilds, the widespread evidence of opportunism as a cause of non-completion, and the important role played by public institutions in dealing with accusations of opportunism all cast doubt on the idea that guilds provided institutional mechanisms that made training relationships work better.
HOW CAREFULLY DID GUILDS MONITOR SKILLS?
There is, however, one other reason guild apprenticeships might have been better than private ones. Guilds could specify skills, examine performance, and provide certificates. The classic picture of a guild is, indeed, of a reservoir of craft expertise with institutional mechanisms for laying down a training curriculum, examining trainees, and certifying proficiency. What do the facts show?
Very few guilds specified a curriculum. It was extremely rare for a guild to lay down what knowledge, skills, performance, attitudes, or values masters were supposed to teach apprentices, or how this teaching was supposed to be undertaken. In fifteenth-century Dijon, for instance, even the highly skilled and technically demanding goldsmiths said nothing in their statutes about what apprentices would have to learn: any apprentice who had done his six years’ training, and any master’s son at all, could become a master by paying a fee and hosting a dinner for the guild’s officers.182 In Antwerp in 1582, when the diamond-cutters applied for guild privileges, they made no reference to the skills involved, except for affirming airily that diamonds were to be “well and carefully cut”.183 Medieval and early modern German guilds, too, made virtually no mention of the knowledge, skills, or teaching methods involved in the training of apprentices.184 In Württemberg, for instance, the national bakers’ ordinance of 1629 said nothing about skills, simply ordering that “we leave it to remain in future with the custom in each locality, as it has been customary up to now”.185 In the early modern Netherlands, few guilds other than the surgeons defined the content of what apprentices were supposed to learn or how they were supposed to be taught.186 Even in painting and bookselling, Dutch guilds defined neither the content nor the skills of apprenticeship training.187
It might be thought that although guilds did not provide a clear description of the skills they required, they might still have monitored proficiency through an examination or “masterpiece”. But not all guilds had such tests. According to Steven Epstein, the distinguished historian of Genoese guilds, tests and masterpieces “seem to have been the exception and not the rule” among guilds in the Middle Ages, and spread only gradually after c. 1500.188 Table 7.18 summarizes the 88 observations in the qualitative guilds database of the date at which guilds introduced examinations and masterpieces. The timing varied across societies and among guilds in the same society, but the evidence broadly bears out Epstein’s chronology. In general terms, guild examinations were rare in the thirteenth and fourteenth centuries, spread gradually in the fifteenth, and became widespread in the sixteenth, but they were still not universal in the seventeenth and eighteenth centuries.
Even in crafts that were highly skilled and technologically demanding, such as those of barber-surgeons, clockmakers, glassmakers, goldsmiths, painters, printers, and shipbuilders, guild examinations and masterpieces were not universal. The glassmakers of Venice, acknowledged as the peerless practitioners of the craft in medieval Europe, mentioned no apprenticeship examination in their 1271 regulations and did not require one until 1441.189 The glassmakers of Altare in northern Italy, also outstanding artisans, made no mention in their 1495 ordinance of any examination or masterpiece.190 The Dijon goldsmiths’ guild required no masterpiece in the medieval period, introducing one only in the sixteenth century.191 The Rome barber-surgeons’ guild was established in the fifteenth century, but did not impose a mastership examination until 1593.192 The high-quality woollen guild of Padua did not impose an examination from 1520 until the 1630s, and even then only introduced the test for makers of cloth and knitwear.193 The Augsburg clockmakers’ guild was at the forefront of skill and technical expertise in the sixteenth century, but as late as 1569 many of its masters had never produced a masterpiece.194 The London Clockmakers’ Company produced some of the most advanced timepieces in Europe, but in 1656 journeymen were becoming masters without any test or “proof-piece”.195 In the “golden age” of Dutch painting, no masterpiece examination was imposed by Dutch painters’ guilds.196 The Northern Netherlands had the most successful shipbuilding industry in Europe in the sixteenth and seventeenth centuries, but none of its 16 ship-carpenters’ guilds required any test of skills until the 1570s; even after that date, only three ever imposed an examination, and this focused exclusively on the working of the wood, leaving untested the much more demanding skill of designing the ship.197
TABLE 7.18: Guilds’ Introduction of Examination or Masterpiece in Different European Countries |
|
Country |
Introduction of guild examination or masterpiece |
Austria |
guild examinations spread 16th c., still not universal 17th c.; some guilds in the 17th and 18th c. allow payment of fee in lieu of examination |
England |
in London, masterpieces not mentioned in medieval period, first mentioned 16th c., become widespread in 17th c., resisted by joiner journeymen in 1615 as entry barrier; some guilds (e.g. clockmakers) still not imposing masterpiece mid-17th c. |
France |
no Paris guild imposes examination in 1260s; Chartres guilds impose examination from 14th c. on; Dijon goldsmiths first impose examination 16th c.; Paris grocers first impose examination late 17th c. |
Germany: Augsburg |
some guilds impose examinations 16th c., not universal among clockmakers 1569, many guilds impose no examination in 18th c. |
Germany: Cologne |
armourers’ guild has no examination in 14th c., imposes examination in late 15th c., abolishes it in mid-17th c. |
Germany: Middle Rhine |
one-third of guilds impose examination in 14th & 15th c.; 43% in 1500–49; c. 70% in 1550–1630 |
Germany: Lüneburg |
no guilds impose examination in 14th c.; first examination imposed 1400 (goldsmiths); examination becomes common during 15th & 16th c. |
Germany: Mainz |
examination becomes widespread in second half of 16th c. |
Germany: Speyer |
examination becomes widespread in second half of 16th c. |
Germany: Verden a.d.A. |
glaziers’ guild does not introduce examination until 1839 (under government pressure) |
Germany: Württemberg |
all guilds introduce examinations in 16th c. |
Italy: Altare |
glassmakers’ guild imposes no examination in 1495 |
Italy: Padua |
woollen guild imposes no examination in 16th c. |
Italy: Palermo |
no guild examination before late 15th c.; a few guilds impose it in 1480s; examination universal by town law 1512 except masters’ sons & sons-in-law |
Italy: Rome |
few guilds have examination in medieval period; barber-surgeons first impose examination 1593, booksellers 1674, butchers 1690 |
Italy: Venice |
glassmakers have no examination in 1271, first impose examination 1441 |
N. Netherlands |
many Dutch guilds never impose examination; 13 of 16 shipbuilders’ guilds never impose examination, 3 introduce examination in 1570s; most printers’, binders’ and booksellers’ guilds impose examination in 17th c. except in Leiden and Amsterdam; Gouda transporters’, bakers’, and groats-grinders’ guilds do not impose examination in 18th c. |
Scotland |
most Scottish guilds introduce examination in 16th or 17th c. |
S. Netherlands |
many Leuven guilds impose no examination in 18th c. (mercers, butchers, grease-mongers, second-hand dealers) |
Spain |
in Castile as a whole, guild examinations are introduced gradually during 15th c., become generalized at beginning of 16th c.; Barcelona, Murcia and Seville introduce first examinations c. 1400 |
Sweden |
examinations not universal until 1720 (national law); even after 1720 many guilds let candidates pay a fee in lieu of examination |
Source: Qualitative guilds database: 88 observations of date at which guilds introduced examinations or masterpieces. |
Examination requirements could vary enormously within the same occupation in the same country. The shoemakers’ guild in the Dutch city of Arnhem, established in the fifteenth century, only introduced a masterpiece in 1674, while the guild of the same occupation in Gouda did not introduce an examination until the end of the eighteenth century.198 Bakers’ guilds in Arnhem and Gouda had no examination requirements, while those in Amsterdam, Utrecht, and Nijmegen did.199 Printers’ guilds imposed mastership examinations in most Dutch towns, but not in Leiden or Amsterdam.200 In 1688, the Seville silk-dyers’ guild had no examination, whereas the silk-dyers of Cordoba and Toledo did.201
These qualitative examples are borne out by quantitative studies. Table 7.19 reports the results of 14 studies of the percentage of guilds with examination or masterpiece requirements. These studies cover a total of at least 342 guilds (the size of some of the samples is not known), are drawn from six different societies, and span a period of 561 years from 1268 to 1829. All but three studies analyzed all guilds in a particular place and time-period. In some cases, all guilds imposed examinations (as in early modern Württemberg and Sweden), while in others none did (as in fourteenth-century Lüneburg). In the towns of the German Middle Rhine, the percentage rose from about one-third before 1500 to over two-thirds after 1600. But across all 14 studies, just 48 per cent of guilds provided mechanisms for examining skills.
TABLE 7.19: Percentage of Guilds with Examination or Masterpiece Requirements, c. 1268–1829 |
|||||
Country |
Place |
Occupations |
Period |
% guilds requiring examination or masterpiece |
Total no. guilds |
France |
Paris |
all |
c. 1268 |
13.0 |
100 |
Germany |
Lüneburg |
all |
1302–99 |
0.0 |
16 |
Germany |
Middle Rhine towns |
all |
1300–1499 |
35.3 |
17 |
Germany |
Middle Rhine towns |
all |
1500–49 |
42.9 |
7 |
Germany |
Middle Rhine towns |
all |
1550–99 |
71.4 |
21 |
Germany |
Middle Rhine towns |
all |
1600–29 |
69.2 |
13 |
Germany |
Nuremberg |
all |
c. 1750 |
59.9 |
110 |
Germany |
Württemberg towns |
all |
c. 1600 |
100.0 |
ng |
Italy |
Palermo |
all |
post-1512 |
100.0 |
ng |
N. Netherlands |
Dutch towns |
ship-carpenters |
pre-1570s |
0.0 |
16 |
N. Netherlands |
Dutch towns |
ship-carpenters |
post-1570s |
18.8 |
16 |
N. Netherlands |
Dutch towns |
painters |
early modern |
0.0 |
ng |
Sweden |
Swedish towns |
all |
post-1720 |
100.0 |
ng |
Switzerland |
Zürich |
all |
1829 |
57.7 |
26 |
Total |
|
|
|
47.7 |
> 342 |
Note: ng = number of guilds in sample is not given. Sources: Quantitative guilds database: 14 observations of percentage of guilds with examination or masterpiece requirements. |
What about the guilds that did have examinations? How accurately and impartially did guilds assess proficiency? The qualitative database contains 97 observations, drawn from over 50 different occupations, revealing some systematic problems afflicting guilds when it came to assessing skills. As Table 7.20 shows, these span more than five centuries, from 1271 to 1800, but the early modern period accounts for 95 per cent of observations, significantly higher than its share of the overall guilds database. This is unsurprising given that, as we have seen, guild examinations became widespread in most societies only after c. 1500. Observations of guilds’ problems in examining skills are found in ten different societies, but Spain and France are significantly over-represented, while England and the Northern Netherlands are significantly under-represented.
TABLE 7.20: Factors Vitiating Guild Examinations, 1271–1800 |
|||||||||||
|
Examination is entry barrier |
Examination favours insiders |
Examination is corrupt |
Examination is inefficient |
Total |
||||||
|
Guild uses exam to exclude entrants |
Guild makes exam costly |
Guild waives exam for masters’ relatives |
Guild sets easier exam for masters’ relatives |
Guild sets easier exam with personal ties |
Guild examiners take bribes |
Guild sells exemptions from exam |
Guild does not clarify exam content |
Guild lets unskilled applicants pass exam |
No |
% |
Country: |
|||||||||||
Austria |
– |
– |
– |
– |
– |
– |
1 |
– |
– |
1 |
1.1s |
Bohemia |
– |
– |
– |
– |
– |
– |
– |
– |
1 |
1 |
1.1s |
England |
1 |
– |
– |
– |
– |
– |
– |
– |
– |
1 |
1.1ll |
Estonia |
– |
– |
– |
– |
1 |
– |
– |
– |
– |
1 |
1.1s |
France |
1 |
7 |
5 |
4 |
– |
2 |
4 |
– |
1 |
24 |
25.5hh |
Germany |
2 |
8 |
1 |
– |
– |
– |
2 |
5 |
3 |
21 |
22.3s |
Italy |
– |
1 |
5 |
– |
2 |
1 |
1 |
6 |
1 |
17 |
18.1s |
Poland |
– |
– |
– |
– |
– |
– |
– |
– |
1 |
1 |
1.1s |
S. Netherlands |
– |
– |
6 |
– |
– |
– |
3 |
1 |
– |
10 |
10.6h |
Spain |
13 |
1 |
– |
– |
1 |
1 |
1 |
– |
– |
17 |
18.1hh |
Period: |
|||||||||||
Medieval |
1 |
– |
2 |
– |
– |
– |
– |
2 |
– |
5 |
5.3ll |
Early modern |
16 |
17 |
15 |
4 |
4 |
4 |
12 |
10 |
7 |
89 |
94.7hh |
Total no. |
17 |
17 |
17 |
4 |
4 |
4 |
12 |
12 |
7 |
94 |
100.0 |
Total % |
18.1 |
18.1 |
18.1 |
4.3 |
4.3 |
4.3 |
12.8 |
12.8 |
7.4 |
100.0 |
|
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The N. Netherlands has zero observations in this table, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database overall. For all other countries not represented in the table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 97 observations of problems with guild examinations of skills. |
The most commonly observed problem (36 per cent of cases) was that guilds used examinations as entry barriers. Many guilds explicitly justified their tests in such terms, as in sixteenth-century Speyer and Frankfurt where a number of guilds introduced examinations to prevent occupations being “over-filled”.202 Other guilds rationed access to the examination process itself, as in seventeenth-century Spanish cities where many guilds refused to examine journeymen in order to prevent them from setting up as masters,203 or in eighteenth-century Rouen where guilds refused to test journeymen from the countryside in order to stifle rural competition.204 Many guilds refused to recognize examinations from other towns, as in Reichenbach in eastern Saxony where the woollen-weavers’ guild declared in 1356 that “anyone who is apprenticed anywhere else and does not learn as is proper in Reichenbach and Görlitz, cannot become a master in Reichenbach”,205 or in early modern Toledo where the tailors’ guild claimed that “there is no locality in Spain where clothing is as well produced as in Toledo, and whoever wishes to settle in the guild here must let himself be examined in the special ways of work of Toledo”.206 Guilds often tried to restrict entry by deliberately imposing an esoteric examination, as in sixteenth-century Koblenz where the smiths’ guild required candidates to make an entire set of horseshoes for an animal led past them three times, whose hooves they could neither measure nor inspect,207 or in sixteenth-century Dijon where the hatters’ guild required candidates to make a masterpiece that was fifty years out of style, so no master even knew how to make it.208 Guilds also obstructed entry by making the test extremely costly, as in early modern Frankfurt where the guilds of the masons, stonemasons, locksmiths, and hose-knitters imposed masterpieces requiring three months or more of full-time work,209 or in eighteenth-century Turin where the masterpiece and examination fees for a non-local shoemaker journeyman cost 88 lire, equivalent to over 90 days’ wages.210
A second widespread problem (27 per cent of cases) was that guilds fixed the tests to favour insiders. In Chartres in the 1590s, for instance, the harness-makers’ guild required a master’s son to make a horse-collar, a master’s son-in-law to make a horse-collar and a saddle, and a journeyman unrelated to any master to make both items, plus a breech strap and “other pieces of work of the said occupation, as they shall be laid down for him”.211 The pewterers’ guild, likewise, required an unrelated journeyman to make three demonstration-pieces but a master’s son or son-in-law to make just one.212 It might be argued that growing up as a master’s son conveyed superior skills which obviated the need for any examination. Even for masters’ sons, however, contemporaries did not accept that privileged status substituted for proper assessment. We see this in seventeenth-century Gera where merchants complained that weavers’ sons lacked skill because they were exempted from any test,213 and in seventeenth-century Venice where the growing competitive threat of Dutch mirrors was ascribed to the mirror-makers’ guild admitting unskilled masters’ sons without examination.214
Not just kinship but other personal ties with guild members also vitiated guild examinations—as might be expected from a qualification system operated by an exclusive professional association. In early modern Spain, for instance, “the mastership examinations resembled less a checking of vocational knowledge than decision based on economic and familial selection”.215 In Padua in 1704, Pasquale Righi proved unable to answer basic technical questions about the wool to be used in making high-quality socks, but the hosiery guild admitted him nonetheless “because other masters and people certified that they had known him from an early age”.216 In eighteenth-century Bologna, the shoemakers’ guild admitted favoured candidates without proper examination, so men got in who were not “well instructed and expert”.217 In Naples in 1800, the consuls of the wool guild were described as “having granted membership without any examination to individuals with whom they had ‘special links’ . . . in order to defend their own personal interests”.218
A third and closely related problem, comprising 20 per cent of cases, was outright corruption. In the Guadalajara gilders’ guild in 1634, for instance, we are told: “It [was] public knowledge that by bribing [the guild examiners] and inviting them to cake and wine, examinations will consist of no tests at all; and when those who come from Guadalajara are asked in Madrid which tests they took, they reply ‘cake and wine’”.219 In Lucca in 1713, the silk-weavers’ guild was admitting applicants through “false exams”, in which candidates bribed the examiners with lavish meals and cheated openly.220 In Bologna in 1766, ordinary masters of the shoemakers’ guild accused the guild council of treating “[t]wo paper models, and a gold coin . . . [as] a sufficient test to obtain a master’s license”.221 More often, the examination was vitiated by the guild’s selling exemptions, as in 1614 when the Paris apothecaries admitted an applicant who paid 3,200 livres (5,333 days’ wages for a journeyman), without which “he would never have been accepted because of his incompetence”,222 in 1634 when the Guadalajara gilders sold mastership diplomas to outsiders who simply mailed in their fees without even visiting the city,223 or in 1768 when the Rouen ribbon-makers admitted one-sixth of new masters by selling them examination exemptions for 100 livres apiece (about a year and half’s wages for a journeyman).224
Even where guilds did not deliberately corrupt the examination process, they often lacked incentives to test skills adequately, as shown by a final 20 per cent of observations. Some guilds imposed examinations where the “answers” were an open secret, as in sixteenth-century Augsburg where the examination papers for clockmakers were described as being “so sketched off and copied from that, what with tracing and copying, everyone knows the way and how of them”.225 Guilds often resisted attempts to toughen or clarify their tests, since that would reduce their discretion to use examinations as an entry barrier. In sixteenth-century Speyer, for example, the tailors refused to specify the skills required and failed candidates without feedback.226 In eighteenth-century Prussia, the masons and carpenters vigorously resisted requests that state representatives be permitted to observe guild examinations.227
Even when a guild went through the motions of administering an examination, incompetent insiders got in easily. In Palermo in 1568, the embroiderers allowed Andrea Bellagamba to pass their examination even though the town council openly voiced doubts about his competence.228 In Wunsiedel in 1677, the cabinetmakers’ guild detected 19 serious errors in the masterpiece presented by Johann Caspar Leypold, but passed him to mastership with only a fine.229 In early modern Breslau, the furriers’ guild passed Georg Kadenbach even though he failed the exam twice for curing rabbit skins poorly and attaching fastenings that repeatedly tore off the fur.230 In Württemberg in 1793, the linen-weavers’ guild passed a candidate whose masterpiece local gossip described as “not masterly, so that it had to be improved by fulling”.231
Many guilds thus failed to test skills in any serious way. They neither laid down what candidates were supposed to be taught, nor examined whether they had learned it. Where they did conduct examinations, they passed candidates according to whether they had the right personal connections or could afford to pay the requisite fees.
Why would guilds behave in this way? Surely they enjoyed an unparalleled capacity to solve imperfections in markets for human capital investment? To understand why the empirical reality deviated so markedly from the ideal, consider the incentives. As a privileged association of masters, a guild had an incentive to certify members’ family members without examining their skills. Agency problems inside the guild meant that officers could profit by setting fraudulent or undemanding examinations for unqualified candidates in return for bribes and favours. Conversely, the guild and its officers had incentives to limit entry by rationing access to the examination process, refusing to recognize examinations from other places, deliberately formulating esoteric examinations, charging high examination fees, or requiring very time-consuming masterpieces. Theoretically a guild might have been unusually effective at testing and certifying skills, but empirically it had strong incentives to use its control over skills assessment to extract rents for its members.
These historical findings on guild examinations shed light on modern policy debates about occupational licensing. To what extent should people have to obtain licenses in order to be allowed to practise particular occupations, and what are the best institutional mechanisms for granting such licenses? Most occupational licensing regimes justify their requirements in terms of protecting the public against incompetent practitioners. Most professional associations claim to be the sole source of expertise for assessing competence. But as the empirical findings on guilds reveal, professional associations can also be motivated by their own interests, even if these conflict with the public good. Associations of entrenched producers have strong incentives to use examinations to limit entry, favour insiders, and collect fees; their officials may prefer to enjoy the perks of office while avoiding the costs of conscientious inspection. The result will be to exclude able practitioners while certifying incompetent insiders. This benefits members of the professional association, but at the expense of outsiders, consumers, and the wider economy.
GUILD TRAINING AND ECONOMIC DEVELOPMENT
How can we assess the overall effects of guilds’ training activities? Did guild training benefit the economy at large? Or did guilds exploit their exclusive privileges over training to extract benefits for their members at the expense of everyone else?
Guild scholars have approached this question from three directions. One focuses on intercontinental comparisons of a statistic called the “skill premium”, the difference in wages between skilled and unskilled workers. A second focuses on comparing performance in the same branch of industry under different training frameworks, some involving guild apprenticeship and some dispensing with it. A third approach focuses on the social distribution of human capital investment, assessing the two countervailing effects guilds had on skills acquisition: facilitating it for insiders (the fortunate minority of young men admitted to the guild) while blocking it for outsiders (such as women, minorities, and those too poor to afford the guild charges).
Intercontinental Comparisons of the Skill Premium
Some scholars argue that the benefits of guild training are demonstrated by the “skill premium”—the gap in wages between skilled craftsmen and unskilled labourers. This gap, they argue, was narrower in Europe than in Asia. The narrowness of the gap in Europe, the argument continues, was caused by superior European training institutions, specifically the fact that European guilds were better than or different from guilds in Asia. Epstein and Prak, for instance, contend that the “relatively low levels of skill premium in Europe, compared to East Asia and India, suggest that the corporate system of professional education must have been generally efficient”.232 Van Zanden argues that the Great Divergence between European and Asian economic growth in the early modern period was caused by European guilds providing skilled training, giving rise to a lower skill premium in Europe.233 The line of argument is as follows: the gap between the wages of skilled workers (craftsmen) and unskilled workers (labourers) was lower in Europe than in Asia; this was because Europe had more skilled workers as a proportion of the labour force than Asia did; this in turn was the product of the better training institutions in Europe; and the key component of Europe’s better training institutions was its guilds.
But does this argument make sense? Assume that the stylized fact is accurate, and that the gap between craftsmen’s and labourers’ wages was systematically narrower in Europe than in Asia. Can we assume this was because Europe had guilds while Asia did not? Probably not.
For one thing, training is not the only determinant of wages. Wage gaps between crafts and labouring are affected by the supply and demand for workers of different characteristics. Supply and demand in turn are influenced by occupational structure, technology levels, capital equipment, demography, institutional privileges, labour bargaining, gender discrimination, organization of firms, and innumerable other factors. This basic principle is illustrated by findings from the modern United Kingdom, where the proportion of 25- to 29-year-olds with university degrees increased from 13 per cent in 1993 to 41 per cent in 2015, but the median wage differential between graduates and school leavers in that age-group—the “skill premium”—stayed flat.234 The relative numbers of skilled and unskilled workers can thus differ greatly without changing the skill premium, which is influenced by a large number of other variables. So, even if the wage gap between skilled and unskilled workers was narrower in early modern Europe than Asia, this does not imply a higher quality or quantity of training in Europe.
Second, institutions are just one determinant of training outcomes. The amount of training workers get depends on numerous factors, notably the demand for trained labour in different occupations, the expected rewards of training, and the supply of experts to do the training. Labour markets convey signals about these factors to workers and employers through wages. The functioning of labour markets and the accuracy of wage signals can be affected by training institutions, but also by many other factors. Even if early modern Europe had more skilled workers than early modern Asia, this does not imply that the abundance of skilled labour was created by training institutions rather than by the many other labour market signals that influence training outcomes.
Third, guilds are only one of many institutions that affect training. As this chapter has shown, in many occupations training was provided by institutional mechanisms other than guilds: over half of all Italian guilds, after all, had no apprenticeship provisions. Many people obtained training in crafts and trades by concluding private, non-guild apprenticeships which were recorded and enforced by notaries, magistrates, municipal offices, courts of law, or other public institutions. Others learned crafts and trades in the household, from parents, spouses, siblings, or other relatives. Still others learned their occupations on the job, from employers, fellow-workers, and in work-gangs. The female half of the labour force was almost completely excluded from guild training, yet many women got vocational skills good enough that they were regarded by guild members as a serious competitive threat. Even if we could be sure that Europe had a lower skill premium because of its training institutions, we could not conclude that the key institutions were guilds, especially considering that large swathes of the workforce were actually denied guild training.
Finally, it may not be justified to regard the industrial labour force as guilded in Europe and non-guilded in Asia. In Europe, many crafts were practised in the non-guilded countryside from the late medieval period onwards, as in the dense textile zone of Hondschoote in Flanders; only in parts of central and southern Europe did rural industries also have guilds. Many urban crafts were practised in a non-guilded framework; only in some parts of Europe, as Chapter 9 discusses, were there virtually no urban centres without guilds. Even in guilded towns, there were many crafts without guilds (as was true of half the artisans in early modern Bordeaux)235 and many with skilled but non-guilded workers (like the highly skilled females and immigrants who were the backbone of the Venetian glass industry).236 European guilds would not have harassed women and other outsiders so relentlessly had these non-guilded workers not been skilled competitors. The skilled part of the European labour force was not comprehensively guilded.
Conversely, the skilled part of the Asian labour force was not comprehensively non-guilded. Guilds existed in a number of early modern Asian societies, including India, Japan, and China. Across Asian societies, as across European ones, the presence, strength, and activities of guilds varied greatly over space and time. A few European scholars have claimed that although Asian guilds existed, they did not behave like European ones.237 But this does not do justice to the scholarship on non-European guilds. Comparative studies of European and Chinese crafts, for instance, suggests that in many sectors of the Chinese economy guilds were strong and provided apprenticeships not dissimilar to those provided by European guilds.238
It is sometimes suggested that guilds in China differed fundamentally from those in Europe because Chinese guilds restricted access to apprenticeship according to family and kinship.239 However, this conjecture is based almost exclusively on nineteenth- and early-twentieth century ethnological descriptions which, on closer inspection, do not support the view that Chinese guild apprenticeships were family- or kinship-based.240 The first source for this view is MacGowan (1886), who declares that “[c]lannishness is a Chinese characteristic”. It then emerges, however, that “clannishness” does not refer to kinship but to geography: many Chinese guilds started as place-of-origin organizations that restricted membership to men from particular towns or regions. Some European guilds also started in this way, as we shall see shortly; and among Chinese guilds it was not universal.241 A second source often adduced to support the idea that Chinese guilds were kinship-based is Morse (1909), who states that as in medieval England, so too in China “some [guilds] allow none but sons and nephews of gild members to learn the trade”. However, Morse then goes on to describe as “very typical” the Wenchow guild of silk-weavers-and-dyers which limited family-based apprenticeship by ordering that “masters may have no more than one member of their own family learning the craft at one time”.242 A third ethnographic source is by Gamble (1921), who makes it clear that not all Chinese guilds limited membership by either kinship or geography. Not a single Peking guild did so, according to his study, and even in central and south China it was the usual practice only among guilds which “demand some particular skill, have special trade secrets, or whose work is especially remunerative”.243
The fourth source given for the kinship-based view of Chinese guilds is Burgess (1928, 1930). But Burgess’s detailed analysis explicitly refutes the idea that Chinese guilds or apprenticeships were kinship-based or fundamentally different from European guilds. He begins by emphasizing that “in no particular are [Chinese guilds] more close in their similarity to the medieval gilds of Europe than in the apprentice system”.244 He lists five conjectures about the origin of Chinese guilds, of which only one is that they developed out of “the family clan organization”, the others being that they originated in religious fraternities, geographical guilds, organizations to protect tradesmen from official exploitation, and organizations for sharing out scarce employment.245 Although in some parts of China towns traced their origins to an apical ancestor so that all families shared a surname, this did not mean that a town’s thousands of inhabitants were all close kin, since the common ancestor lay centuries in the past.246 Burgess makes it clear that apprentices did not typically learn from their fathers or close relatives, stating that “[w]hile some masters treated their apprentices with undoubted cruelty, others treated them as members of the family . . . apprentices, journeymen, and master would eat together at the same table”.247 Just one of the 42 Peking guilds Burgess studied, that of the porters, restricted admission to sons and brothers of guild members, but this policy was new and replaced apprenticeship altogether.248 Many Peking guild masters deliberately recruited their apprentices from rural areas because “Peking boys are not wanted because their families are near by and hinder their work”.249 In an account reminiscent of European guilds, Burgess describes how, “[d]uring the period of apprenticeship the master has full authority over the boy . . . taking the place of his father”.250 Chinese guild apprenticeships thus apparently substituted for kinship relationships rather than being based on them.
Just one recent case study is adduced in support of the idea that Chinese guilds were kinship-based. This is a study by Moll-Murata (2013) of a potters’ guild in Jingdezhen formed after 1674 by a group of migrants. This guild was subsequently described as including only “24 surnames”, which Moll-Murata interprets as showing that access to the occupation was restricted according to “family connections”.251 But as we saw in Chapter 3, pre-modern Europe also had groups of migrants who formed guilds which restricted access according to kinship. The Viennese chimney-sweeps’ guild, for instance, was founded in 1664 by nine masters originating from four neighbouring valleys of the Ticino and the Grisons; for the next two centuries, this guild limited apprenticeship and mastership to members of just 14 families.252 The 24 names in the Jindezhen potters’ guild are hardly adequate grounds for characterizing the Chinese guild system as family-based, any more than the 14 names in the Viennese chimney-sweeps’ guild should be used to characterize the European guild system as family-based. For Chinese crafts more widely, Moll-Murata describes guilds and extended families as distinct alternatives, with guild-free zones arising where industries “were managed entirely within extended families, so that no guilds were necessary or desired”.253 The available literature on Chinese guilds does not, therefore, provide comprehensive support for the idea that Asian guilds and guild apprenticeships were family-based and European ones were not.
Intercontinental comparisons of the skill premium thus permit no conclusions about guilds. Much more evidence would be needed on the determinants of wages, the determinants of training outcomes, the importance of different training institutions, and the training activities of guilds in different European and Asian economies, before it could be concluded that Asia was comprehensively inferior to Europe in this respect. In the current state of research, comparisons of the skill premium cannot be viewed as evidence that European guilds fostered high human capital levels, let alone that these caused the Great Divergence in economic growth between early modern Europe and Asia.
How the Same Industry Performed with Different Training Institutions
A second approach to evaluating the wider effects of guild training is to compare the same industry in different places with different training institutions. Many pre-modern industries had guild training in some European societies, non-guild training in others, and no formal training in still others. Comparing the same industrial activity in places with different training institutions can establish whether guild training and good industrial outcomes were associated, even if it cannot pin down a definitive causal link.
Textiles were by far the largest branch of pre-modern industry, so the performance of textile industries is central to assessing the efficacy of craft training.254 The largest and most successful textile sectors before the Industrial Revolution were linens and worsteds, which expanded rapidly from the late medieval period onwards, producing low-cost, fashionable products that appealed beyond traditional elites to mass markets, and exported wares across Europe, the Near East, Asia, and the Americas. But even smaller, luxury sectors such as high-quality woollens and silks participated to some extent in the “proto-industrialization” of European textile production from the fifteenth century onwards.255 In these textile sectors, fast growth and commercial success show no discernible association with strict guild training requirements. Most textile wares could be successfully produced and sold in export markets by producers who did not have formal guild training.
This emerges clearly from the large and rapidly growing European linen industry. Medieval and early modern linen regions ranged from the strongly guilded (Württemberg, Swabia), through the moderately guilded (Strasbourg, Bern), to the weakly guilded (Bohemia, Lusatia, Silesia), and the wholly non-guilded (Ireland). Even guilded urban linen-weavers did not all require apprenticeship: the Strasbourg linen-weavers’ guild did not impose a compulsory apprenticeship requirement until as late as 1484.256 But those urban linen industries that did require apprenticeship complained bitterly of being out-competed by rural cottage workers who had no formal training. In the Silesian town of Neumarkt in 1572, for instance, the linen guild lamented that untrained rural weavers were successfully depriving its members of merchant custom.257 In 1577, likewise, the Bern linen-weavers’ guild complained that huge numbers of rural weavers who had never undergone any apprenticeship were enjoying great entrepreneurial success by enticing customers away from the urban guild masters.258 Other successful linen industries were either wholly non-guilded from beginning to end (as in Ireland),259 or were primarily practised by enserfed rural cottage workers who worked part-time without having undertaken guild apprenticeships (as in Silesia and Bohemia).260 The Irish and Silesian linen regions were in fact the most successful in Europe, with no guild training requirements for their predominantly rural producers; the success and growth of these industries certainly surpassed that of the more strongly guilded linen industries of Württemberg or Swabia.261 The largest branch (linen) of the largest industrial sector in pre-modern Europe (textiles) was thus widely and successfully practised with no guild training.
Similar findings emerge for the worsted (and worsted-woollen hybrid) industry, where new entrants set up in business with little or no training—at least where guilds did not prevent them.262 In fourteenth- and fifteenth-century Flanders, the rural agglomeration of Hondschoote surpassed the moribund textile industries of the guilded Flemish towns, expanding rapidly and exporting its wares internationally without guilds or apprenticeship requirements, which were only introduced in the sixteenth century during Hondschoote’s decline.263 In the Württemberg Black Forest, an export-oriented worsted industry sprang up in the 1560s without any guilds or apprenticeships, and it expanded so successfully that in 1582 disgruntled urban guild masters complained that peasants and men of other crafts, “here and there also joined by women”, were setting up as worsted-weavers and selling in export markets with no guild training; only from c. 1590 onwards did urban male masters manage to organize guilds to limit entry, requiring a three-year guild apprenticeship which was closed to females and other outsiders.264 In the Thuringian principality of Schleiz, likewise, by 1592 smiths, goldsmiths, tanners, tailors, bakers, and shopkeepers were weaving worsteds in large quantities and exporting them through the Leipzig fairs; only later did weavers in the towns get guild privileges and impose training requirements, though these remained quite undemanding.265 In eighteenth-century Somerset, contemporaries observed that several thousand weavers were operating in and around Taunton, and “not Half of them have served Apprenticeships to the Weaving Trade”.266 In the West Riding of Yorkshire, eighteenth-century observers commented that “every man that wolde had libertie to be a clothier”,267 and in 1800 a witness to a Parliamentary inquiry declared that “nineteen out of twenty have not served regular apprenticeships in the textile industry”.268 Worsted production was thus another massive branch of the largest industrial sector in Europe where guild training was neither necessary nor sufficient for good economic performance: non-guilded medieval Hondschoote out-performed the guilded Flemish towns; non-guilded Yorkshire out-performed strongly guilded Württemberg.269
It might be argued that guild apprenticeships were unnecessary in low-quality, mass-market sectors such as linens and worsteds, but were critical in high-quality branches such as woollens or silks. But a remarkable array of successful, high-quality woollen industries casts doubt on this proposition. The Nuremberg woollen industry, for instance, enjoyed its golden age between the fourteenth and the mid-sixteenth century, yet during this period imposed no compulsory training.270 Only during its long stagnation after c. 1550 did the Nuremberg woollen-weavers’ guild begin to require a minimum three-year apprenticeship and a three- to four-year journeymanship, although still no examination or masterpiece.271 In medieval and early modern Florence, too, the powerful Wool Guild, dominated by merchants, had no mastership examination and ignored occasional agitation among weavers to restrict entry through apprenticeship, instead regarding “ex post controls [as] adequate to ensure the quality of the Florentine product”.272 The same was true in sixteenth- and early seventeenth-century Padua, where the wool guild did not define the scope of training, there were no guild examinations, masters and apprentices signed private training contracts, skill certification was provided by testimonials from other masters, and “the market established the level of skills”.273 The Kentish Weald was one of the most flourishing woollen broadcloth regions in Europe, exporting successfully to the continent throughout the sixteenth and early seventeenth centuries, on a wholly rural basis, with no guilds and no guild apprenticeships.274 Even the most skilled operations in the high-quality woollen industry, such as shearing and finishing, were successfully practised by people excluded from guild training, as in 1663 when the Aachen shearers’ guild protested against non-guilded shearing and finishing in the neighbouring hamlet of Burtscheid, notably by a female Anabaptist whose woollen press the guild was intent on destroying as a serious competitive threat.275 By the eighteenth century, the shortage of journeymen led the Aachen town council to overrule the shearers’ guild by opening the labour market to non-guild-trained journeymen shearers, who were snapped up as workers despite their lack of guild training.276 Across the woollen sector as a whole, therefore, we observe the Kentish Weald flourishing on a wholly non-guilded basis, while Florence and Padua flourished with guilds but no mandatory guild training. Strongly guilded Nuremberg, by contrast, experienced its golden age without mandatory guild apprenticeship, while Aachen’s industry survived only by outsourcing production to non-guild-trained workers.
In the silk sector, too, industries flourished without requiring their practitioners to get guild training. In late-fourteenth- and early-fifteenth-century Venice, for instance, the silk industry employed some 200 non-apprenticed female silk-winders; only after 1410, in response to an industrial downturn, did the guild impose a three-year apprenticeship as an entry barrier.277 Vicenza developed a high-quality, export-oriented silk industry in the fifteenth and sixteenth centuries in an entirely non-guilded context, attracting and training a specialized workforce skilled enough to operate hydraulic silk-throwing mills, some of the most complex mechanical devices of the era.278 The Krefeld silk industry in the German Rhineland achieved phenomenal success from the late seventeenth century through to the early nineteenth with on-the-job training and no guild apprenticeship. When the Prussian government proposed in 1766 that Krefeld adopt the Berlin silk ordinance with its minimum apprenticeship requirements, the Krefeld manufacturers petitioned “in all graciousness to spare us this and to maintain us further in the freedom we have hitherto enjoyed”.279 When a new Prussian ordinance again sought to impose an examination regime in 1849, the Krefeld masters stated uncompromisingly that each of them was aware from his own knowledge “whether his journeyman is capable of being a master or not”.280 In early modern Lyon, the silk industry was regulated by a “sectoral” guild combining merchants and craftsmen: the craftsmen members enforced guild apprenticeships but were constantly complaining that street hawkers were willing to accept “goods made by simple garçons en chambres [lit. “boys in rooms”] who had not even passed an apprenticeship”.281 The non-guilded silk industry of the seigneurie of Avignon and the surrounding Comtat Venaissin was regarded as a competitive threat by the silk guilds of Lyon, whose customers it easily attracted even though “apprenticeships are not required in the Comtat”.282 Lyon flourished as a silk centre partly because, although its guilds enforced apprenticeship requirements, merchants and consumers eagerly purchased from non-guilded suppliers, “indifferently from all hands without asking”.283
In non-textile branches, as well, producers who had not completed—or even begun—guild apprenticeships were skilled enough to please employers and customers. One example is the making of silvered and gilded wire in the style of Lyon. This industry was introduced into the region around Nuremberg in southeastern Germany from 1569 onwards by a religious refugee from Lyon, Anthoni Fournier, who trained successors in a number of different localities that, because the region was territorially fragmented, were subject to different institutional systems. One was Nuremberg itself, where Lyon-style wire-drawing was practised successfully for 130 years without a guild organization. Only in 1696 did the Nuremberg wire-drawers finally obtain guild privileges, confirmed and extended in 1719, at which point the Nuremberg wire industry entered a long decline.284 Another southeast German centre of Lyon-style wire-drawing was the small market-town of Freystadt, governed first by the Electors of the Palatinate and later by Bavaria. Here, too, the industry was initially non-guilded, and Fournier passed on his skills without the formality of apprenticeship, as shown by a conflict which arose in the 1650s when one of the chief Freystadt wire-makers proved unable to display any apprenticeship certificate even though he had been taught by the old master Fournier himself. It was only under pressure from other Freystadt guilds that Lyon-style wire-drawing was granted guild privileges in that town in 1657, in an ordinance which permitted the training of apprentices and journeymen exclusively to a privileged group of “elder masters”, whose maximum number was fixed at six.285 However, this decision was clearly dictated by politics, and not by the need for better training institutions, since for more than half a century the wire-drawers had transmitted their skills to a number of successful practitioners in Freystadt, without formal apprenticeship, let alone guild regulation.
Outside the textile and metal sectors, we repeatedly observe the same lack of association between industrial success and guild training, even in skilled activities such as printing, painting, glassmaking, medicine, shipbuilding, and precision-instrument making. In sixteenth-century Toledo, the heretical publications prosecuted by the Inquisition were printed by a torrent of immigrants from France, Savoy, and Germany, attracted by the lack of printers’ guilds which meant that they could find work without serving full apprenticeships or surmounting the onerous entry barriers of printers’ guilds back home.286 In seventeenth-century Madrid, Diego Velázquez complied with the painters’ guild’s ban on slave apprentices by refusing formal training to his long-time mulatto slave Juan de Pareja, who nonetheless “was so clever that, behind his master’s back and by cheating himself of sleep . . . [he] came to do works of painting worthy of great esteem”.287 In seventeenth-century Seville, likewise, a mulatto slave owned by Bartolomé Esteban Murillo was barred from apprenticeship by the painters’ guild but “sought to imitate his master in his spare time after having served his master, whereby he became an accredited painter, who displayed a good taste in color, a mastery of union in his canvases and sufficient exactitude in his use of drawing”.288 In seventeenth- and eighteenth-century Venice, the glassmaking industry only maintained its international position against technically superior French, English, and Bohemian competitors thanks to tens of thousands of female workers who congregated in glass-bead-making and by 1750 vastly outnumbered guild-trained male workers. As Francesca Trivellato points out, “women were the sole bearers of skills transmitted outside any formal apprenticeship and thus indispensable to the large urban domestic industry of glass bead making”.289 In shipbuilding from the 1670s onward, the fastest growing industries in Europe (those of England and the Zaanstreek of Holland) operated completely outside a guilded framework, while the German industry with its compulsory guild apprenticeships stagnated.290 In precision-instrument-making, the London industry was the European leader, despite its lack of compulsory guild training, far surpassing the Paris trade with its comprehensive guild apprenticeship requirements.291 Even in medicine, many patients preferred to patronize non-guilded encroachers, often females who could not have got guild training even if they had wanted it, much to the fury of the surgeons’ guilds of eighteenth-century Frankfurt292 and Chateaubriand.293
Pre-modern European crafts and trades, including many of those requiring the greatest skill, thus provide plentiful evidence that guilds were neither necessary nor sufficient for providing training at levels sufficient to attract customers and foster industrial dynamism. Comparisons of performance across different institutional regimes show no association between industrial success on the one hand and guild-mandated training on the other.
The Social Distribution of Human Capital Investment
A third approach to assessing the aggregate impact of guild training is to examine how it affected the social distribution of human capital. Evidence from modern developing economies suggests that human capital investment has the greatest economic impact among those with the lowest pre-existing levels of training, particularly females, migrants, peasants, and the poor. This implies that the institutions that best foster economically relevant skills in poor societies are not those that add to the capacities of the better off, but rather the ones that open access to training for disadvantaged groups. In particular, as Chapter 5 discussed in detail, women’s human capital has a large effect on economic performance. This is partly because women are typically denied training in traditional economies, with the result that the marginal return from such investments are high. Partly, too, it is because maternal human capital is a major determinant of the human capital of offspring. Institutions that stifle economically relevant human capital investment by females are likely also to stifle economic growth.294
Against this background, how do we assess the impact of guilds on aggregate human capital investment? A guild invested in, and probably often increased, the human capital of the young men it admitted to apprenticeship. But a guild also had the exclusive right to decide who was allowed to be apprenticed to that occupation. Guild masters had a local monopoly not only over producing particular goods and services, but also over producing new practitioners by taking on apprentices. This gave guilds an incentive to restrict the supply of apprenticeships and the associated human capital, so as to increase the price that could be charged. And that is what they did.
Most guilds, as Chapter 5 showed, excluded females from apprenticeship. This was not because women were inherently unable or unwilling to practise crafts: many girls wished to invest in vocational training in crafts and commerce, their families were willing to pay the fees, and the girls were able to learn and practise the trade. Furthermore, many pre-modern European societies females could expect to spend part or all of their lives outside marriage, as either spinsters or widows. Secondary- and tertiary-sector occupations were, moreover, well suited to women’s physical endowments and to any domestic responsibilities they might have. Nonetheless, females were almost completely excluded from guild apprenticeship for institutional reasons: guilds that admitted girls to apprenticeship were extremely rare and became even rarer in the course of the early modern period. As a German jurist stated in 1780, “no female may properly practise a craft, even if she understands it just as well as a male person”.295 What mattered was not that one “understood” the craft, but that one belonged to an identifiable group which the guild could justify excluding.296 Young women were eager to invest in occupation-specific human capital, as shown by the large number of female apprentices in places and occupations where access to training was not regulated by guilds. But women were largely prevented from entering formal training in any occupation or place in which a guild had the legal right to decide who could and could not be admitted to apprenticeship.
Guilds did not just restrict the supply of occupational human capital to the 50 per cent of people who were female. They also, as Chapter 3 showed, restricted its supply to the many young men who wanted to undertake apprenticeships in guilded activities but could not surmount the entry barriers for one or more reasons: they were immigrants; they could not afford the guild fees; they were illegitimate; they were serfs from the countryside; they held the wrong religious beliefs; they spoke the wrong language; their skin was the wrong colour; their parents or grandparents practised a defiling occupation; the existing guild masters did not like them. These young men were often eager to invest in their own human capital, but were prevented from doing so by the fact that a guild had the local monopoly over training in that occupation. Contemporaries recognized that guilds restricted young men from investing in their human capital, as a landmark English legal judgment of 1614 makes plain. When the Ipswich tailors’ guild tried to close down the business of a man who was working as a tailor without having served his apprenticeship in the guild, the justices of King’s Bench decided against the guild on the grounds
[that under] the common law, no man could be prohibited from working in any lawful trade, for the law abhors idleness, the mother of all evil . . . and especially in young men, who ought in their youth (which is their seed time) to learn lawful sciences and trades, which are profitable to the commonwealth, and whereof they might reap the fruit in their old age, for idle in youth, poor in age; and therefor [sic] the common law abhors all monopolies, which prohibit any from working in any lawful trade.297
Conversely, where guilds continued to control entry, contemporaries recognized that a guild apprenticeship was not a good investment for a poor youth, since he was unlikely to surmount guild entry barriers without the financial resources of a middle-class family or inheritance. Parish administrators in early modern France, for instance, refused to allocate welfare funds to place poor youths in guild apprenticeships, because they perceived the guild system as closed to outsiders and the outcome of guild apprenticeship as too doubtful for those without private means. The welfare authorities of Saint Eustache, for instance, wrote in 1684 that “the fees for mastership and apprenticeship being high, and the utility of this help being too uncertain to be undertaken by the [welfare authorities], no [charity] tickets will be given . . . to put anyone in apprenticeship”.298 The contrasting situation in the Northern Netherlands, where orphanages deliberately sought to place their charges in guild apprenticeships, was made possible by urban governments which negotiated with guilds or overruled normal guild entry barriers by declaring orphans to be exempt from paying apprenticeship fees and other contributions.299 But even in the Netherlands, guilds restricted the number of young men who could invest in occupation-specific human capital. In 1766, for instance, the Leiden cloth-shearers, one of the few textile occupations in Leiden with a guild organization, complained that the limit the guild placed on the number of apprentices each master could have was preventing their trade from expanding.300 It is thus unsurprising that even in the Netherlands, the abolition of the guilds noticeably increased access to craft apprenticeships for orphans by removing guild ceilings on apprentice and master numbers.301 The guild system restricted investment in occupation-specific human capital even by poor males.
Guilds were thus precisely the wrong sort of institution to increase human capital investment in a developing economy: in the conceptual categories established in Chapter 1, they were particularized rather than generalized institutions. They added to the capacities of a better-off minority while limiting training for the less advantaged majority, including social groups with the lowest pre-existing levels of human capital: females and poor males. Guilds thus compelled a limited group of young men to undertake more years of formal vocational training than they actually needed, and prevented a much larger group of young men and women from obtaining any such training at all. Guilds might have benefited human capital for the privileged few, but they harmed it for the disadvantaged many.
CONCLUSION
Did guilds promote human capital investment in pre-modern economies? The findings in this chapter suggest they did not. In theory, cartel privileges might have created enticing rents and efficient monitoring, giving guilds the incentive and capacity to solve market failures in the provision of training. But cartel privileges also gave guilds incentives to exploit their control of training to profit their own members at the expense of the rest of society.
On the one hand, guilds did not guarantee skills well, failing to exploit their theoretical advantages in correcting failures in training markets. Many guilds included no provisions for apprenticeships in their statutes: in the largest available sample, covering Italy over five centuries, 60 per cent of guilds did not even mention training in their ordinances. Many guilds did not require any examination or masterpiece, so they did not seriously certify skill. Those guilds with examinations often failed to define their content and most resisted outside pressure to clarify or improve their test regimes. Some guilds imposed a detailed examination but left its requirements unchanged for centuries, during which techniques, equipment, and commercial practices changed in fundamental ways. Still other guilds administered examinations corruptly or openly passed unqualified candidates who possessed the requisite personal ties, political clout, or financial means. Many guilds also failed to realize their potential advantages in preventing opportunism between trainer and trainee. As organizations to promote the interests of masters, guilds turned a blind eye to masters who failed to provide training and exploited apprentices as cheap general labourers. This in turn encouraged apprentice opportunism, particularly high non-completion rates, a suggestive indicator of the failure or economic irrelevance of guild training. Institutions that could solve the moral hazard problem in the master-apprentice relationship might be key determinants of human capital investment and economic growth, but guilds did not succeed in solving this problem and often did not even try.
Guilds cannot have been the efficient institutions for vocational training. Alternative institutional mechanisms to facilitate training existed alongside guilds from the earliest period and were widely preferred to them. Private apprenticeships were widespread in guilded and non-guilded occupations alike, wherever guilds did not prohibit their use. Revealed preferences of apprentices and masters indicate that notarial, legal, municipal, and state institutions provided acceptable protections against opportunism in training contracts from the medieval period onwards. That guilds were not necessary for providing skilled training is shown by the fact that many mainstream crafts were practised successfully by females whom guilds, as we saw in Chapter 5, excluded from obtaining guild training. Females and other outsiders lacking formal guild training were attacked as a serious competitive threat in most guilded crafts, indicating the adequacy of their skills in the real economy.
Guilds did, however, use their local monopolies over providing training in their occupations to extract rents for their members. They imposed formal apprenticeship as an entry condition in a large number of occupations generally recognized as not highly skilled, including labouring, farming, cleaning, selling, carrying burdens, and coal-picking. Even in mainstream crafts, guilds often mandated apprenticeships longer than were needed to learn the occupation, a practice that cannot be explained as a bond on apprentice opportunism given that most new apprentices were productive enough to cover their consumption costs and alternative monetary and legal bonds were available. Rather, excessively long apprenticeships increased barriers to entry. Guilds also used apprenticeship requirements as entry barriers in other ways by denying training to otherwise capable applicants (such as Jews and women), charging high fees to those they admitted to training, exempting masters’ relatives known to lack expertise, and selling exemptions from training requirements to completely untrained persons. Thus many guilds administered their training regulations, where they had them, to serve as barriers to entry, pretexts for other privileges, or licenses to be sold, rather than to increase occupation-specific human capital. This casts doubt on romantic views that guild apprenticeships were superior to non-guild ones and benefited economic growth because they ensured dissemination of knowledge.
It is therefore unsurprising that guild apprenticeship requirements failed to improve economic performance. Claims that the skill premium was lower in Europe than Asia tells us nothing about guilds, since wage differentials are caused by many labour market characteristics other than skills, skills are determined by many factors other than training institutions, training was provided by many other institutions than guilds, and too little is known of Asian guilds to conclude that they provided inferior training to European ones. Comparisons across European industries, by contrast, show that guild training was neither necessary nor sufficient for high productivity and rapid growth. In all major branches of manufacturing, industries which were either completely non-guilded or whose guilds did not mandate apprenticeship outstripped guilded industries with mandatory apprenticeships.
Guilds did not administer a training system open to all capable applicants. Instead, to benefit their members, they decided who was allowed to enter training, and restricted the supply. Guilds denied apprenticeship to large groups in society based purely on their identity—on considerations that had nothing to do with their individual capacity to learn the skills involved. The groups that guilds excluded from training included not just almost all females, but also large numbers of males—Jews, bastards, gypsies, former serfs and slaves; most members of other religions, ethnicities, and nationalities; those without the right parentage in the guild or community; those with an ancestor who had practised a “defiling” occupation; and anyone who could not afford the entrance fees. Guilds enabled a privileged minority to obtain vocational skills but excluded large numbers of unprivileged men, and nearly all women, from investing in their own human capital. It is therefore unsurprising that guild activities in the sphere of human capital show no positive relationship with economic outcomes.
Epigraph sources: Paris dyers’ guild, Depping 1837, 402; Iglau woollen-weavers’ guild, Werner 1861, 30; Venice wood-carvers’ guild, Rapp 1976, 43; Report to Governor of Upper Austria, Commenda 1959, I:111; Nuremberg apprentice’s complaint, Grießnger and Reith 1986, 166.
1 Mokyr 2009; Allen 2009; McCloskey 2010.
2 For a survey, see Ogilvie and Küpker 2015, 21–24.
3 Dennison and Ogilvie 2014, 680.
4 See, e.g., Jacob 2014, 185–219; Squicciarini and Voigtländer 2015.
5 See, e.g., Humphries 2003, 2013; Minns and Wallis 2012, 2013.
6 See De Munck, Kaplan, and Soly 2007, 14–15; and Zeev, Mokyr, and Van der Beek 2017.
7 Epstein 1998; Epstein 2008; Pfister 2008 [Craft Guilds, Theory], 26; Epstein and Prak 2008, 4, 23; Prak and Van Zanden 2013.
8 Epstein 1998, 688–92; Pfister 1998, 14, 18; Lucassen, De Moor, and Van Zanden 2008.
9 Pfister 2008 [Craft Guilds, Theory], 27.
10 Epstein 1998, 691–92.
11 Epstein 1998, 687, 690–92, 706.
12 Quoted in MacKay 2006, 133.
13 Epstein and Prak 2008, 6.
14 Depping 1837, 402.
15 Quoted in Werner 1861, 30.
16 Quoted in Wesoly 1985, 249.
17 Rule 1987, 107.
18 Wiesner 1991, 777.
19 Hafter 1995, 44.
20 Vainio-Korhonen 2000, 62–63.
21 Sonenscher 1987 [Mythical], 31–32.
22 Ehmer 2001 [Artisans], 820–21.
23 The medieval period accounts for 32.4 per cent of observations in Table 7.2, compared to 27.2 per cent in the guilds database, a difference that is of borderline statistical significance (it is significant at the 0.0724 level).
24 As a reminder, throughout this book, “significant” means that the null hypothesis of no difference cannot be rejected at the 0.05 level.
25 Crews 2005–2006, unpag.
26 On brickmakers’ guilds in Italy and the Northern Netherlands, see Kessler and Lucassen 2013, 271, 284–85; on brickmakers’ guilds in the German territory of Württemberg, see Hoffmann 1905, 11.
27 Hoffmann 1905, 12 n. 4.
28 Kessler and Lucassen 2013, 284–85.
29 Kessler and Lucassen 2013, 270–85.
30 Kessler and Lucassen 2013, 284.
31 Ben-Amos 1991 [Failure], 155; Wallis 2008, 844; Davies 1956, 259–67.
32 MacKay 2006, 134.
33 Reyscher 1828ff, vol. 13, 1178, 1180; Hoffmann 1905, 10, 12 n. 1.
34 Wagner 1987; Ehmer 2000, 200; Steidl 2007, 142.
35 Marsh 2006, 123–24.
36 Marsh 2006, 126.
37 Marsh 2006, 127.
38 Marsh 2006, 132–33.
39 Hafter 2007, 153.
40 Epstein and Prak 2008, 6,
41 Epstein 2008, 165–66.
42 Epstein 2008, 166 with nn. 51–53.
43 Quoted in MacKay 2006, 133.
44 Epstein and Prak 2008, 8, say that this is why apprenticeship length varied so much across guilds in the same occupation.
45 Schulz 1985, 248; Reith 2007, 183.
46 For two occupations, means could not be compared because no observations were available for the medieval period.
47 Quoted in Wesoly 1985, 54.
48 Wesoly 1985, 54.
49 Wesoly 1985, 54.
50 Davies 1994, 196–97; Davies and Saunders 2004, 55.
51 Quoted in Ben-Amos 1991 [Failure], 170.
52 Ben-Amos 1991 [Failure], 170.
53 McClendon 2015, 364, 366, 368–69.
54 Vainio-Korhonen 2000, 61.
55 Steidl 2007, 149.
56 Kluge 2007, 156.
57 Aubin 1942, 160
58 Troeltsch 1897, 10–11; Ogilvie 1997, 140.
59 Caracausi 2014, 134, 147–50, 156.
60 Musgrave 1990, 49–50.
61 Quoted in Lipson 1965, 60.
62 Heaton 1965, 310–11.
63 Smith 1776, vol. I, x, c, 5.
64 Epstein 1998, 687, 690–92, 706; Epstein 2008, 165.
65 Wallis 2008, 836–37.
66 For early modern Germany, see Ogilvie 2003, 99–102; for the early modern Netherlands, see Van Nederveen Meerkerk and Schmidt 2008, 719; for eighteenth-century Turin, see Zucca Micheletto 2014, 147–48.
67 All averages reported are unweighted since information is not available on the size of the samples.
68 Quoted in Rapp 1976, 43.
69 Nicholas 1995, 1109.
70 Reith 2007, 190.
71 Nicholas 1995, 1123.
72 Quoted in Davies and Saunders 2004, 59. Wage rate from Allen n.d., London database; master building craftsman’s day wage is 8d.
73 Wesoly 1985, 70.
74 Epstein 1988, 120.
75 Michaud-Fréjaville 2005, 29 (Table 11).
76 Wallis 2008, 847.
77 Colavizza 2017, 1–2, 6–8.
78 Reith 2007, 190.
79 Schalk 2014, 35–38; Schalk 2017, 741–45.
80 Epstein 1988, 127, 129.
81 Nicholas 1995, 1117; Kluge 2007, 155.
82 Stabel 2007, 163.
83 Epstein 1991, 110.
84 Kluge 2007, 154.
85 Kluge 2007, 155.
86 Schalk, Wallis, Crowston, and Lemercier 2017, 135.
87 Epstein 1991, 78.
88 Reith 2007, 192.
89 Quoted in Werkstetter 2004, 171.
90 Werkstetter 2004, 175.
91 Quoted in Werkstetter 2004, 176.
92 Quoted in Martinat 2011, para. 9.
93 Quoted in Werkstetter 2004, 173.
94 Anishanslin 2016, 28; Thunder 2004; Rothstein 1994, 111.
95 Quoted in Werkstetter 2004, 166.
96 Quoted in Werkstetter 2004, 166.
97 Quoted in Werkstetter 2004, 167.
98 Quoted in Lanza 2007, 103.
99 Werkstetter 2004, 169.
100 Lanza 2008, 74.
101 Ogilvie 2004 [Guilds], 303–304 (Table 3); Ogilvie 2003, 260.
102 For examples of such arguments, see Epstein 1998, 687, 690–92, 706; Van Zanden 2004; Van Zanden 2009; Pfister 1998, 14, 18; Lucassen, De Moor, and Van Zanden 2008.
103 Epstein 1991, 77–78.
104 Wallis 2008, 851; Elbaum 1989; Elbaum and Singh 1995.
105 Nicholas 1995, 1107 (quotation); Epstein 1991, 77.
106 Bednarski and Courtemanche 2009, 114–15, 128.
107 Gestwa 1999, 466–67.
108 Gestwa 1999, 467 with n. 72.
109 Snell 1985; Simonton 1991; Lane 1996.
110 Epstein 2008, 161 n. 26.
111 Davids 2003, 7 (quotation); Davids 2007, 71–72.
112 Davids 2007, 69.
113 Nicholas 1995, 1107; Stabel 2007, 163.
114 Loats 1997, 17–18.
115 Hilaire-Pérez 2007, 138.
116 Mocarelli 2006, 10; Caracausi 2014, 149–50; Clemente 2014, 323–24 (quotation).
117 Nicholas 1995, 1125.
118 Zucca Micheletto 2014, 131; Mackenney 1997, 38; Colavizza 2017, 2.
119 Epstein 1991, 76, 108.
120 Wallis 2012, 795–803; Epstein 1991, 108.
121 Pfister 2008 [Craft Guilds, Theory], 26.
122 Unger 2013, 185.
123 On the practical and rhetorical context within which guild ordinances were promulgated, see Ehmer 1998 [Traditionelles], 39; Ogilvie 1997, 39–45; Ogilvie 2004 [Guilds], 292–93.
124 Nicholas 1995, 1107.
125 Menjot 2002, I:442.
126 Davids 2007, 69.
127 Quoted in Lipson 1915, I:292, 295.
128 Archer 1991, 114–55; Carlin 1994, 226.
129 Caracausi 2014, 134, 147–48, 150, 156; Caracausi 2017 [Reassessment], 88.
130 Wesoly 1985, 67.
131 Ogilvie 1997, 175–78.
132 Schalk, Wallis, Crowston, and Lemercier 2017, 147–48.
133 Ben-Amos 1991 [Failure], 155; Wallis 2008, 844; Davies 1956, 259–67.
134 Hill 2002, 30.
135 As is argued in Lucassen, De Moor, and Van Zanden 2008, 16.
136 Saint-Léon 1922, 553, 552, 554, 556.
137 Nicholas 1995, 1126.
138 See, e.g., Grießinger and Reith 1986, 153; Lombardo 2001, 105–106.
139 Fischer 1966, 62.
140 La Force 1965, 93 n. 22.
141 Pestalozzi 1829, 100.
142 Poni 1989, 97.
143 Poni 1989, 101.
144 Manninen 1984, 160–61.
145 Faivre 1949, 170.
146 Destouches 1809, 263.
147 Grüll 1953, 282.
148 Kaplan 1981, 271; see Rosenband 2000, 99–104, on journeymen’s wages.
149 Dell’Orefice 1998, 124, 127.
150 Kaplan 1981, 263.
151 Kaplan 1986 [The Character], 640.
152 Kaplan 1986 [The Character], 640–41.
153 Epstein 1998, 687, 690–92, 706.
154 Schalk, Wallis, Crowston and Lemercier 2017, 140.
155 See Rappaport 1989, 313; Davies 1994, 195.
156 Schalk, Wallis, Crowston and Lemercier 2017, 140, 142.
157 Ben-Amos 1991 [Failure], 170; Davies 1994, 196–97; Davies and Saunders 2004, 55.
158 Schalk, Wallis, Crowston and Lemercier 2017, 135, 142.
159 Quoted in Caracausi 2017 [Reassessment], 106.
160 Quoted in Commenda 1959, 111.
161 Quoted in Grießinger and Reith 1986, 166.
162 Schalk, Wallis, Crowston and Lemercier 2017, 144–45.
163 Minns and Wallis 2012; Schalk, Wallis, Crowston and Lemercier 2017, 146–48.
164 Collantes de Terán Sánchez 1993, 98.
165 Caracausi 2017 [Reassessment], 108.
166 Davies and Saunders 2004, 40.
167 Wallis 2012, 800, 802.
168 Kisch 1989, 163.
169 De Kerf 2014, 249–52.
170 Reith 2007, 192–93, 199 n. 100.
171 Marquardt 1975, 57–58.
172 Ben-Amos 1991 [Women], 233 (on use of apprentices for housework); Ben-Amos 1991, 157 [Failure]; and Wallis 2008, 839 (on non-completion rates).
173 Stabel 2007, 161.
174 Quoted in Grießinger and Reith 1986, 166.
175 Kaplan 1981, 284.
176 Furió 1996, 32 with n. 28.
177 Quoted in Reith 2007, 193.
178 Commenda 1959, 111; Reith 2007, 192.
179 Ogilvie 2004 [Guilds], 311.
180 Molas Ribalta 2002, 220.
181 Quoted in Reith 2007, 192–93, 199 n. 100.
182 Farr 1997, 66.
183 Quoted in Dupré 2014, 144.
184 Wesoly 1985, 75; Kluge 2007, 161.
185 Reyscher 1828ff, vol. 12, 969.
186 Davids 2007, 67.
187 Rasterhoff 2012, 118; Rasterhoff 2014, 182.
188 Epstein 1991, 125.
189 Maitte 2014, 37, 51–52; Trivellato 2006, 159.
190 Maitte 2014, 37.
191 Farr 1988, 21 n. 35.
192 Rodocanachi 1894, 2:249.
193 Caracausi 2014, 134, 147–50, 156; Caracausi 2017 [Reassessment], 88.
194 Groiss 1980, 58.
195 Unwin 1908, 347.
196 Prak 2008, 155, 157.
197 Unger 1978, 76, 186, 194–95; Unger 2013, 186.
198 Davids 2008, 380.
199 Davids 2007, 67; Davids 2008, 380–81.
200 Davids 2008, 381.
201 MacKay 2006, 43–44.
202 Quoted in Wesoly 1985, 249.
203 MacKay 2006, 62–63.
204 Horn 2006, 39.
205 Quoted in Anon 1866 [Fortsetzung], 108 n. 196.
206 Quoted in Leonhard 1909, 749–50.
207 Berlepsch 1852, vol. 7:76–77.
208 Farr 1988, 49.
209 Wesoly 1985, 250.
210 Cerutti 2010, p. 591 n. 53 (Turin journeyman shoemaker’s daily wage of 18 sous, in a currency where a lira contained 20 sous (soldi)); 603 n. 87.
211 Quoted in Aclocque 1917, 42 n. 7.
212 Aclocque 1917, 42.
213 Finkenwirth 1910, 68.
214 Mackenney 1997, 42.
215 Molas Ribalta 2002, 221.
216 Carcausi 2014, 150.
217 Poni 1989, 97, 100 (quotation).
218 Dell’Orefice 1998, 124, 127.
219 Quoted in MacKay 2006, 41.
220 Sabbatini 1998, 241 n. 50.
221 Poni 1989, 103.
222 Larmour 1967, 479.
223 MacKay 2006, 41.
224 Hafter 1997, 5–6. See Rosenband 2000, 99–104, on journeymen’s wages c. 1780.
225 Quoted in Groiss 1980, 67–68, 78.
226 Wesoly 1985, 255.
227 Bode 2002, 15.
228 Lombardo 2001, 107
229 Kluge 2007, 238–39.
230 Thoß 2012, 102.
231 Hauptstaatsarchiv Stuttgart, A573, Bü. 100, 1793, fols. 22v–23r.
232 Epstein and Prak 2008, 11.
233 Van Zanden 2009, 122.
234 Blundell, Green, and Jin 2016, 3–5 (esp. Fig. 3).
235 Horn 2015, 30.
236 Trivellato 2006, 163; Trivellato 2008, 217.
237 Van Zanden 2009, 140.
238 See, for instance, MacGowan 1886, 175, 178–79; Morse 1909, 31–34; Burgess 1928, 71; Gamble 1921, 31; Moll-Murata 2013, 234.
239 Lucassen, De Moor, and Van Zanden 2008; Prak and Van Zanden 2013, 15.
240 The four classic references are MacGowan 1886; Morse 1909; Gamble 1921; and Burgess 1928.
241 MacGowan 1886, 181 (Chinese “clannishness”); Ogilvie 2011, 15, 23–30 (European alien merchant guilds as place-of-origin organizations).
242 Morse 1909, 33.
243 Gamble 1921, 168.
244 Burgess 1930, 75.
245 Burgess 1928, 70–71.
246 Burgess 1928, 71.
247 Burgess 1930, 75.
248 Burgess 1928, 84, 124, 129.
249 Burgess 1928, 156.
250 Burgess 1928, 15.
251 Moll-Murata 2013, 234.
252 Ehmer 1997, 184.
253 Moll-Murata 2008, 244 (quotation); see also Moll-Murata 2013, 238.
254 Soly 2006, 2; Lis and Soly 2008, 95.
255 Jenkins 2003; De Vries 1976.
256 Kluge 2007, 151.
257 Aubin 1942, 160.
258 Bog 1968, 78.
259 Boldorf 2006.
260 Cerman 1996.
261 For a comparative consideration of Silesia and Ireland, see Boldorf 2009; on Württemberg see Medick 1996.
262 Ogilvie 2004 [Guilds], 303.
263 Coornaert 1930, 28.
264 Troeltsch 1897, 10–1
265 Finkenwirth 1910, 30–1.
266 “Petition of serge, worsted and other woollen weavers”, Journals of the House of Commons, 4 Dec. 1702 [http://www.bopcris.ac.uk/bop1688/ref1506.html].
267 Quoted in Heaton 1965, 102.
268 Quoted in Heaton 1965, 310–11.
269 Ogilvie 2004 [Guilds]; Heaton 1965; Lipson 1965.
270 Bog 1968, 72.
271 Wiest 1968, 192–96 (Tab. F).
272 Ammannati 2014, 72–73.
273 Caracausi 2014, 134, 147–50, 156.
274 Zell 1994.
275 Kisch 1989, 165.
276 Ebeling 2004, 126–27.
277 Trivellato 2008, 213.
278 Demo 2014, 85.
279 Quoted in Kriedte 1993, 266.
280 Quoted in Kriedte 1993, 266.
281 Quoted in Hafter 2007, 140.
282 Quoted in Horn 2015, 53.
283 Quoted in Hafter 2007, 140.
284 Braun and Burger 2008, 164–65.
285 Braun and Burger 2008, 166.
286 Griffin 2005, 77–78.
287 Quoted in Rodriguez 2015, 133.
288 Quoted in Rodriguez 2015, 134.
289 Trivellato 2006, 163 (quotation); Trivellato 2008, 217.
290 Unger 1978, 64, 66, 82, 84–85, 110–12, 116.
291 Daumas 1953, 124ff.
292 Brandt 2003, 12.
293 Locklin 2007, 60.
294 World Bank 1998, 45–46, 52, 55; Knowles et al. 2002.
295 Weisser 1780, 99–100.
296 For a detailed discussion of how arbitrarily distinguishing identifiable groups helped guilds reduce competition, see Stuart 1999, esp. 189–221.
297 Quoted in Siegan 2001, 23–24.
298 Quoted in Crowston 2007, 54.
299 Davids 2007, 74.
300 Schalk 2015, 70.
301 Schalk 2017, 749–54; Schalk 2015, 71–75.