Top Banner
Economic History Association What Do Bosses Really Do? Author(s): David S. Landes Reviewed work(s): Source: The Journal of Economic History, Vol. 46, No. 3 (Sep., 1986), pp. 585-623 Published by: Cambridge University Press on behalf of the Economic History Association Stable URL: http://www.jstor.org/stable/2121476 . Accessed: 04/01/2012 14:25 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Cambridge University Press and Economic History Association are collaborating with JSTOR to digitize, preserve and extend access to The Journal of Economic History. http://www.jstor.org
40
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Landes - What Do Bosses Really Do

Economic History Association

What Do Bosses Really Do?Author(s): David S. LandesReviewed work(s):Source: The Journal of Economic History, Vol. 46, No. 3 (Sep., 1986), pp. 585-623Published by: Cambridge University Press on behalf of the Economic History AssociationStable URL: http://www.jstor.org/stable/2121476 .Accessed: 04/01/2012 14:25

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

Cambridge University Press and Economic History Association are collaborating with JSTOR to digitize,preserve and extend access to The Journal of Economic History.

http://www.jstor.org

Page 2: Landes - What Do Bosses Really Do

THE JOURNAL OF ECONOMIC HISTORY

VOLUME XLVI SEPTEMBER 1986 NUMBER 3

What Do Bosses Really Do? DAVID S. LANDES

If employers make so much money, why don't workers hire machines and expertise and make the money instead? This question has generated a large body of writing, including Stephen Marglin's much-cited article "What Do Bosses Do?" Marglin draws on history to argue that the employer, who added nothing to technical efficiency, used specialization of tasks to divide labor and impose himself as boss, thereby creating an artificial, unproductive role. These arrange- ments were embodied in domestic industry and were reinforced when employers turned to the factory system as a more effective disciplinary mode. This article argues that such a thesis misreads history and is essentially ideological.

If Marxist rumors about surplus value were true, one would have to assign a large, indeed inordinate share of product to profit. This

implicitly poses an interesting problem: if bosses make (cream off) so much money, why shouldn't boss-free enterprises (cooperatives, col- lectives, small self-employed workers, and the like) be able to outdo those capitalistic units that pay so heavy a toll to owners and managers? Why shouldn't workers be able to get together, share tasks and responsibilities, hire capital and expertise as needed, and sell for less? The exploration of problems suggested by these questions has given rise to a large and still growing body of literature, from the theoretical and hortatory to the empirical and historical.

As my title indicates, the paper that follows takes up a theme first presented by Stephen Marglin in 1974 in a seminal essay entitled "What Do Bosses Do?"'1 In a later article on the same subject, Marglin wrote that this essay had circulated in scholarly circles as samizdat, that is, as a kind of contraband literature.2 It is hard to imagine how such a fate

Journal of Economic History, Vol. XLVI, No. 3 (Sept. 1986). ? The Economic History Association. All rights reserved. ISSN 0022-0507.

The author is Professor of Economics at Harvard University, Cambridge, Massachusetts 02138. A reply by Stephen Marglin will appear shortly in this JOURNAL. ' "What Do Bosses Do?: Part I," The Review of Radical Political Economy, 6 (Summer 1974),

pp. 60-112. The article was republished as "What Do Bosses Do?" in Andre Gorz, ed., The Division of Labour: The Labour Process and Class Struggle in Modern Capitalism (London, 1976), pp. 13-54. Page references in this article are to the reprint in Gorz.

2Stephen Marglin, "Knowledge and Power," in Frank H. Stephen, ed., Firms, Organization and Labour: Approaches to the Economics of Work Organization (London, 1984), pp. 146-64; on samizdat, p. 156. In fact, the article is part of a larger stream of revisionist literature on the

585

Page 3: Landes - What Do Bosses Really Do

586 Landes

could befall the work of a professor of economics at Harvard Univer- sity; but if it will make Stephen Marglin feel better, I shall do my best here to take note of his contribution to the history of capitalism and industrial enterprise, which has had far more resonance than he seems to recognize or cares to admit.

In the beginning there was Smith, and Smith told us about the division of labor. Book I, Chapter I of The Wealth of Nations (1776) opens: "The greatest improvement in the productive powers of labour, and the greater part of the skill, dexterity, and judgment with which it is any where directed, or applied, seem to have been the effects of the division of labour."3

Smith, of course, was not the first. Stephen Marglin takes us back to Plato-"the attribution of the division of labour to efficiency antedates Adam Smith by at least two millenia [sic]."4 And Edwin Cannan, in his note to this passage, reminds us that Bernard de Mandeville, in his well-known but well-neglected Fable of the Bees (1714), had already marked this phenomenon, noting, if not distinguishing between, the social or occupational division of labor on the one hand and process division or specialization within an occupation on the other.5 But it was Smith who explicitly analyzed the advantages of the procedure, which Mandeville took as almost self-evident, and made it the linchpin of his model of economic growth.

Although not unattentive to division of labor as occupational differ- entiation, Smith focused his scrutiny on intraoccupational specialization and chose for his example the pin manufacture. He could have found an even more advanced example of division in the watch trade, which most of his predecessors had taken for illustration;6 but the very apparent

organization of labor and the structure of the firm and has drawn notice from mainstream as well as from radical economists. See, for example, Oliver E. Williamson, "The Organization of Work: A Comparative Institutional Assessment," Journal of Economic Behavior and Organization, 1 (March 1980), pp. 6-11; and Charles P. Kindleberger, "The Historical Background: Adam Smith and the Industrial Revolution," in Thomas Wilson and Andrew S. Skinner, eds., The Market and the State: Essays in Honour of Adam Smith (Oxford, 1976). Marglin's new article does make some changes in his earlier thesis, and I shall advert to these when relevant to the discussion that follows.

3Page references in The Wealth of Nations are to the Modern Library edition, edited by Edwin Cannan (New York, 1939).

4 Marglin, "What Do Bosses Do?" p. 47, n. 2. Kindleberger, "The Historical Background," argues that Smith's emphasis on the division of labor actually came from Plato, rather than from his immediate predecessors or from experience; he cites on this point Vernard Foley, "The Division of Labour in Plato and Smith," History of Political Economy, 6 (1974), pp. 220-42.

5 As does Marx, who never misses a chance to diminish Smith: Capital, vol. 1, p. 335, n. 1. Page references are to the Moscow edition, Progress Publishers, which is a translation of the third German edition, edited by Friedrich Engels.

6 Thus Henry Martyn (Martin), Considerations on the East India Trade (1701); and Bernard de Mandeville, The Fable of the Bees (5th ed., London, 1728).

Page 4: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 587

simplicity of the pin as an object made the degree of specialization in its manufacture that much more impressive. It was, he wrote, "a very trifling manufacture; but one in which the division of labour has been very often taken notice of."7

"The division of labour," he went on, "so far as it can be introduced, occasions, in every art, a proportionable increase in the productive powers of labour." These gains in productivity have three sources:

(1) The specialization of the workman increases his dexterity. (2) Specialization saves the workman the time and trouble of setting

up for new tasks. (3) The simplification of tasks opens the way to improvements in

technique, among other things, to the invention of machines.8 Concen- tration on a single operation concentrates the worker's mind and hands, and suggests "easier and readier methods of performing" the work. In this connection, note that Smith does not take account of the contribu- tion of simplification to mechanical imitation or reproduction of the work process. Karl Marx understood the link better: "The manufactur- ing period simplifies, improves, and multiplies the implements of labour, by adapting them to the exclusively special functions of each detail labourer. It thus creates at the same time one of the material conditions for the existence of machinery, which consists of a combination of simple instruments. "9

Smith's emphasis on the psychological advantages of specialization- what we may call the inspiration of monotony-has exposed him since to criticism and some mockery. Marx, for example, who accepted the general thesis of the gains to efficiency from division of labor, de-

Smith, Wealth of Nations, p. 5. Kindleberger, "The Historical Background," p. 7, remarks that the consequences of division of labor "can be found worked out complete with references to pins" in Carl, Traitg de la richesse des princes et de leurs 9tats et des moyens simples et naturels (1722).

8 Smith's "machines" were not necessarily engine-powered devices. He specifically remarks that the hand mill is also a machine. Lectures, p. 167, cited by Cannan in Smith, Wealth of Nations, p. 9, n. 17. In his discussion of the gains to division of labor, moreover, Smith makes no explicit reference to what we could call a factory, that is, a large manufacturing unit using machines driven by nonhuman power. There is no evidence that Smith had ever seen such a unit, which had only recently appeared in cotton spinning. But such units had existed since early in the century in silk throwing, and Smith must have known of their existence. Even so, my sense is that Smith's world was prefactory, prepower machinery; and that he had no awareness of the Industrial Revolution, then in its inception. There is some disagreement on this: most economic historians say he was unaware; historians of economic thought say he was not. Like Charles Kindleberger ("The Historical Background") I think the "unawares" have it. For a similar point of view, see Hiram Caton, "The Preindustrial Economics of Adam Smith," this JOURNAL, 45 (Dec. 1985), pp. 833-53.

For this reason, I find incongruous the statement that "Marglin and Smith agree in allowing early factories, both mechanized and nonmechanized [that is, powered and nonpowered], considerable potential for measured productivity growth." Kenneth L. Sokoloff, "Was the Transition from the Artisanal Shop to the Nonmechanical Factory Associated with Gains in Efficiency?: Evidence from the U.S. Manufacturing Censuses of 1820 and 1850," Explorations in Economic History, 21 (1984), p. 352. Not that Smith would have had trouble incorporating factories-that is, true, powered factories-into his division-of-labor paradigm.

9 Marx, Capital, vol. 1, p. 323.

Page 5: Landes - What Do Bosses Really Do

588 Landes

nounced him for assigning to "common workmen" (Smith's words) inventions due rather to "learned men, handicraftsmen, and even peasants. "10 In fairness to Smith, though, he sought his inventors in a wider sphere than hired labor alone. Thus, although in the passage cited above he does not speak of invention by employers, only by workmen, he would surely have argued that simplification gave ideas to both, if only because, in small shops, the employer was one of the workmen. And on the very next page he recognized that inventions might well have to come from other sources:

All the improvements in machinery, however, have by no means been the inventions of those who had occasion to use the machines. Many improvements have been made by the ingenuity of the makers of the machines, when to make them became the business of a peculiar trade: and some by that of those who are called philosophers or men of speculation, whose trade it is not to do any thing, but to observe every thing; and who upon that account, are often capable of combining together the powers of the most distant and dissimilar objects. In the progress of society, philosophy or speculation becomes, like every other employment, the principal or sole trade and occupation of a particular class of citizens. Like every other employment too, it is subdivided into a great number of different branches, each of which affords occupation to a peculiar tribe or class of philosophers; and this subdivision of employment in philosophy, as well as in every other business, improves dexterity, and saves time. Each individual becomes more expert in his own peculiar branch, more work is done upon the whole, and the quantity of sciences is considerably increased by it.1"

Marx's principal objection was not to Smith's economics, for he essentially repeats the arguments pro division of Smith & Co., but to the social and moral consequences of a system that deprived the workman of his independence and converted him "into a crippled monstrosity, by forcing his detail dexterity at the expense of a world of productive capabilities and instincts."12 The one thing that Marx added to the economics, he picked up from such bourgeois writers as Andrew Ure and Charles Babbage. This is the cost advantage that accrues to the disaggregation of a productive process into tasks of different degrees of skill and complexity. It is this disaggregation that enables the employer to select workers according to their capacity, paying none more than the nature of the task warrants. Thus Babbage:

The master manufacturer, by dividing the work to be executed into different

?0Ibid., p. 329, n. 4. 1" Smith, Wealth of Nations, I-i 10. Marx's criticism (Capital, vol. 1, p. 329, n. 4), then, is not

justified. But then, Marx saw in Smith what he wanted to see. 12 Capital, vol. 1, p. 340. And in a rhetorical fluorish: "As the chosen people bore in their

features the sign manual of Jehovah, so division of labor brands the manufacturing workman as the property of capital" (p. 341). A characteristic jibe by Marx, who bore in his own features "the sign manual of Jehovah." As is generally true with people who run from their origins, so with this reincarnation of a biblical prophet, Jewish by descent, Christian by parental decision, atheistic and antireligious by conviction, and vocally anti-Jewish because of this deliberate break with a distinguished Jewish lineage.

Page 6: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 589

processes, each requiring different degrees of skill or of force, can purchase exactly that precise quantity of both which is necessary for each process; whereas, if the whole work were executed by one workman, that person must possess sufficient skill to perform the most difficult, and sufficient strength to execute the most laborious of the operations into which the article is divided.13

To the best of my knowledge, it is not until our own time that the economic-as opposed to the moral-basis of the division of labor has been challenged. This revisionist view has come from the Left, a somewhat vague, heterogeneous cover term that I use to compensate for my ignorance of the sectarian alignments. Among the first critics was Stephen Marglin, in "What Do Bosses Do?" As part of an argument tending to demystify the capitalist ideal of technological efficiency, Marglin wrote that Smith's advantages were not what they seemed. Let me take the points in his order (which is not Smith's):

(1) It is true that division of labor saves set-up time; but this does not necessarily entail specialization by task. "To save the time that is commonly lost in passing from one species of work to another, it is necessary only to continue in a single activity long enough that the set-up becomes an insignificant portion of total work time." 14

(2) Smith's third argument regarding the encouragement to innovation "is not terribly persuasive."" As Smith himself notes, repetitive labor does not stimulate creative faculties; rather it stultifies, making the worker "as stupid and ignorant as it is possible for a human creature to become. "16

(3) As for Smith's first advantage, the gain in dexterity due to practice, it may make sense when speaking of musicians or dancers or surgeons, but not pinmakers. The argument, by implication, is that for run-of-the-mill industrial work, there is no significant difference be- tween tasks that would warrant specialization by age, experience, or strength. Marglin does not deal specifically with the Babbage-Ure analysis, but in effect rejects it by omission. My own sense is that it raises questions of cost efficiency that he prefers to avoid or deems irrelevant.

This equal-skill thesis is not entirely consistent with the distinction Marglin draws between division of labor and specialization, or between "separation of tasks" and "capitalist division of labor." The former, he recognizes, is found in all systems of production; by implication, therefore, it has a raison d'etre in the nature of complex work. This, he

13 On the Economy of Machinery and Manufactures (London, 1832), chap. 19. See also the comments in Harry Braverman, Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century (New York and London, 1974), pp. 79-82. Braverman's revival of this issue was badly needed.

14 Marglin, "What Do Bosses Do?" p. 18. 15 Ibid. 16 Smith, Wealth of Nations, p. 734.

Page 7: Landes - What Do Bosses Really Do

590 Landes

says, is not necessarily bad: "In precapitalist societies, industrial production was organized according to a rigid master-journeyman- apprentice hierarchy, which survives today in anything like its pure form only in the graduate departments of our universities."17 A good reference? Marglin finds in such relations important virtues: that they are what he calls linear, with provision for promotion up along the line; further, that the master "works along with" the apprentice, instead of simply telling him what to do.'8 He contrasts this with pyramidal arrangements, in which a few direct many and only a handful are chosen for advancement to higher status.

This is what we have in capitalism: the separation of tasks becomes what Marx called detail labor, that is, labor in which different workers carry out different pieces of work and produce nonproducts, defined as objects that have no "wide market." The producer, in these circum- stances, writes Marglin, is obliged "to make use of the capitalist as intermediary to integrate his labour with the labour of others and transform the whole into a marketable product."19 "Why," he then asks, "did the division of labour under the putting-out system entail specialization as well as separation of tasks? In my view the reason lies in the fact [sic] that without specialization, the capitalist had no essential role to play in the production process."20 And further: "The capitalist division of labour, as developed under the putting-out system, embod- ied the same principle that 'successful' imperial powers have utilized to rule their colonies: divide and conquer.''21 Create a problem and then solve it, thereby inventing a role-a highly lucrative role.

The economic historian who encounters this kind of a priori argu- ment, with a few swift allusions to the pin industry thrown in by way of obeisance to empirical evidence and a potent anti-imperialist metaphor by way of bona fides, is hard pressed to know where to begin. Three questions at the least call for consideration. The first is the bread-and- butter question: Is this the way it happened? The second is an exercise in the sociology of knowledge: Where is Marglin coming from? The third is: So what?

On what actually happened-or, what did bosses really do?-I shall

17 Marglin, "What Do Bosses Do?" p. 16. 18 There is an irresistible temptation on the part of social critics to idealize past arrangements by

way of contrast with an undesirable present. (See Friedrich Engel's view of the putting-out system in The Condition of the Working Class in England) In fact, the traditional apprentice contracted to obey his master and did what he was told-even to the extent of non-work-related errands and tasks for the master's household. That was in the nature of an open-ended relationship between "parent" and "child." As for graduate departments in universities today, it is true that the students, unlike apprentices, have a great deal of autonomy. There are limits, however, as in a large course in which professor and teaching assistants disagree about method or content.

19Marglin, "What Do Bosses Do?" p. 16. 20Ibid., p. 20. 21 Ibid.

Page 8: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 591

follow Marglin's order of the advantages, alleged or true, of the division of labor:

(1) Set-up time: It is true that a skilled worker could in principle perform all the tasks that went into the manufacture of finished commodities, from the preparation to the working to the finishing. But no skilled worker would, if he could help it, perform those simple tasks that would waste his time and experience, or whose character rendered them dirty, noxious, or disagreeable. (Comparative advantage, we know, holds in interpersonal as in international exchange.) Unless strength requirements imposed the use of an adult male, there were women and children to do those jobs. This was indeed the kind of specialization that characterized the household (the oikos, which gave us our word economics); and the workshop, which was initially an enlargement or extension of the household producing unit.

It should not be thought, moreover, that this specialization, which is almost as old as work itself, was simply a matter of efficiency and convenience. It was at the same time the vehicle for instruction and training. This was the way youngsters learned the techniques of the masters, by doing the simple and rough jobs to start with and graduating to more difficult tasks. In this sense, it is no accident that handicraft shops, reflecting the organization of the household, brought master and apprentices together in principle as father and children. There was even an established calculus of training and advantage: it was a common- place, for example, in the clock and watch trade that in the first years the apprentice could not be entrusted with the more highly skilled tasks: he would only botch the job. By the third and fourth years, though, the apprentice could begin this kind of work, and the loss from bad jobs was compensated by the low wage. And then, in the last year or two, the master appropriated an increasing surplus over and above the cost of labor. Hence some very bitter litigation between masters and those apprentices who left too early: such departures deprived the master of the best years of the apprenticeship contract.22

In this sense, the speculation that skilled workers could reduce set-up time by a serial arrangement of tasks is simply fantasy-what we might call economoneirics, or dream economics. They could, but they wouldn't. Not unless they could not find children to hire, or women, or illegal aliens, or such other labor as was ready to work for a lower wage.

(2) Invention: It is true that repetitious labor can be stultifying, so that detail workers may not be a very good source of new ideas. I say "may not be," because I am not sure whether this is invariably true. At least we are often told that workers, with appropriate encouragement and

22 See, for example, E. Develle, Les horlogers bhesois au XVIe et au XVIIe sikle (3rd ed., Nogent-le-Roi, 1978) pp. 58-59.

Page 9: Landes - What Do Bosses Really Do

592 Landes

incentives, can be a rich source of suggestions regarding improvements in technique, and that employers would do well to mine this vein.

Yet as we noted above, Smith himself realized that the effects of repetitious labor might be deleterious and that inventions might well have to come from other sources. And even that misses the point, which is that simplification suggests and facilitates the imitation by machines of manual skills.

(3) The increase of dexterity: Every witness we have testifies to the gains to skill accruing from specialization. Even hostile witnesses: the poet Robert Southey, recollecting the children in a Manchester cotton mill, spoke of "the unnatural dexterity with which the fingers of these little creatures were playing in the machinery."23 We have a proverb to this effect: practice makes perfect.

Marglin does not really question this, but argues that some skills are so elementary that anyone could learn them. We are not talking here, after all, about neurosurgery.

To be sure; and indeed, it is precisely because so many tasks, once simplified, can be learned quickly, that it pays to reduce them to elemental forms. Not only can the employer thereby hire low-wage, unskilled labor, but he can shape it to the habits and performance desired. This is no small matter with old hands, who can offer an intense resistance to new ways, and it is no accident that entrepreneurs desirous of launching new techniques have often established themselves in strange places and drawn on an untrained labor force.24

But Marglin takes the argument here a step further. If these tasks are so simple, why specialize at all? Why can't everyone learn every task? And he takes his example from Smith's pinmaking industry: "It would

23 Cited in C. Aspin, Lancashire, the First Industrial Society (Helmshore, 1969), p. 36. The book has no footnotes, even for quotations.

24 Some critics of the concept of proto-industrialization have tried to discredit it by noting that areas of cottage industry have not always (or have often not) moved on to the factory stage (Marx's Modern Industry). But this is precisely what one would expect. For an example of this kind of reasoning, see Rab Houston and K.D.M. Snell, "Proto-industrialization? Cottage Industry, Social Change and Industrial Revolution," The Historical Journal, 27, no. 2 (1984), pp. 490-91. In fairness to Messrs. Houston and Snell, the word "proto-industrialization" does seem to imply a promise, and that promise is not always fulfilled. We might have done better to stick with good old "cottage industry" and "putting out."

On new locations and new labor, see Gregory Clark, "Authority and Efficiency: The Labor Market and the Managerial Revolution of the Late Nineteenth Century," this JOURNAL, 44 (Dec. 1984), p. 1077 and n. 31; David S. Landes, Revolution in Time: Clocks and the Making of the Modern World (Cambridge, Mass., 1983), pp. 289, 380 Table 2, on the introduction of machine manufacture into the Swiss watch industry and the incorporation of a new labor force; and Carol Heim, "Structural Transformation and the Demand for New Labor in Advanced Economies: Interwar Britain," this JOURNAL, 44 (June 1984), pp. 585-95. The emphasis of economists on the value of "firm-specific" skills, therefore, is one-sided. Such skills do matter within a given technological context; but change the technology (introduce new machines, for example) and an asset turns into a handicap. On the value of firm-specific skills, see Oliver Williamson, Markets and Hierarchies: Analysis and Anti-Trust Implications (New York, 1975).

Page 10: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 593

appear to be the case that the mysteries of pinmaking were relatively quickly learned, and the potential increase in dexterity afforded by minute division of tasks was exhausted. Certainly it is hard to make a case for specialization of workmen to particular tasks on the basis of the pin industry."25

Really? In that case, why hire adult males at all? If Marglin's analysis were true, this would imply a strangely irrational pattern of entrepre- neurial behavior. Marglin relies here on an article by T. S. Ashton on pinmaking in the early nineteenth century, which he says shows that "there were no special skills" in the industry and "no great discrepan- cies among the various branches of pin production." 26 Fortunately for entrepreneurial rationality, a reading of Ashton's article shows some- thing very different. There were apparently major skill differences among the component tasks. Some were reserved to men who earned 20 shillings and more a week. Some could be performed by children getting as little as Is. to Is. 6d. And these children generally worked with and under the eye and hand of women who earned five times as much, partly for their work but even more for their role as supervisors of the child labor. And for their role as beaters: the treatment of working children was sometimes brutal, as much in cottage industry as in the later factory system; and parents were not much kinder than strangers: "When I was five, my mother took me to lace school and gave the mistress a shilling. She learned me for half an hour, smacked my head six times, and rubbed my nose against the pins." The younger the better: "Six is the best age, you can beat it into them better then. If they come later, after they have been in the streets, they have the streets in their minds all the while." And the more frightened the better; in the words of a pacemakers' ditty:

There's three pins I done today, What do you think my mother will say? When she knows I done no more, She'll take and turn me out of door, Never let me come in any more.

Even a hard mother was better than a stranger.27

When all is said and done, then, Smith and the other advocates (admirers) of the division of labor seem to have had a realistic sense of

25 Marglin, "What Do Bosses Do?" p. 20. 26Ashton, "The Records of a Pin Manufactory, 1814-21," Economica, 15 (Nov. 1925), pp.

281-92. 27 The most common ailment was a nervous stomach. The above quotations are from G.F.R.

Spenceley, "The Health and Disciplining of Children in the Pillow Lace Industry in the Nineteenth Century," Textile History, 7 (1976), pp. 166-69. Depressing but important. See also the classic study by Ivy Pinchbeck, Women Workers and the Industrial Revolution (London, 1930), pp. 232-35.

Page 11: Landes - What Do Bosses Really Do

594 Landes

its economic advantages; but less of its social disadvantages (the eighteenth century thought work was good for children, that it built character). It did make possible important savings in the cost of production and, given elastic demand, opened new markets that in turn promoted further specialization. Smith, writing about the pre-Industrial Revolution world, made this dynamic the engine of economic growth in the days before mechanization. And that it was. Was the system more "efficient" than unspecialized production?-Marglin's question. To appreciate his answer, one has to understand that he is speaking of technological efficiency, defined as output over input, with input of labor measured not by cost but by labor hours weighted for effort. His comments on this subject are relevant to the shift from cottage work to supervised work, rather than from the craft workshop to the cottage: "Even if labour is the only input, a new method of production might require more hours of labour, or more intensive effort, or more unpleasant working conditions, in which case it would be providing more output for more input" and would not necessarily be technologi- cally superior.28 A change in technique that substituted child for adult labor, or unskilled for skilled, and thereby achieved the same output at lower cost, falls outside this conceptual framework.

Technological efficiency is, he says, crucial to his argument, but to an economic historian trying to understand the motivation of entrepreneur- ial choices, it is only part of the story. The historian wants to know the options as they presented themselves to a manufacturer, and here what mattered was cost efficiency and predictability (enforceability) of out- put. The Babbage-Ure analysis is right on target. The organizer of production did not ask whether skilled adult males could work as fast and do as much as a mix of people of varying skill. He simply knew that a team of detail laborers could do the job for less and undersell unspecialized manufacture; also, as we have seen, that women and children could be compelled to "deliver" work that adult males could not be counted on to perform.

For Marglin, however, costs are irrelevant and detail labor a fig- ment-an artificial construct by "capitalists" who saw here a way of making a place for themselves they would not otherwise have had: "without specialization, the capitalist had no essential role to play in the production process. "29 I am not sure what this means: that the capitalist's role was useless and could be dispensed with? No one with any experience of business would make such an argument. The ability to combine the factors of production in such a way as to make goods cheaper is one of the central aspects of entrepreneurship.30 The state-

28 Marglin, "What Do Bosses Do?" p. 16. 29 Ibid., p. 20. 30 In his second article ("Knowledge and Power," p. 150), Marglin explicitly recognizes this:

Page 12: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 595

ment has to be turned around: in a premachine era, the ability of "capitalists" (merchant-manufacturers) to break down production into a number of simple tasks and assign them to workers of different degrees of skill and experience, hence working for different wages, was the essential task of the capitalist.

Marglin further suggests that even if the capitalist had played a useful role in organizing new forms and modes of production, once that task was accomplished, the worker could have dispensed with him:

If each producer could himself integrate the component tasks of pin manufacture into a marketable product, he would soon discover that he had no need to deal with the market for pins through the intermediation of the putter-outer. He could sell directly and appropriate to himself the profit that the capitalist derived from mediating between the producer and the market.31

Yes and no. The question is, sell directly to whom? To final users or retail shops? That kind of thing is time-consuming and could play havoc with production. To distant markets? That took even more time and often called for language skills not commonly found among industrial producers. And none of this was good for the health: in Britain, at any rate, the commercial traveler went by horseback from inn to inn, on bad roads and in all kinds of weather, carrying (and loading and unloading) stock, while imbibing nostrums for "coulds," "coffs," "the ague" (now there's a word one doesn't hear any more), and unspecified "fevers." The boldest men could turn to jelly at the thought of another season on the road: I'd rather take up the file, wrote J. B. Vacheron to his partner in Geneva. He had just worked his way over the Apennines, in a coach to be sure (but before the construction of the autostrada, this was, and is, a miserable route, even on paved roads in well-sprung automobiles); and as any moviegoer knows, nothing was so attractive to a brigand as a stage coach.32

The manufacturer could, of course, put his goods in the hands of an intermediary, a wholesaler; but then he would be back at square one, making the goods for others to sell. Or he could engage a traveling salesman to sell his work; but he would have to produce or buy up a large and diverse stock to keep a specialized salesman busy, and then, lo and behold, our small, independent, complete pinmaker would have turned into a capitalist himself. As might his employee: given the semi-independence of cottage workers, the salesman could buy stock on

"the essence of the capitalist's contribution is not capital, but organizing ability." But then he goes on to say that that ability was unnecessary: it is only because the capitalist has "imposed" an organizational form that makes him essential that he is able to "secure a reward" from the production process. So the capitalist could (and can) be dispensed with. See also Andre Gorz, "The Tyranny of the Factory: Today and Tomorrow," in Gorz, ed., The Division of Labour, p. 55.

31 Marglin, "What Do Bosses Do?" p. 20. 32 Landes, Revolution in Time pp. 252-53.

Page 13: Landes - What Do Bosses Really Do

596 Landes

his own (conceivably including goods made from materials furnished by and embezzled from his own employer), sell it alongside that of his employer, and eventually become a merchant-manufacturer himself. Employee turnover was rapid. The only reliable salesman was a partner, better yet a blood relative.33

All of this may seem in its turn hypothetical, but this was the way of trade and trades. Take, for example, the clock and watch manufacture. It is one well worth considering because it requires work of widely varying skill and hence lent itself from the beginning to specialization and division of labor. Indeed, most of the early writers on this subject used it as the ideal example of such division. Besides, it is an industry I know something about, and it is always a good idea to write about things one knows about.

In clocks and watches, the work was originally done in shops organized along classical handicraft lines (in German: das alte Handwerk). Each shop had its master, assisted by one or more apprentices and possibly one or more journeymen, that is, craftsmen who had completed their training but were not yet ready or able to open shop for themselves. These shops came early to be governed by corporate or guild regulations, designed partly to ensure the quality of work, but even more to restrict competition without and within. In particular, shops were limited in size and masters were forbidden to purchase and resell work done outside the shop. These constraints were, of course, a response to market changes threatening the stability of the manufacture and the theoretical parity of its members.34

These sources of instability came on both the demand and supply sides. On the one hand, consumers preferred the work of some masters over others, so that some shops had more to do than they could, while others lacked for orders. On the other hand, the formation of appren- tices tended to outstrip the openings for masters, so that there devel- oped a class of skilled journeymen looking for work and ready to produce for others in spite of rules to the contrary. Some masters, responding to this disequilibrium, became in effect capitalists. The best of them, or perhaps the most enterprising, soon found themselves devoting an inordinate part of their time to selling rather than making. They signed their clocks and watches, but made no part of them. They did not have to invent specialization to make a place for themselves. The specialization was already there; the gains to specialization made the market; and producers responded to market opportunity.

Alongside these workers turned merchant-manufacturers were mer- chants looking for workers. Some of these entrepreneurs came from

3 There is good material in S.R.H. Jones, "The Country Trade and the Marketing and Distribution of Birmingham Hardware, 1750-1810," Business History, 26 (March 1984), pp. 24-42.

34 See Landes, Revolution in Time, chap. 13: "The Good Old Days that Never Were."

Page 14: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 597

related trades-jewelers and goldsmiths, for example-whose clients also wanted watches, or who, in the course of attending the fairs and traveling from city to city, found buyers for other articles. The medieval and early modern merchant was ajack-of-all-trades, ready to buy cheap and sell dear and resistant to any efforts to put him in a box-business, like love, laughs at locksmiths. It is these merchants who came back from Istanbul and brought large contracts for "Turkish watches" to the makers of Blois and Geneva-themselves usually etablisseurs, that is, merchant-manufacturers; or who planted themselves at Canton and sent back to correspondents in the Jura orders for "mandarin watches," thereby laying the basis for the important Fleurier section of the Swiss watch manufacture. These merchants were not intruders, levying an unnecessary, hence unjustified, tax on producer victims. On the con- trary, they brought in the business and made it possible for the producers to work. They also brought in opportunities for growth based on a continuous tendency toward division of labor cum specialization. By 1830 the mountain Swiss had broken up the manufacture of a watch into fifty parties brisees and this in the absence of mechanization.35 Nor was this the limit: the pressure toward simplification continued, because it was the key to lower costs; also to the removal of craftwork obstacles to the pace and character of the production process.

If one wants to find instances of "artificial," parasitical hierarchy in industrial relations, one of the best places to look is among serf producers in eastern Europe.36 These peasants were bound personally to their lord and did such work as he required. In the early modern period, more and more of this work was industrial-in particular spinning and weaving. Flax yarn, for example, was sold by peasants to "yarn collectors," themselves often peasants or peasants-turned- publicans and -innkeepers, who sold the yarn at weekly markets to village weavers. These in turn sold the cloth to merchants, who sometimes moved about and inspected the pieces on the loom. The lord meanwhile levied his tax on all transactions-a tax that was nothing more than a "feudal" rent: he was renting the labor of his serfs.

In this system, then, the lord was a parasite; the peasant, a proletar- ian-pure source of labor power; and the merchant, an enterpriser. The availability of this large source of cheap labor (even with fees to the lord) made central and eastern European serf manufactures highly

35Ibid., p. 451, n. 30. A report of 1798-1799 stated that one hundred and fifty workers contributed in one way or another to the making of a watch-probably an exaggeration, but still strong evidence of the strategy that made Switzerland synonymous with watches.

36 The paragraphs that follow draw heavily on Robert Millward, "The Early Stages of European Industrialization: Economic Organization under Serfdom," Explorations in Economic History, 21 (1984), pp. 406-28. An important piece, but marred by heavy jargon. Do economic historians feel that they have to sell their competence to economists? They would do better to communicate their results to historians as well. It might help recruit people into the discipline.

Page 15: Landes - What Do Bosses Really Do

598 Landes

competitive. Output grew fastest, of course, precisely in those districts with the poorest soil, for it was there that cultivation was least remunerative and the marginal utility of labor lowest. The richer peasants (those with animal teams and subject to full corvee) were typically not involved; but cottars and bordars and day laborers (Hausler, Kdtner, and Kossaten) were fodder for industry.

At this point, one might imitate Marglin's approach and ask a hypothetical: Why didn't the lord manage this whole operation and collect more than his fees? Why did he let these outside merchants come in, distribute and collect, and make the much bigger profits derived from buying and selling? It was not, after all, for lack of experience: if one goes back in time, the peasants in these areas had long delivered manufactures as part of their seignorial dues; the lord consumed some of these and resold the rest. But as opportunities opened up for sale to distant markets, outside merchants were allowed to enter the estate and take over the role of commercial intermediary.

The answer is twofold. To begin with, the merchant knew far more about market opportunities and the composition of consumer demand than the lord did: it paid the lord, therefore, to give the merchant a hiring as well as a hunting license. But equally important was the code of social values, which defined some activities as appropriate to gentility (nobil- ity) and some as not. To use Marglin's "more neutral language," the lord's snobbery was a preference for graceful, dispendious leisure. It is no accident that large estates always hired stewards (the word originally meant the keeper of the pigsty); day-to-day supervision of dusty and muddy tasks was not appropriate work for a proper gentleman. And this sometimes meant that the steward became richer and more powerful than his master: styward gave us steward gave us Stewart or Stuart, the name of Scotland's royal family.

But one does not have to look to such class- and status-bound societies as those of central and eastern Europe for hierarchy cum intermediation. In eighteenth-century Britain, middlemen putters-out sprang up between manufacturers and weavers-like the man who advertised that he "would be willing to engage with any Manufacturer to put out Checks by Commission . .. as he is fix'd in a town famous for making a remarkable stout Article and can command a Set of the very best Workmen. He could get some Cotton Yarn."'' And middlemen hosiers intervened between the merchant-manufacturers, owners of the knitting frames, and the knitting workers, who had traditionally rented these directly. These middlemen hired numbers of frames and re-rented these to the workers. The frame owners thereby settled for less, while

37A. P. Wadsworth and Julia de L. Mann, The Cotton Trade and Industrial Lancashire, 1600-1790 (Manchester, 1931), p. 277.

Page 16: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 599

the knitters paid more; the middlemen took the difference.38 Again a hypothetical: Why did the large hosiers do this? These were, after all, experienced businessmen; they didn't have to put up with an "artifi- cial" intruder. The answer, of course, is that it paid them to do so-not in income per working frame, but in stability of return and in trouble avoided (they too had their leisure preference). And that's why it pays a giant like the telephone company to sell blocks of time to wholesalers, who then resell these to consumers at retail rates lower than those practiced by the telephone company itself. Or why it pays the town of Brookline, Massachusetts, to rent municipal parking lots to private operators, who rent the same old space to the same old parkers at a new and higher price.

But back to the early modern period and to the cottage manufacture that was the primary focus of industrial expansion. Guilds and guild regulations were urban phenomena; and although violations took place all the time, the city was not a place to increase output, if only because the overproduction of skilled artisans became undersupply when the market extended far beyond the local area. The answer lay in tapping a new supply of labor, outside the guild domain and necessarily less skilled than trained journeymen. This was found in the countryside, among rural populations whose irregular work calendar left them time for industrial occupations. Merchant-manufacturers had no illusions about turning these people into masters by means of apprenticeships of several years' duration. They wanted results quickly, and the answer lay in assigning simple tasks, fragments of the larger process. With such simplification, as we have seen, inexperienced workers were actually advantageous, in that they offered a tabula rasa: the employer could train them to his needs. Whereas a highly skilled worker had acquired habits and normative standards that might oppose novelty, these newcomers would do as they were told. No one was more articulate on this point than Josiah Wedgwood, who had very strong ideas on what he wanted:

You observe very justly that few hands can be got to paint flowers in the style we want them. I may add, nor any other work we do-We must make them. There is no other way. We have stepped forward beyond the other manufactures and we must be content to train up hands to suit our purpose. Where among our Potters could I get a complete Vase maker? Nay I could not get a hand through the whole Pottery to make a tableplate without training them up for that purpose, and you must be content to train up such Painters as offer to you and not turn them adrift because they cannot immediately form

38 On middlemen in hosiery, John Rule, The Experience of Labour in Eighteenth-Century English Industry (New York, 1981), p. 141f., who may not be saying quite the same thing as I am. See also similar arrangements in hardware in Jones "The Country Trade," p. 25.

Page 17: Landes - What Do Bosses Really Do

600 Landes

their hands to our new stile, which if we consider what they have been doing all their life we ought not to expect from them.39

The Wedgwood enterprise, as Josiah himself pointed out, was excep- tional- "beyond the other manufactures." Both source and conse- quence of its lead were its innovations in management: the design (and sometimes creation) of a multitude of administrative and supervisory functions at both shop and office levels. Wedgwood hired clerks, weighers, scribes, packers, and similar specialists, inventing tasks and procedures to promote rational use of all the factors of production. In all this, he was exceptional only by his precocity: the growing complexity of managerial responsibilities, anything but artificial, finished by impos- ing its own division of labor.40 Some of these white-collar employees, we have seen, became employers in their turn; some of them became managers-a new profession. When workers learn that they can do without the capitalist, it is because they have become capitalists themselves.

It is this widespread pattern of occupational and functional mobility that gives meaning to Marglin's rhetorical question: "Why didn't some enterprising and talented fellow organize producers to eliminate the capitalist putter-outer?" His answer is that such an enterpriser could not capture the gains and remain a producer, that is, a worker. His co-workers might have bought him a gold watch, but not much more. "To glean rewards from organizing, one had to become a capitalist putter-outer!""4 Exactly. But one didn't have to become a capitalist; such an organizer was one already. And that was the point of the exercise.

In such flexible, fuzzy circumstances of transition, of course, it was not uncommon for producers to continue working while hiring others to augment output. Worker and capitalist were one. This combination of old role and new was frequent in the watch trade, where most putters-out were old hands at the bench. But increasing responsibilities on the management and commercial sides tended to pull such people away from production, except insofar as they continued to design layouts and check quality.42 Besides, highly skilled work is very

9 Letter of Josiah Wedgwood to Thomas Bentley, 19 May 1770, in Ann Finer and George Savage, eds., The Selected Letters of Josiah Wedgwood (London, 1965), p. 92.

40 On the development of "white-collar" specialties, see Marie B. Rowlands, Masters and Men in the West Midland Metalware Trades before the Industrial Revolution (Manchester, 1975), pp. 84-87.

41 Marglin, "What Do Bosses Do?" p. 21. 42 A further encouragement to the separation of roles is the desire to avoid the confusions of

fellowship and intimacy. Such a consideration could be important for both sides, especially in situations of disagreement and conflict of interest. See Richard Price, Masters, Union and Men: Work Control in Building and the Rise of Labour, 1830-1914 (Cambridge, 1980), p. 129: what was probably the first business manual for builders urged employers not to work along with the men, because that would confuse the true character of the relationship.

Page 18: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 601

demanding, and idleness does not make fingers nimbler. As the concert pianist put it, "If I don't practice one day, I know it. A second day, and the critics know it. A third day, and the public knows it."

By comparison with customary guild wages, of course, rural labor power came cheap; but from the point of view of these farm people, the new sources of income were a major addition to well-being and independence. So effective was this symbiosis that village life was transformed: population grew rapidly as industry migrated to areas of easy settlement, because that was where their burgeoning workforce could move; young people, accustomed to waiting long years before they could achieve the independence of farm ownership, now could support themselves, marry early, and buy out their parents.43

It was this expansion into rural areas that made European and especially English (later British) industrial products competitive world- wide and laid the foundations of the Industrial Revolution-not the big manufactories (protofactories without central power, hence equipped with hand tools and machines), for all their resemblance in physical appearance and spirit of organization to the later factories; but the cottage manufacture, which built on specialization, found the cheapest labor available, and cut the cost of the final product.44

And from the workers' point of view, see Hobsbawm quoting Stedman-Jones quoting Hodgskin: masters who belong to the "useful classes" are "labourers as well as their journeymen," and insofar as they are needed "to direct and superintend labour, and to distribute its produce," are all right; but they are also "capitalists or agents of capitalists, and in this respect their interest is decidedly opposed to the interests of their workmen." E. J. Hobsbawm, "Artisan or Labour Aristocrat," Economic History Review, 2d ser., 37 (Aug. 1984), p. 361.

4' The best description of this pattern is to be found in Rudolf Braun, Industrialisierung und Volksleben: Verinderungen der Lebensformen hunter Einwirkung der verlagsindustriellen Heimarbeit in einem ldndlichen Industriegebiet (Zurcher Oberland) vor 1800 (Erlenbach-Zurich, 1960; 2d ed., G6ttingen, 1979). This was written before the invention of the term proto- industrialization, but it was the first major study of this phenomenon in its social and cultural as well as economic aspects, and it is still the best book on the subject. Further on this: Franklin F. Mendels, "Proto-Industrialization: The First Phase of the Industrialization Process," this JOUR- NAL, 32 (1972), pp. 241-61; and Peter Kriedte, Hans Medick, and Jurgen Schlumbohm, Industri- alization Before Industrialization: Rural Industry in the Genesis of Capitalism (Cambridge and Paris, 1981).

44 Houston and Snell, "Proto-industrialization?" (cited in fn. 24, above), are skeptical, because they assume that if cottage industry promoted the development of modern industry, one should find factories appearing in areas of dispersed manufacture. They miss the implications of lower-cost rural manufacture for the extent of the market and rising demand, hence as a stimulus to a change in the mode of production, not necessarily in situ.

Karl Marx, who had no access to the information we now have on the historical role of cottage industry, stressed the role of handicraft shops and the manufactory as the preliminary to modern industry. He may have been influenced in this by the history of manufacture on the Continent, where efforts to industrialize from above accorded the manufactory a special role as vehicle for investment and technological change. But Marx would surely have been pleased to incorporate putting-out into his stage schema, for it would have provided him with a congenial social explanation for the shift from hand tools and machines to powered equipment, in terms of the internal contradictions of the older mode. See David S. Landes, The Unbound Prometheus: Technological Change and Industrial Development in Western Europe, 1750 to the Present

Page 19: Landes - What Do Bosses Really Do

602 Landes

The expansion of industry due to putting-out and its success in drawing new and cheap labor into manufacture set the stage for the next step: the introduction of a new mode of production-the factory system. For this, it is crucial to specify the termini a quo and ad quem. The move took place in the cotton manufacture, and it was a move from putting-out to the factory, not from craft shops or large workshops or manufactories to the factory.

Here I think Marglin gets the story right. I ought to think so, because he adopts the model of explanation that I put forward in The Unbound Prometheus. This, I confess, is a source of considerable satisfaction, partly because an endorsement (even an implicit endorsement) by so keen a critic is no mean reference, partly because one rarely has the pleasure of seeing one's ideas still current some fifteen years after publication-at least not in so active a field as economic history.

Let Marglin state the thesis:

Thus the very success of pre-factory capitalism contained within it the seeds of its own transformation. As Britain's internal commerce and its export trade expanded, wages rose and workers insisted in [sic] taking out a portion of their gains in greater leisure. However sensible this response may have been from their own point of view, it was no way for an enterprising capitalist to get ahead. Nor did the capitalist meekly accept the workings of the invisible hand.45

Instead, the capitalist tried to compel the worker by fines (and worse) to finish his tasks promptly and deliver the finished goods as promised, undiminished by theft of material; but these efforts failed, as incentives to embezzlement grew and competition for labor increased. Marglin goes on: "It is no wonder that, as Landes says, 'the thoughts of employers turned to workshops where the men would be brought together to labour under watchful overseers.' "46

Marglin agrees, then, that the motivation for the move from putting- out to factory was the employer's desire to gain control over the work process. Factory equaled discipline cum supervision. To quote Marx quoting Ure: " 'Order' was wanting in manufacture based on 'the scholastic dogma of division of labour,' and 'Arkwright created or- der.' "48 Where Marglin and I disagree, however, is that he thinks that this was enough to give the factory an edge and make it a dominant mode of production.

This is clearly not so. Before entering into reasons, however, I want

(Cambridge, 1969), pp. 58-60. Marx's discussion of the problems caused by worker insubordina- tion refers to skilled handicraft labor (Capital, vol. 1, pp. 346-47).

45 Marglin, "What Do Bosses Do?" p. 35. 46 Ibid., p. 37. 47 I learned this very important point from my teacher Abbott P. Usher. He hinted at it in his still

valuable Introduction to the Industrial History of England (Boston, 1920). And he stressed it in class.

48 Marx, Capital, vol. 1, p. 347.

Page 20: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 603

to clarify a point that is the source of some confusion. A factory, as noted above, is not simply a large production unit or workshop. A factory uses power-driven machines, and such units do not appear in Britain before the eighteenth century-first in silk, then in cotton. Large workshops, however, go back centuries before; and it is worth consid- ering where and why they appear. We find them in fulling, dyeing, brewing, tanning, papermaking, glassmaking, ceramics, metallurgy, shipbuilding, and, on an ad hoc basis, in construction (the chantier). A glance at the technology quickly explains this tendency to concentra- tion: these branches are characterized by one or more of the following features:

(1) They consume relatively large amounts of heat or other energy. (2) The work requires more space than is available in a room or

cottage, either because the object produced is very large (a ship or a building), or because the manipulations take room (large tables for shearing or printing), or because stocks of raw materials and semifinished objects take room (furniture, hats).

(3) The materials employed are so costly (precious metals, mercury, best-quality silk or wool) that one cannot afford the embezzlement that invariably occurs with unsupervised labor.

(4) The tools and machines (hand-driven) employed are too expensive for most workers (smallwares manufacture using the "engine loom").

(5) Their noxious or noisome character limits locational opportunity: not everyone will have them (tanning).

(6) There are special security considerations (arsenals; secret tech- niques of manufacture, as at the Murano glassworks near Venice).

In such circumstances it pays to bring workers together in larger- than-handicraft-shop units. Some of these processes, to be sure, can be effected by one or two artisans working alone: one can dye fabrics, for example, or make hats to order. But as soon as one starts producing "on speculation," that is, making goods in advance for customers some- where "out there," it pays to work cooperatively, to assign tasks by strength and skill, and to grow to such size as will exhaust economies of scale.

These considerations were not relevant to the most important indus- try of all in terms of factors employed and value of product: the textile manufacture, or more exactly, spinning and weaving. Here the instru- ments of production, the wheel and the loom, were such that there was

49These economies, as Marx pointed out, are determined in large part by the appropriate combination of specialists: an efficient unit is one that exhausts the potential contribution of each worker (Capital, vol. 1, pp. 326-29). On the tendency to growth of handicraft shops in the United States, see the Sokoloff article cited in fn. 8, above. This is an analysis of census data and establishes conclusively the tendency to growth and the gains to efficiency from increasing size. On the other hand, the association between size and efficiency is not an explanation, and the article invites attention to the techniques of the branches studied.

Page 21: Landes - What Do Bosses Really Do

604 Landes

nothing one could do in a large shop that one could not do as well and cheaper in the home of the worker. Dispersed manufacture, indeed, offered important cost advantages to the merchant-manufacturer. He was able to shift capital expenditures (plant and equipment) to the worker, along with the risk of fluctuation in demand. If orders fell off, the putter-out had only to stop buying. He had no machines to stand idle eating up the interest on capital. For another, labor costs were lower in cottage industry. This was the more so because workers preferred to work at home rather than in someone else's place under supervision. At home, wife and children could be of help when needed: the household, as we saw, constituted its own little realm of specialization. And at home, it was the worker who set the pace-worked when he wanted, rested when he wanted, ate when he wanted. It took higher wages to lure or seduce these people into workshops; and given the established preference for leisure, high wages were not enough. No wonder, then, that cottage manufacture held on tenaciously even after the factory rendered it technologically uncompetitive. It was ideally suited to accommodate the variable portion of demand, and numbers of factory enterprises, in wool for example, hired and fired cottage weavers while keeping their own loom sheds occupied.

Analogous was the earlier tendency (late eighteenth century) in the wool manufacture to bring some hand processes into concentrated units. Marglin makes much of the appearance of such manufactories in the woolen branch, arguing that this proves that "factory spinning did not depend for its success on a superior machine technology" and that supervision alone was enough to make such units competitive.5S Yet this was not part of a larger effort to impose discipline on wayward workers, and it is highly unlikely that hand spinning or weaving sheds by themselves could have competed with cottage labor. Rather such units were a response to the introduction of machines into the earlier, preparatory stages of manufacture (carding and subbing) and aimed at ensuring throughput to the heat-intensive, power-driven finishing pro- cesses (fulling, dyeing, printing). If clothiers were going to invest in factories at both ends of the production process, they wanted to be able to keep plant and equipment busy.51

50Marglin, "What Do Bosses Do?" p. 31. 51 In the West Country woolen manufacture, large workshops were almost invariably linked with

beginning and end processes that used power and heat, while adopting the new machines (carding and scribbling engines and the jenny) in sizes (eighty-spindle jennies) that exceeded the resources of the cottage worker. In this branch of the trade, manufacturers were slow to introduce water and steam power, using instead adult males (blind men sometimes) to turn the wheels; or horse gins, a dirtier and more troublesome source of power than the water wheel. There is considerable material in Kenneth Rogers, Wiltshire and Somerset Woollen Mills (Edington, Wilts., 1976).

Even in the more advanced Yorkshire woolen manufacture, the first mills did not appear until the 1770s and then multiplied as the technical problems of machine treatment of the recalcitrant natural fibers were progressively solved. See the tables and maps in D. T. Jenkins, The West Riding Wool Textile Industry, 1770-1835 (Edington, Wilts., 1975), pp. 15, 17, 20-21; also Jenkins, "Early

Page 22: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 605

The penetration of machines into the wool manufacture posed a problem to that sector of the industry that did its work in small, autonomous units rather than by putting out to dispersed cottages and that would therefore seem best to embody the Marglin ideal. This was the woolen (as opposed to the worsted) branch around Leeds and Wakefield in the West Riding of Yorkshire, and ever since Daniel Defoe described it in his classic Tour thro' the Whole Island of Great Britain, it stood as the archetype of small-scale, nonhierarchical, independent enterprise:

this Division of the Land into small Pieces, and scattering of the Dwellings, was occasioned by, and done for the Convenience of the Business which the People were generally employ'd in, and that, as I said before, though we saw no People stirring without Door, yet they were all full within; for, in short, this whole Country, however mountainous, and that no sooner were we down one Hill but we mounted another, is yet infinitely full of People; those People all full of Business; not a Beggar, not an idle Person to be seen.... This Business is the Clothing Trade.... almost at every House there was a Tenter, and almost on every Tenter a Piece of Cloth, or Kersie, or Shalloon.... Among the Manufacturers Houses are likewise scattered an infinite Number of Cottages or small Dwellings, in which dwell the Workmen which are employed, the Women and Children of whom are always busy Carding, Spinning, &c. so that no Hands being unemployed, all can gain their Bread, even from the youngest to the antient; hardly any thing above four Years old, but its Hands are sufficient to it self.52

Yet the Yorkshire woolen manufacture, as Defoe's description tells, was something less than the "boss-less" democracy it was reputed to be. As G.D.H. Cole writes in his introduction to the 1927 reprint edition, "Commentators have often written as if De Foe had described the country round Halifax as a paradise of prosperous artificers, each earning a good competence by the sale of his own products in the market at Leeds or Halifax, owning no master and treating no man as servant." "In fact," he goes on, "De Foe's West Riding manufacturer is not an independent craftsman so much as a small employer of labour, the scale of whose productive operations is still limited by the absence of power-driven machinery." 53

Factory Development in the West Riding of Yorkshire, 1770-1800," in N. B. Harte and K. G. Ponting, eds. Textile History and Economic History (Manchester, 1973), pp. 247-80.

52 Daniel Defoe, A Tour thro' the Whole Island of Great Britain .. . (1st ed., 3 vols., 1724-1726; reprint 2 vol., London, 1927), vol. 2, pp. 600-2.

53 Ibid., vol. 1, pp. xiv-xv. Even so, he was significantly different from the clothiers of the rapidly growing worsted branch around Bradford, who gave out all their work to dispersed cottage weavers, employed far greater numbers, disposed of much more capital, and were far quicker to move from putting-out to factory manufacture. On the divergent character of the two branches, see the excellent and well-documented article of Pat Hudson, "Proto-industrialisation: the Case of the West Riding Wool Textile Industry in the 18th and Early 19th Centuries," History Workshop, no. 12 (Autumn 1981), pp. 34-61. One caution: Hudson, like Houston and Snell (see footnote 24, above), makes much (too much?) of rectifying what she sees as an erroneous "proto-industry model," which is portrayed as linear: rural industry is essentially homogeneous and progresses normally and naturally to the next stage, factory production. Hudson clearly demonstrates that this is not so and (following in Braun's footsteps) that the agrarian context shapes both the forms of rural industry and its development.

Page 23: Landes - What Do Bosses Really Do

606 Landes

The small clothier in woolen, much though he loved the old ways, could not remain indifferent to the new technology. He had known and come to terms with power machinery in the form of fulling mills as far back as the Middle Ages. But these were always seen and used as complementary to the spinning and weaving that were at the heart of the production process. Only when mechanization began to touch these central stages did it pose a challenge to the older mode of production. The new carding and scribbling machines (used in preparing fibers for spinning) were incompatible with cottage or house shops because they were power-driven, took a lot of room while growing bigger from one generation to the next, and cost more money than a modest clothier could afford. Their capacity, moreover, exceeded the needs of the usual Yorkshire unit and would have constituted gross overcapitalization unless they could service a multiplicity of cottage shops. Which is exactly what the clothiers arranged to do: drawing on small accumula- tions of capital and borrowing on land held in freehold or copyhold, the clothiers pooled their resources to create what Pat Hudson calls company mills ("producer cooperatives" would perhaps be better) to provide them with these mechanized services. The participants in the co-op were expected to give it all their work; and the co-op in turn extended favorable terms of payment and credit in hard times.54

Thanks to this symbiosis, the small Yorkshire clothier with cottage weaving shop was able to survive into the second half of the nineteenth century. But one should not overestimate the efficiency and competi- tiveness of these enterprises. As machines got bigger and better, more and more cloth was made in factories, and increased supply pressed on prices, such that these tenacious, ingenious producers held out at the price of long hours, low wages, and increasing recourse to inferior materials.55 One defense of beleaguered small shops is an insistence on high quality and individuality in opposition to mass production. But the low road is also there, and even those who want to maintain standards are often forced to trade them for survival. Small bosses are bosses too.

No, what made the factory successful in Britain was not the wish but the muscle: the machines and the engines. We do not have factories

"Pat Hudson, "From Manor to Mill: the West Riding in Transition," in Maxine Berg, Pat Hudson, and Michael Sonenscher, eds., Manufacture in Town and Country before the Factory (Cambridge, 1983), pp. 124-44. Hudson notes that similar services were supplied, at least in an earlier period, by mills built and financed by landowners, who linked their operations to the needs of tenant clothiers and saw the combination of land, shop, and access to mill as a package designed to attract renters and enhance the revenue of their estates. The biggest such operator, Lord Dartmouth, owned nineteen mills in 1805 (p. 139). Such were some of the varieties of enterprise in an enterprising society.

55 On hard times in the woolen trade, see ibid., p. 143.

Page 24: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 607

until these were available, because nothing less would have overcome the cost advantage of dispersed manufacture.

But this implies an important question: Where were those early millowners to find their hands? Cottage workers, we know, were not ready to enter these establishments if they could help it-although the women and children who did the spinning and helping were more mobile, willy-nilly, than were the adult males who did the weaving. The problem is pertinent because of the later reliance on involuntary labor- that of young women and children, especially of parish apprentices assigned by poor law authorities as the equivalent of indentured servants. Marglin does not raise this issue, but his emphasis on the significance of institutionalized power in foisting "artificial" arrange- ments on an ostensibly or nominally competitive market system makes such a question relevant to an appreciation of the economic and commercial "validity" of the entrepreneurial role- "what bosses do."

To answer the question requires some knowledge of the technological sequence. The earliest inventors of textile machinery, in the late seventeenth and early eighteenth centuries, took it for granted that there would be more than enough free labor to work their devices. They gave no thought to the possibility of conscription; on the contrary, they worried (because others were worrying) about the eventual displace- ment of labor, and they sought to justify their projects by minimizing their expected impact on employment or by noting that their machines would provide work for otherwise idle hands: the poor, the very young (to be saved from mischief and vice), the blind and lame. With hindsight, to be sure, we look upon many of these as involuntary recruits; children, we have seen, did as they were told. But one has to put oneself in the mind of the day and recognize that it was not only their product that made such workers so attractive and deserving of concern; it was also the perceived philanthropy in furnishing tasks to those spiritually (even more than materially) imperiled by idleness.

By the 1760s, however, fears of technological unemployment were irrelevant: they were not raised, for example, against Richard Arkwright, "inventor" of the water frame (1768). The major centers of manufacture were suffering from an undersupply of yarn, and workers were being energetically wooed by manufacturers and intermediaries. An outrider who had under his command a large stable of productive spinners could and did advertise his wares and took his pick of customers. The introduction of the jenny (from 1767 on) into this kind of labor market did not initially pose a threat to domestic spinning. On the contrary, it increased output in the cottages, and there was a moment of hectic prosperity for those who could buy or rent the new device. We are talking of a short time-perhaps a decade.

The invention of the water frame did not initially change that picture, because the water frame made tough warp yarn, while the jenny turned

Page 25: Landes - What Do Bosses Really Do

608 Landes

out the softer weft. The two were thus complementary. But the water frame, as the name indicates, was powered from the start, hence used in factories; while the jenny rapidly grew bigger to where it was no longer suitable for small cottages. In 1779, moreover, the invention of the mule (so-called because it was a crossbreed of jenny and water frame) made it possible to spin weft with powered machinery. The first mules were small, with sixty spindles; but they were already more than even a strong man could drive. A new kind of work force was needed.

The first solution was to hire men. There was a pool of adult males used to working in branches where technology had already imposed cooperative labor in large units. Arkwright, for example, moved imme- diately (1768) from Lancashire, where workers were hard to find, to Nottingham, home of the hosiery manufacture. There he found capital, labor, and a friendly environment. For one thing, Nottingham did not spin yarn and had to import from outside, indeed from as far away as India; so there was no local opposition to the introduction of machines. For another, there was in the hosiery trade a population of spinners used to working with hand-driven machines (the knitting frame) under supervision, and it was among these stockingers that Arkwright found his first hands.

But this was no more than an interim arrangement. Two years later Arkwright set up an additional and much larger installation at Cromford (Derbyshire) on the Derwent, and this became the prototype for numerous competitive establishments, all of them planted along streams for their water power. This first generation of country mills was located in large part away from centers of population, hence far from the big labor pools. But population density was uneven, even in the seemingly empty countryside, and employers, as always, went after people: "I should go from north to south in search of a place that was the most inhabited, where I could have the greatest choice of workmen, and where I conceived labour would be cheapest. '56

In the water-frame branch of the industry, these "workmen" were largely women and children. Where were they to come from? In Derbyshire, we are told, they were found in the families of lead miners already working in the district: "When Mr. Arkwright established his works... ., he did not establish them where the people had been in the habits of spinning at all; but he established them at Cromford . .. where till that time the People [women and children] had been almost wholly unemployed, except in the washing of lead.' 7 In other instances-the most famous example is that of Samuel Oldknow at Mellor-families

56 Testimony of a leading Manchester fustian manufacturer to the House of Lords in 1785, cited in Stanley Chapman, "Workers' Housing in the Cotton Factory Colonies, 1770-1850," Textile History, 7 (1976), p. 118.

57 Ibid.

Page 26: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 609

were hired as units. Wife and children worked in the mill, and husband was occupied on roads or in transport or on a farm cultivated expressly because it furnished employment to the most expensive member of the team. But Oldknow also took casual (tramp) labor, and parish appren- tices from as far away as London.58 Increasingly, manufacturers found their readiest supply in poorhouses, which were only too happy to ''apprentice out" dozens of young boys and girls by way of relieving the ratepayers of their statutory burden. (Note that this was not a new practice: small, independent masters had been taking on parish appren- tices for years, thereby filling in the gaps in their home labor supply. It would be too much to expect every weaver to have the children he needed. And those who had more than they needed hired them out to neighbors as industrial workers and domestic servants.)

Here was a relatively cheap and stable work force. (I say "relatively" because these involuntary workers were not so careful and diligent as free agents; and numbers of them fled their indentures because of the hardships imposed by mean masters and hard supervisors.) There was a saving in capital as well: when these country mills were built in isolated, undeveloped places, it was often necessary for the employer to build lodgings for his employees, and these young people were normally housed in dormitories-more economical obviously than regular family housing would have been.

The construction of large jennies (up to eighty spindles by 1784) and the invention of the mule gave the key position in machine spinning to the adult male: the need to throw back the spindle carriage called for greater strength than women or children could furnish. The second generation of factory labor, then, placed less reliance on forced labor, especially after and perhaps even before legislation restrained access to and use of parish apprentices.59 Employers had to find workers in the free market, and one formula again was to hire family units: father to run the mules, children to help as piecers of broken threads, women to prepare the fibers for machine spinning. Such evidence as we have, however, admittedly local, indicates that only a small minority of

58 On the significance of tramp labor for the early textile mills, see M. M. Edwards and R. Lloyd-Jones, "N. J. Smelser and the Cotton Factory: A Reassessment," Textile History and Economic History (Manchester, 1973), p. 309.

59 Involuntary labor was not so productive as free labor, for understandable want of motivation, to say nothing of active hostility. See the fascinating testimony of James McConnel to the Children's Employment Commission of 1841-1842, in Parliamentary Papers, 1843, vol. 14, b-63. The firm of McConnel & Kennedy, then the largest in the kingdom, had acquired a country mill worked by parish apprentices and decided to offer them their freedom-to leave or stay on as free wage labor. The reason for this apparent benevolence was purely practical: these indentured young women were more trouble than they were worth-obviously unhappy, sensitive and quarrelsome in the smallest matters, ripe for rebellion. They sang, he said, revolutionary songs as they marched in ranks to and from the mill.

Page 27: Landes - What Do Bosses Really Do

610 Landes

children were employed or supervised by a relative. The bulk of the new work force, then, came from a more undifferentiated pool.60

Access to this pool was facilitated by the shift of the locus of manufacture: with the adaptation of Watt's steam engine to rotary motion, power production was freed of dependence on site and could move to the centers of population-Manchester first and Glasgow and other cities after. One effect of this move was the appearance of what we might call the multi-cellular mill: a building housing a number of spinning rooms, each a semi-independent fief run by a master spinner assisted by boys often linked to him by blood or neighborhood ties.

These master spinners were from one point of view workers-the core of what came to be known as the labor aristocracy. But they were also bosses, small subcontractors taking their share of surplus product. Following the Marglin method, we may well ask whether their role was in some way artificial-an imposition on a system that did not need them. It depends, I suppose, on the value that millowners placed on such delegation of functions and authority: it was obviously worth it to them to shift the responsibility for much of the hiring and firing, as well as the bulk of the supervision, to others. It is a system that reflects the limitations of early factory management and was eventually to disap- pear. But it lingered in the textile manufacture to the end of the nineteenth century; and in a field such as shipbuilding the so-called butty system was still dominant in the twentieth.61

' On labor recruitment for the early factories and the creation of a factory work force over a period of generations, see Frances Collier, The Family Economy of the Working Classes in the Cotton Industry, 1784-1833 (Manchester, 1964; this is the printed version of her M.A. thesis of 1921); and Neil Smelser, Social Change in the Industrial Revolution: An Application of Theory to the Lancashire Cotton Industry, 1770-1840 (London, 1959).

For a critique of the Collier-Smelser thesis, see Edwards and Lloyd-Jones, "N. J. Smelser and the Cotton Factory," pp. 304-19. Edwards and Lloyd-Jones used the 1816 report to the House of Commons to show that at Preston (Lancashire), only 11.6 percent of the children were employed by a parent, brother, or sister. (What about friends and neighbors?) In a recent article based primarily on the Lords' report of 1819, Herman Freudenberger, Frances J. Mather, and Clark Nardinelli ("A New Look at the Early Factory Labor Force," this JOURNAL, 44 [Dec. 1984], pp. 1085-90), come to the same conclusion: "the evidence suggests that individuals, not families, were the basic units of labor in cotton factories" (p. 1087).

61 Such arrangements reflect the inability of centralized management to extract "full" labor for wages: a small group bidding for the work and assigning tasks by consent, or working under the direct eye of a producer-entrepreneur, would deliver more product for the money. The shift to centralized control was a response to efforts by workers to constrain or withhold performance, for good and bad reasons, and called forth new, more effective techniques of supervision and record-keeping. On "monitoring" and similar matters, see Armen A. Alchian and Harold Demsetz, "Production, Information Costs, and Economic Organization," American Economic Review, 62 (Dec. 1972), a seminal theoretical analysis that has brought forth a stream of comment and reaction; also Williamson, Markets and Hierarchies, and "The Organization of Work"; and the thoughtful article by Gregory Clark, "Authority and Efficiency: The Labor Market and the Managerial Revolution of the Late Nineteenth Century," this JOURNAL, 44 (Dec. 1984), pp. 1069-83. Clark writes: "If radical economists are correct (that giving control of production to workers is more efficient) then, ironically, the efficient organization disappeared because workers used control of

Page 28: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 611

Could power-machine technology have succeeded had it not been possible to recruit forced labor? Here one has to distinguish between the forced labor of parish apprentices and the implicitly forced labor of dependent women and children. Apprentice labor was in one sense easier to recruit en masse, but left much to be desired from the point of view of productivity. In the beginning, ease of recruitment was probably more important, and the availability of such indentured servants added to the profitability of these enterprises. But had they not been available, the mills would have found what they needed, paid the price, and still made money because the carding machines, water frame, and mule were so efficient by comparison with hand labor. (One possible recourse was foreign labor: the Irish, for example, were not normally employed in English mills, but they could have been, as they were in Scotland, and at wages well below the English level.)62 What is more, this difference grew with astonishing rapidity as new machine builders entered the industry, flouted the efforts of such patent holders as Arkwright to protect their monopoly, and introduced that stream of small unpatent- able improvements that constituted the greater part of subsequent technical progress.

The steepness of this learning curve justifies, moreover, the argument that even if the first machines had not been dominant, the new technology would have made its way, with or without parish appren- tices. It is true that the failure of an innovation can and often does serve as a deterrent to pursuit of a given line of technical experiment; also that there are in principle alternative lines of advance, such that one can at least imagine the possibility of inventions more compatible with older forms of industrial organization, as indeed the early jennies were. Yet this is not, in my opinion, what would have and could have happened here. In the first place, technological advance does not normally proceed like a feedback robot, changing course at each obstacle. On the contrary, it typically takes the form of a series of probes, feeling out a particular line, withdrawing and revising the better to go forward, building and improving on previous failures; and this line of experiment is normally suggested by need and opportunity. This is in fact what happened in the eighteenth-century textile industry: the first devices (Wyatt and Paul) were built for wool, which proved relatively intracta- ble; then switched to cotton, at that time a far less important textile fiber. Even so they ran into all manner of technical difficulties, so that

production to drive up wages and limit output. If workers had not managed to exercise power within firms, they would have had more control over their production activities" (p. 1072).

62 There was good precedent for the employment of alien workers in large shops using new technologies: thus in the silk industry; and in the smallwares trade, in connection with the introduction of the Dutch loom, to the point where agitation by conventional weavers against this "develishe invention" took the form of xenophobic riots. Wadsworth and Mann, The Cotton Trade, pp. 101-2.

Page 29: Landes - What Do Bosses Really Do

612 Landes

the first enterprises failed. This did not stop numerous inventors from plugging away at what was clearly a major bottleneck, until success was finally achieved in the 1760s. This is not to say that some people did not feel discouragement; even Arkwright had his moments of disbelief. But so keen was the interest and so great the supposed advantages that there was no lack of mechanics to try their hand or capitalists to advance funds.

Secondly, the nature of the techniques tried is to a large extent dictated by the task undertaken and the character of known devices. Technology has its preferences. In the case of textiles, the models were furnished by the existing spinning wheels, and the aim of the machine was to multiply these in such fashion as to allow a single person or, better yet, an engine to work many spindles at once.63 The new and hardest part of the task was somehow to imitate the motion of the human hand drawing and twisting the thread. The two systems eventu- ally employed were already in use on the hand machines: the older and more delicate (the jenny) imitated the distaff by turning the spindle while letting the thread come off its end; the second (the water frame), quicker but rougher, ran the thread to the spindle through a flyer rotating at a different speed; the difference imparted the twist. It was this legacy that set the terms of the innovations and opened the door for further improvements ofjust that sort that made factory manufacture dominant. Once the problem had been solved for more than one spindle, it was essentially solved for a hundred or a thousand. And so it was.

This brings me to a further major point of disagreement with Marglin: the contribution of the factory mode to technological change. This has always been something of a given, and it is useful to have Marglin question the conventional wisdom and compel us to think about the issue. Marglin is not prepared to concede intrinsic innovative superior- ity to the factory. Insofar as it furnished a greater stimulus to invention, this too was nothing more than an artifact: "the factory's superiority in this domain rested in turn on a particular set of institutional arrange- ments, in particular the arrangements for rewarding inventors by legal monopolies vested in patents."64 Nothing, says Marglin, compelled British society to reward inventors in this manner. There are other

63 There were several precedents that served as models: thus the Dutch loom or engine loom, invented in 1604 and used in the manufacture of smallwares (tapes, ribbons, garters, and such), though a weaving rather than a spinning machine, set the example of multiplying the working parts by replication so that one motor could drive them and one person could make a dozen or more pieces at a time. The earliest, experimental spinning devices followed a similar logic: see the patent granted to Richard Dereham and Richard Haines in 1678, which promised a device that would drive six to a hundred spindles, by human power no less (ibid., pp. 98-106, 413-14).

4 Marglin, "What Do Bosses Do?" p. 33.

Page 30: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 613

incentives-prizes, for example-and these might have tilted the bal- ance by encouraging those inventions useful in private workshops and cottage industry. "Had the patent system not played into the hands of the more powerful capitalists by favouring those with sufficient re- sources to pay for licenses (and incidentally contributing to the polar- ization of the producing classes into bosses and workers), the patent system need not have become the dominant institutional mode for rewarding inventors."65

Most of this is simply wrong. It rests on two assumptions. The first- that patents were indeed an incentive to invention-has never been proven empirically, at least not to my satisfaction; but it does seem logical and plausible. The second assumption-that it is easier to pay for and capture gains from patents in factory manufacture than in home and shop industry-is more troublesome. Like the first, it is based more on a priori reasoning than on empirical evidence. On the basis of what I know of one handicraft branch, the clock and watch manufacture, the expected difficulty of recovering royalties-or for that matter, the cost of obtaining a patent-does not seem to have deterred inventors. On the contrary, this is a distinctively creative branch, full of invention and emulation, and the spur of fame seems to dominate pecuniary self- interest. Precisely because it is so inventive, moreover, patents have limited value: watchmakers, like economists, have always known that there is more than one way to skin a cat. When Thomas Earnshaw invented in 1783 the mechanism that was to become the standard controller for precision chronometers, he could not afford the hundred pounds sterling needed to secure a patent. But this did not stop him from boasting of his device and, when he had talked too much and found himself in danger of being preempted, he found a wealthier watchmaker to secure the patent for him. Earnshaw, of course, had to reimburse his stand-in. He did this by making a series of watches using the new escapement and letting the stand-in sell them under his own name; and other watches for still other "makers," who paid a royalty of one pound per watch. Meanwhile those unwilling to pay Earnshaw had the option of using a similar, imitative device, invented by Samuel Peto precisely for this purpose. Earnshaw was contemptuous of his rival, who had gone, said he, to a lot of trouble to get in by the back door when the front door was already open. He was right about that, but that was the whole point: the Peto cross-detent escapement was an alternative.

What are we to infer from this story? That patent rights were worthless? Or that they were collectible? Or both? The world of watchmaking was a small one: it was hard not to know what other people were doing, and evasion of patent rights may have been more difficult than in other trades. On the other hand, the watch is a small and

65 Ibid., p. 33f.

Page 31: Landes - What Do Bosses Really Do

614 Landes

highly personal object, and one could argue that it is precisely the kind of commodity that lends itself to concealment and evasion. Certainly it would have been easier to defraud a patent on the hidden movement of a pocket timepiece than, say, on steam engines built to drive mills. Yet we know that Boulton & Watt were engaged in a ceaseless effort to enforce their patent rights, and the evidence indicates that about one out of every two horsepower produced at the end of the eighteenth century was generated by illicit engines. We also know that James Watt himself had to find an alternative to a preemptive patent on a crank transmission for converting reciprocating to rotary motion. It is hard to believe that the British patent office, sometimes sticky to the point of unreason, was ready to grant protection to a device that went back to the Middle Ages. Perhaps Watt could have fought the patent successfully on the ground of unoriginality. But it was easier to bypass it, and so he did with his sun-and-planet gearing.

On balance, patents were not the major incentive to invention. The biggest and surest source of gain was the application of invention in one's own enterprise-a Schumpeterian headstart-and the role of the patent, if obtainable, was to discourage other people from using the same or similar techniques. But there were limits to discouragement: every successful innovation called forth imitations, as the author of a history of Nottingham cheerfully recognized: "Various patents have been obtained . . . for making warp lace; but so numerous are the pretensions set up for the invention of different movements and various formed meshes . .. that the patentees and other inventors mutually laugh at and invade each other's schemes."66 The result was a criss- cross of litigation that made at least the lawyers rich. But that was the point: to increase the cost of imitation and protect those headstart gains. Meanwhile a large part, perhaps the greater share, of productivity increases in factory manufacture was the result of the accumulation of small, unpatentable improvements-so unprotectable that it often paid, as it did in iron and steel, simply to make them available to competitors in return for reciprocity.67

In those instances where such altruism was not encouraged or commanded by reason-in those arts and crafts, for example, that had once been known as mysteries-patents were not always the best way to protect knowledge. Instead, inventors preferred to try and keep

I J. Blackner, History of Nottingham (Nottingham, 1816), cited in S. D. Chapman, "Enterprise and Innovation in the British Hosiery Industry, 1750-1850," Textile History, 5 (Oct. 1974), p. 25.

67 On the importance of unpatented invention in advancing technique in the British iron industry, see Robert C. Allen, "Collective Invention," Journal of Economic Behavior and Organization, 4 (1983), pp. 1-24. Allen lays stress on information-sharing within the industry: "If a firm constructed a new plant of novel design and that plant proved to have lower costs than other plants, these facts were made available to other firms in the industry and to potential entrants." And he offers some valuable insights into the material advantages of altruism.

Page 32: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 615

devices and techniques secret, sometimes by so dividing the process that no one worker could penetrate the technique. This is what the great French watchmaker Abraham-Louis Breguet proposed to do when he planned the mass production of watches by means of power tools and interchangeable parts: the aim was not discipline (for that, one could hire married women), but security.68

It is not easy to sum up such impressionistic observations, which need the underpinning of empirical research. If I were to reason a priori, I would argue that patents were probably more useful in putting out than in craft shops, manufactories, or factories, simply because secrecy was unenforceable in dispersed manufacture. As for the chances of main- taining confidentiality in shops large or small, the better the secret, the greater the temptation to exit and betrayal. Besides, a valuable secret implies growth potential: a small enterprise that made a better mouse- trap would have been hard pressed to keep from growing into something bigger, perhaps a factory. A factory, on the other hand, could go a long way on its own unpatented improvements. This, more than anything else, explains the factory bias of technological change: that was where the money was. For one thing, the saving in labor costs was greater, because factory wages were higher. For another, the accumulation of small, incremental improvements was a function of the volume of investment: new plant meant new and often better equipment.69 Be- cause of the constant interaction, moreover, between work force and supervision, the factory environment was more favorable to the percep- tion of possible improvements and their introduction. One final consid- eration: the logic of technology was moving toward ever-wider mecha- nization, toward doing more and faster, thereby enhancing the advantage of mass production and the factory system. Each success invited others-first, by showing the way and, second, by creating an ever-changing pattern of bottlenecks that promised new rewards to innovation.

Here the role of the entrepreneur was crucial-another point of disagreement with Marglin. No one else was in so strategic a position to see the needs and opportunities of technological change, for no one else was in a position to look both upstream and downstream, as well as to competitors on either side. Not all capitalists played this role well; that's

58 France, Archives Nationales, F'2 1325A, memorandum of A. L. Breguet to the Controller- General, 6 Sept. 1786.

69 See Allen, "Collective Invention," pp. 13-14; Jacob Schmookler, "The Level of Inventive Activity," Review of Economics and Statistics, 36 (May 1954), pp. 183-90, and "Changes in Industry and in the State of Knowledge as Determinants of Industrial Invention," in Richard R. Nelson, ed., The Rate and Direction of Inventive Activity: Economic and Social Factors (Princeton, 1962), pp. 195-232.

Page 33: Landes - What Do Bosses Really Do

616 Landes

what business competition is all about. But many, including those who lacked the technical know-how to prescribe remedies, could and did see possibilities that others could not; and where necessary, they hired the technical help they needed. Note that sometimes the role of strategist fell to others than the owners of enterprise-to managers or, as in late-nineteenth-century Germany, to those investment bankers who were bankrolling the operation and had access to a wider body of information than even the best-trained managers had. In socialist economies, technicians and planners play a similar role. But role there is, and one has to be placed high enough to have the wide view that is not visible on the shop floor.

As against this positive view of entrepreneurship and hierarchy, Marglin's is clearly disapproving. This transpires in his choice of words: the term cognoscenti, which stands for capitalists in the 1984 version of his model, "is perhaps an unfortunate shorthand, for ambition and greed, single-mindedness and disregard for the needs and concerns of others, adventurousness and rootlessness, ability to command and difficulty in relating outside the hierarchical model: all these play a role along with technical knowledge in shaping the ability to organize production.'"70 Now I am not prepared to make an issue of any of these alleged characteristics, except perhaps for rootlessness, which is not true for many entrepreneurs and in any case is a word that the Soviets and their lackeys in Eastern Europe have attached to those groups such as the Jews who would like to leave that working-class paradise. Why don't we try "mobility"? Or "responsiveness to opportunity"? Be that as it may, it is certain that just about all entrepreneurs have been characterized by one or another of these vices (virtues?), and that some have been characterized by all of them. But in generalizations of this kind, tone matters. One could, for example, think of another set of entrepreneurial characteristics: ambition (if it were so, 'twere a grievous fault) and aspiration; desire for self-enrichment by serving others, whether by job creation or the production of socially valued goods and services; an urge to build monuments more lasting than bronze; the habit of command; alertness and flexibility; intelligence and ingenuity; diligence, dedication, thrift, high seriousness, and those other Weberian virtues; et cetera, et cetera. Or, to take the stick from the other end: Marglin writes, as part of his condemnation of selfish individualism, that "in a culture in which the dominant value is every man for himself, only the most exceptional among the upwardly mobile will stop at self- sufficiency. "71 One might as easily say that only the laziest among the upwardly mobile would so behave; or the most easily satisfied; or the

70Marglin, "Knowledge and Power," p. 151. 71Ibid., p. 150.

Page 34: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 617

fools; or the saints; or the would-be saints. And there would be truth in all of these.

Clearly Marglin does not like capitalists as a type or entrepreneurship as a role. Who needed these people? What for? To supply capital? It is true, he says, "that machinery was too costly for the individual workman, and the group was to all intents and purposes nonexistent. But before that time, machinery was not prohibitively expensive, and since then the [labor] union has become a force that might have offset the high cost of machinery-for the group if not for the individual."72 But the union, he says, is part of the problem. It has preferred the comparative peace of negotiation within the existing arrangements, and has accepted hierarchy and division of labor because this is the easiest way to operate in a growing, changing economy. If I may put words in Marglin's mouth, labor has sold its birthright for a mess of transactional pottage.

Let us not blame labor. In the last analysis, it is the capitalist who is not forthcoming. "[My] model," Marglin writes in his second essay,

provides a response to the question of why capitalists are not in general content to act as rentiers, leasing capital goods to workers. In the present model, the essence of the capitalist's contribution is not capital, but organizing ability; to secure a reward for this service, he must impose an organizational form that makes him essential to the production process on a continuing basis. A rentier class is a class ripe for expropria- tion.73

Such a statement poses questions of both fact and interpretation. The historical record shows that capitalists have always been ready to lease equipment to workers when it pays, that is, when leasing yields a secure income. One has only to think of the Nottingham hosiers and the knitting frame in the seventeenth and eighteenth centuries; or of lace manufacturers in the nineteenth, who sold machines worth hundreds of pounds on the installment plan and rented factory space and steam power to the purchasers; or of truck-leasing arrangements today.74 As for expropriation, experience shows that when political change permits or brings about expropriation, no capital is safe, whether operated by the owner or leased to someone else. To be sure, some forms of capital are more mobile than others and flee when storms approach; while others cannot move, so that the capitalist is easy game. The story of rent control is a good example of legalized confiscation and the involuntary

72 Marglin, "What Do Bosses Do?" p. 25. 73 Marglin, "Knowledge and Power," p. 150. 74 On the lace industry, see the introduction by Stanley D. Chapman to Felkin's History of the

Machine-Wrought Hosiery and Lace Manufactures (Newton Abbot, 1967), p. xxviii. Cf. the comparable transfer of ownership (and risk) in the American silk manufacture-a remarkably effective way of turning contentious workers into motivated capitalists. Philip J. McLewin, "Labor Conflict and Technological Change: The Family Shop in Paterson," in Philip Scranton, ed., Silk City: Studies on the Paterson Silk Industry (Newark, 1985), pp. 135-58.

Page 35: Landes - What Do Bosses Really Do

618 Landes

transfer of wealth from one group to another-"what's yours is mine" economics. What matters, in other words, is the political and legal context. Capitalists, as Marx understood full well, are in it for the money. They have no philosophical or moral or practical objection to worker-managed enterprises per se, as their readiness to unload unsuc- cessful plants at bargain-basement prices shows.

Nor do they have some kind of collective repugnance for rentier status, whether for fear of expropriation or other reason. On the contrary, the technological changes that have occurred over the last century have devalued the role of the capitalist, shifting much of his entrepreneurial contribution to managers and technicians. This has dramatically altered the lines of control in business enterprise, to the discomfiture of many employers now reduced to the role of absentee owners. But this has not stopped capitalists from investing their fortunes in new techniques: they understand as well as anyone that on the management level as on the production level, division of labor and specialization are concomitants of growth. Insofar as some capitalists have opposed this trend, they have failed or been ousted for poor performance. Their own fellow capitalists have joined in firing them or retiring them or buying them out. So much for models, such as this one, that reify a capitalist class and present it as moved by some unitary sense of its function and needs.

By way of consolation, I do not think that small-scale industry on a nonhierarchical, nonspecialized basis was a real alternative, at least not in the economic sense. On this point it is instructive to reread the report in 1806 of the House of Commons Committee on the Woollen Manu- facture.75 The Committee noted that the woolen industry showed three principal forms of organization: factory, putting-out, and small-scale "independent" weaving. The last, we saw, was as close as one might get to Marglin's worthy ideal. These were small units, employing at most a few journeymen, and where necessary, subcontracting for those operations that could be done economically only in a larger, powered installation. What is more, entry was fairly easy:

as it has been expressly stated to Your Committee, a young man of good character can always obtain credit for as much Wool as will enable him to set up as a little Master Manufacturer, and the public Mills, which are now established in all parts of the Clothing District, and which work for hire at an easy rate, enable him to command the use of very expensive and complicated Machines, the construction and necessary repairs of which would require a considerable capital. Thus, instances not unfrequently occur, wherein men rise from low beginnings, if not to excessive wealth, yet to a situation of comfort and independence.

75 Parliamentary Papers, 1806, vol. 3 (268). The next three passages quoted are from pp. 10, 12, and 13.

Page 36: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 619

The Committee took pleasure from the advantages such a system offered to health and morals, especially by contrast with the factory system. But they also noted the technological and commercial advan- tages of the large enterprise:

for, it is obvious, that the little Master Manufacturers cannot afford, like the man who possesses considerable capital, to try the experiments which are requisite, and incur the risks, and even losses, which almost always occur, in inventing and perfecting new articles of manufacture, or in carrying to a state of greater perfection articles already established. He cannot learn, by personal inspection, the wants and habits, the arts, manufactures, and improvements of foreign countries; diligence, economy, and pru- dence are the requisites of his character, not invention, taste, and enterprise.

And they recognized that many of the small clothiers felt threatened by the new mode of production.

Even so, they were not prepared to recommend government inter- vention on behalf of what they called the domestic system. Freedom of investment was the constitutional right, indeed the birthright, of every Briton, the basis of national prosperity. Commerce and manufactures had always "flourished in free, and declined in despotic countries." Besides, they wrote, factories contributed to the well-being of the domestic system by furnishing new designs and qualities:

It is besides an acknowledged fact, that the Owners of Factories are among the most extensive purchasers at the Halls, where they buy from the Domestic Clothier the established articles of manufacture, or are able at once to answer a great and sudden order; while at home, and under their own superintendence, they make their fancy goods, and any articles of a newer, more costly, or more delicate quality, to which they are enabled by the Domestic system to apply a much larger proportion of their capital.

In other words, symbiosis. But it was an asymmetrical relationship. One partner was riding the wave of technological change; the other was trying to keep his head above water. The Committee made its report just about the time the power loom was making its presence felt in cotton. Wool still had some decades to go. But the same techniques that worked in cotton would eventually be perfected in wool, and once that hap- pened, the independent clothiers would be forced into the shrinking upper reaches of the market; or they would serve as mere auxiliaries of the- factory system. Their fate, like that of the handloom weavers in cotton, was a calvary of independence in dependence.

The workers themselves understood this: Eric Hobsbawm, explaining the failure of socialism to capture the imagination and loyalty of mid-nineteenth-century workers, points out that the chief reason was "the very advance of the British capitalist economy over the rest, which already made an economy of small commodity producers, individual or

Page 37: Landes - What Do Bosses Really Do

620 Landes

collective, somewhat implausible or economically marginal."76 Division of labor, in other words, with specialization and hierarchy, was there to stay, at least in those major branches producing in quantity for mass consumption. No complex operation could be effected at low, compet- itive cost without such arrangements. Marglin recognizes this when he asks whether factory employment was "better than alternative forces of productive organization that would have allowed the worker a measure of control of product and process, even at the cost of a lower level of output and earnings."" Only a political decision could have inverted the pressures toward low-cost mass production.

Here Marglin and I come together again, though not comfortably. His main point is that the key to an understanding of technological choice is power, that is, political power; and that this applies as much to socialist as to capitalist economies. Technology, he says, is not autonomous (his word is "exogenous") and inexorable. Man makes technique and then chooses; or dictates what techniques to look for. I think this is right, up to a point. That point is, the choices are not accidental. As I implied above, there is an inherent logic to technological change, which is governed by the law (condition) of minimization of inputs; or con- versely, of maximization of output. The aim is to get the most for the least. This is not to say that innovations that satisfy that condition are intrinsically good or "better"-to use Marglin's word. They just pay more-in money and power, depending on the economic and social system. This is why socialist countries seem to obey the same techno- logical tendencies as capitalist nations: they define their goals differ- ently, but they also want more for less.

If, then, we want to abrogate this condition, we have to make a deliberate, collective decision to do so; and in some areas we do. But we

76E. J. Hobsbawm, "Artisan or Labour Aristocrat," Economic History Review, 2nd ser., 37 (Aug. 1984), p. 362.

77 My italics; Marglin, "What Do Bosses Do?" p. 37. In his 1984 essay, Marglin admitted that he had in fact underestimated the ability of labor to retain some control over the work process and pace, even in a factory environment. See "Knowledge and Power," p. 164, n. 12. But it must be recognized that historically, much of this countervailing power has been exercised by a small fraction of the work force, composed of those highly skilled craftsman whose dexterity and experience remained valuable if not indispensable even in the presence of machinery, and who were therefore in a position to open or shut the flow of work. Much of the sharpest labor conflict of the late nineteenth and twentieth centuries has focused on attempts of management to free itself of this constraint by substituting capital for labor and thereby deskilling these specialties. Even so, the organized power of such crafts was such as to maintain their leverage even in the absence of technological indispensability. But as Hobsbawm points out in "Artisan or Labour Aristocrat" (pp. 367-68), this was a precarious power base, the more so as these privileged workers often set themselves off from, and sometimes against, the unskilled remainder of the work force. To paraphrase Marglin, otiose property rights are an invitation to expropriation. For a recent study of this battle in the American machine-tool industry, with special reference to General Electric, see David F. Noble, Forces of Production: A Social History of Machine Tool Automation (New York, 1984), and my review, "Machines and Their Masters," in The New Republic, Nov. 19, 1984, pp. 37-41.

Page 38: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 621

should not kid ourselves about such important matters. Technology does have its own reasons, and we ignore them at a price. We ought to know the price before we buy the goods.

The second of the questions I posed at the start of this essay was this: Where is Marglin coming from? He has to be situated, I think, in a larger stream of criticism of the structures and conditions of work going back to the Industrial Revolution itself. The criticism is multiple: a denunci- ation of the exploitation of labor (not only in the Marxist sense, which is tautological, but in the real sense of extracting labor on imposed terms); an attack on specialization and the consequent alienation of the worker from his work; a rejection of hierarchy, which lays the economic foundation of social and psychological inequality.

This multiple criticism has as its counterpart a dream and a program. The dream: Why can't things be better? Why can't work be enriching and fulfilling? Why can't workers do what they enjoy doing and receive a wage appropriate to their performance, or more ideally, to their need? Why can't all of us be equal and run our collective activities as equals? The program? Well, that necessarily varies. It is easier to find fault than remedy, easier to unite around a negative banner than an explicit set of goals.

In the last twenty or so years, these utopian aspirations have been renewed and have found expression in economics in a growing corpus of revisionist literature on the structure of enterprise. Most of this writing is of a theoretical character-as are most of the neoclassical analyses. The authors operate in the traditional economoneiric mode: markets are competitive; workers, like bosses, are rational maximizers; all workers are potentially equal, though they may differ in strength and ability; management is just one more skill (here radicals and business schools come together); if capital can buy it, so can labor; and so forth. That kind of theoretical speculation is intellectually stimulating and sugges- tive, but less than useless if not combined with empirical observation, preferably over a long experience.

Here lies the merit of the Marglin contribution: he has sought to study history and enlist it in his argument. To be sure, his history is not what it should be, as I have tried to show; and there are those historians who would ignore the discussion on that account alone. Yet this essay is itself evidence that I think that that is a mistake. For one thing, however incomplete or inaccurate Marglin's history, his concerns and intent must be given serious consideration. He expresses, as we have seen, discontents of long pedigree and considerable justification, and nothing that old and tenacious should be taken lightly. For another, it is the historian's job to respond. His primary task is to demythify the past and prevent others from mythifying it, to come as close as possible to telling

Page 39: Landes - What Do Bosses Really Do

622 Landes

it like it was. Of course, one can never be sure how close one has come to such an ideal. History is too complex to allow us the kind of near-certainty available to some of the natural sciences; and no histo- rian is exempt of bias. The combination can be damaging. Under the circumstances, the best one can hope for is a high degree of plausibility and persuasiveness. The historian is like the attorney addressing a jury-whether the jury of other historians or a larger public. He makes the best case he can, musters the best evidence, reasons by inference whenever possible, controls for alternatives. And then he waits for the verdict, which comes in over a long period of time.

Which brings me to the last of the questions I posed at the start of this essay: So what? By this I meant: even if Marglin were right about the artificial, prestidigitarian origins of industrial capitalism, would it then be "an open question whether or not hierarchical production is essential to a high material standard of living"?78 In other words, would that justify us in thinking that we can give up hierarchical technology today-as Marglin points out, in both capitalist and noncapitalist systems-at little or no cost?

My own logic (I'm not talking about history now) tells me that the answer is no. There is nothing in the character of a previous event or decision-the choice of one path rather than another-that implies reversibility. Even an accident changes the future irremediably. Per- haps we should never have taken the road that led us through special- ization and putting-out to the factory system and mass production, but we did and have gone a long way; and we cannot turn the clock back. We can take steps to improve such arrangements as we have inherited: to increase worker participation in management decisions, to diversify tasks and rotate assignments. But we are not going to abolish degrees of responsibility and authority in any operation of any size. Even in small groups committed to absolute equality, it takes a tenacious effort and much euphemism to repress or conceal distinctions. One very popular device that we are all familiar with is the effort to paper over differences by using more dignified designations for what are perceived or defined as the lower rungs of the ladder. But these inventions do not change the substance: in the long run, they only add to the vocabulary of pejorative designations.

The difference at bottom between him and me, I suspect, is not so much that we disagree about work structures and discontents, though we do, at least in part; or about the role of the capitalist entrepreneur, which we do; or about the character and distribution of individual and social virtue and vice, which we do; but rather that we diverge on issues of social and political remedy. He is an economic theorist, and to theorists what is conceivable is possible. I am an historian, and

78 Marglin, "What Do Bosses Do?" p. 14.

Page 40: Landes - What Do Bosses Really Do

What Do Bosses Really Do? 623

historians tend to be disenchanted by the record of human experience. We are suspicious of promises. Employers (capitalists) useless? Spe- cialization useless? The bigger the promise, the more suspicious we get.

In the last analysis, then, I see Marglin's essays as exercises in optative economics, useful for the historical questions they raise, but directed primarily to true believers as a vision of what might have been and ought to be. His aim, implicit if not explicit, objective if not subjective, is to delegitimize the capitalist today-hence the present tense of his title-and thereby to encourage and justify some unspecified act of expropriation. That is in the nature of revolutionary propaganda: to accentuate the negative. As Moses Hess, then editor of the socially furious Neue Rheinische Zeitung, replied to one of his bourgeois backers (who else?) who had complained about the pessimistic tone of the newspaper, that was the point: "Alles in Bewegung" -everything in motion. So here: first hold out the promises. There will be time enough for reckoning later.

But as every good economist knows, there is no such thing as a free utopia.