¹ See for instance Postan, M. M. in Postan, M. M. and Habakkuk, H. J., eds., The Cambridge Economic History of Europe, I, (2d ed.; Cambridge, 1966), 571Google Scholar. This series will henceforth be referred to as Cehe.

² I use this term, as economists do, to signify a set of functional relationships that yield different results in different circumstances; the purpose of a model is thus not to ignore variety but to account for it. This notion is of course quite different from that of a model as a mere simplified description or archetype, which is to the preceding as “everyone eats beef is to “beef consumption depends on incomes and prices.”

³ See for instance Duby, G., Rural Econom and Country Life in the Medieval West (Columbia, S.C., 1968), pp. 263–264, 341Google Scholar; Miller, E., The Abbey and Bishopric of Ely (Cambridge, 1951), p. 105Google Scholar; Postan, CEHE, I, 581; Smith, R. A. L., Canterbury Cathedral Priory (Cambridge, 1943), pp. 114, 141Google Scholar; Titow, J. Z., English Rural Society 1200–1350 (London, 1969), p. 43Google Scholar.

⁴ North, D. C. and Thomas, R. P., “The Rise and Fall of the Manorial System: A Theoretical Model,” Journal ofEconomicHistory, XXXI (1971), 777–803Google Scholar; see also The Rise of the Western World (Cambridge, 1973), pp. 29 ff., 38 ff., 59 ffGoogle Scholar.

⁵ Fenoaltea, S., “The Rise and Fall of a Theoretical Model: The Manorial System,” Journal ofEconomicHistory, XXXV (1975), 386–409Google Scholar.

⁶ In this discussion, “the lord” should be understood to include the lord's agent or the firmarius. I am not here concerned with the issue of “farming” the tenants' rents or the demesne, though I shall touch on it below.

⁷ The sanction may be a termination of the employment agreement, or a suitable fine; for further discussion see my “The Rise and Fall,” n. 10.

⁸ The specification of individual duties—pushed to an extreme, for instance, in the servile familia of the ancient estate—can thus be seen to reduce supervision costs by narrowing responsibilities. When work is individual, and the quality and quantity of output readily ascertained, the worker can be paid by the piece; his incentives are then those of an independent worker (tenant), and not or an employee.

⁹ Because the abuse of animals does not produce immediately apparent changes in their condition, individual contributions to their care are most easily evaluated by making a single individual responsible for them over extended periods. Because animal care is a continuous job, and because it was thus best kept undivided, labor dues systems were rarely found without the admixture of a staff of bovarii and the like whose subsistence was produced for them. See M. M. Postan, The Famulus: The Estate Labourer in the Twelfth and Thirteenth Centuries (Economic History Review supplement no. 2, 1954).

¹⁰ See my “The Rise and Fall,” pp. 392–94.

¹¹ “[Dopsch has demonstrated that] demesne fanning and independent cultivation with money rents exist side by side from Carolingian times on; rented land was already more extensive than demesne land in the Carolingian period …” E. R. Wolf, “The Inheritance of Land among Bavarian and Tyrolese Peasants,” Anthropologica, new series, XII (1970), 106.

¹² See Postan, M. M., “The Chronology of Labour Services,” Transactions of the Royal Historical Society, 4th series, XX (1937), 169–93CrossRefGoogle Scholar.

¹³ Transaction costs—and specifically the costs of supervising dependent labor— also appear to account for the scattering typical of peasant holdings; see my “Risk, Transaction Costs, and the Organization of Medieval Agriculture,” Explorations in Economic History XIII (April 1976), part VGoogle Scholar.

¹⁴ Note that the fundamental (voluntary or extortionate) nature of the arrangement between lord and peasant is here immaterial: we can consider the lord a pure tax-collector, and examine the alternative of taxing the peasant in goods (rent) or in time (labor services, if the tax is less than the surplus, and wage work, if it is more than that). The transaction-costs problem is unchanged, whether in the absence of the tax the peasants consume their surplus over subsistence as extra goods or as free time: if the tax is collected in goods, the peasant is automatically motivated to produce them (if need be by giving up free time to restore his after-tax output, and consumption, to a minimum subsistence level); if instead the tax is collected in time, that time must then be turned into goods by somehow “artificially” motivating the workers to be diligent—and that cannot be cheaply done.

¹⁵ I am assuming that individuals behaved “rationally,” by which we mean only that they would choose the option they preferred from among those open to them. I am also assuming—and here there is rather more room for meaningful disagreement —that medieval men were motivated much the way we are; while this assumption is perhaps more natural to economists than to historians, even the latter might wish to conclude from this paper (and from its counterpart on “Risk, Transaction Costs”) that it cannot be excluded out of hand.

¹⁶ In the case of monopoly, instead, the deadweight loss is established only by formal analysis; and even those who know it typically lack the opportunity to strike an efficiency-producing bargain with the monopolist.

¹⁷ See however the example of (skilled) workers renting space in a factory (and selling their goods rather than their time) cited by Engels, F., The Condition of the Working Class in England, Henderson, W. O. and Chaloner, W. H., trans, and eds. (Oxford, 1958), p. 224Google Scholar.

¹⁸ Migrants who consider themselves temporary may well try only to sell their labor, rather than acquire immobile capital even of a familiar type; coolie labor on tropical plantations appears to be the major example of this type in agriculture.

¹⁹ On Roman agrarian institutions see for instance Gras, N. S. B., A History of Agriculture (New York, 1925), ch. illGoogle Scholar, and C. E. Stevens, CEHE, I, ch. ii; note also Bloch, M., French Rural History (Berkeley, 1966), pp. 67–68Google Scholar.

²⁰ This is reported by Tacitus in ch. xxv of the Germania.

²¹ L. J. M. Columella, Res Rustica, Book I, vii. 3. Chapters vii and viii provide an Eloquent description of the incentive problems associated with agriculture by dependent labor, as well as of the even greater costs of technical suboptimality. See also Gras, A History of Agriculture, p. 62.

²² Bloch, French Rural History, p. 68; also Duby, Rural Economy, p. 221.

²³ In later centuries, it was common for tenants' labor obligations to be sold or commuted, and the demesne worked by wage-eamers. Such an arrangement is only a juxtaposition of unsupervised and wholly supervised work; it is quite distinct from the labor-sharing system considered here.

²⁴ A given amount of manure, for instance, would have increased total output more by being distributed over all the cultivated area than by being concentrated on part of it; the increased output of the villeinage could be captured as rent, and it would more than offset the decline in the output of the demesne. The critical notion here is that of opportunity cost; it is all too often ignored in the discussion of these arrangements.

²⁵ Because tools and animals (at least with “their” servant) can readily be transferred, the qualities of the soil appear to be the primary fixed factors of production.

²⁶ For references to demesnes as consolidated and/or as recent assarts throughout the Middle Ages, see for instance Duby, Rural Economy, pp. 47, 182, 200, 266; Gras, A History of Agriculture, p. 85; Ernie, Lord, English Farming Past and Present, 5th ed. (London, 1936), p. 28Google Scholar; Fossier, R., La terre et les hommes en Picardie (Paris, 1968), p. 217Google Scholar; Harvey, P. D. A., A Medieval Oxfordshire Village (Oxford, 1965), p. 20 ff.Google Scholar; Hilton, R. H., A Medieval Society (London, 1966), pp. 19, 23Google Scholar; and Seebohm, M. E., The Evolution of the English Farm (Cambridge, Mass., 1927), p. 162Google Scholar. T. A. M. Bishop also relates consolidated demesnes to the use of dependent labor; see “Assarting and the Growth of Open Fields,” Economic History Review, VI (1935), 21, and n. 13 aboveGoogle Scholar.

²⁷ See for instance Duby, Rural Economy, p. 90; Ernie, English Farming, p. 8; Harvey, A Medieval Oxfordshire Village, pp. 46 ff., 130; Smith, Canterbury, p. 41; also Postan, CEHE, I, 600 ff., and Ch. Parain, Ibid., p. 159 ff.

²⁸ See Ault, W. O., Open-Field Farming in Medieval England (London, 1972), pp. 15–16Google Scholar; van Bath, B. H. Slicher, The Agrarian History of Western Europe (London, 1963), p. 61Google Scholar; Maitland, F. W., Domesday Book and Beyond (Cambridge, 1897), p. 149Google Scholar; and Postan, CEHE, I, 572–573. Differences in crop types between adjoining strips would, however, create negative externalities and an incentive to consolidate the demesne; see also my “Risk, Transaction Costs,” part V.

²⁹ Even land that is relatively fertile if properly worked may be considered inferior if it is relatively unforgiving of improper practice. Moreover, again on grounds of opportunity cost, one should be wary of the assumption that the lord kept the best land for himself. Imagine a plantation owner who has decided to hut his slaves: if suboptimal techniques reduced output more on infertile land than on fertile land, the master would maximize the surplus by keeping the poorer land “in demesne.”

³⁰ Parain, CEHE, I, 126. On the relative scarcity of supervised agriculture in Italy see for instance Slicner van Bath, Agrarian History, p. 47, and Duby, Rural Economy, pp. 53, 376 ff. The regular clergy s predilection for demesne farming is noted by Vinogradoff, P., English Society in the Eleventh Century (Oxford, 1908), p. 356Google Scholar, and by Maitland, Domesday, pp. 321, 392; the latter attributed it to the superior technical knowledge of “the church's reeve [who] was a professional agriculturist.” I wish to thank Fred Cheyette for bringing these points to my attention.

³¹ In the Carolingian period, tenants appear to have spent perhaps one eighth or one ninth of their working time at the lord's bidding. The Polyptyques indicate a standard labor obligation of three man-days per mansus; each mansus supported at least one household and usually more, the figures in the secondary sources suggesting an average of perhaps one household and a half; and the households appear to have averaged 2.3 adults and 2.5 children. These figures suggest an available labor pool equivalent to perhaps 4 to 4.5 adult male workers per mansus, or 24 to 27 man-days per mansus per six-day week, placing the lord's share at about 12 percent of the available labor. This estimate may yet be too high, if (as in thirteenth-century England) a labor-dues “day” was only a half-day. See Duby, Rural Economy, pp. 12–13; Bloch, French Rural History, p. 73; Slicher van Bath, Agrarian History, p. 183; and Perrin, Ch.-E., “Observations on the Mansus in the Region of Paris in the Early Ninth Century,” in Cameron, R., ed., Essays in French Economic History (Homewood, Ill., 1970), esp. p. 29Google Scholar.

³² See for instance Stevens, CEHE, I, 105.

³³ It would also seem reasonable to attribute the prolonged weakness of higher (peace-keeping) authority to its inability to guard the frontiers against the pressure of raiders and invaders. To the extent that that pressure varied in response to developments in those people's own home areas, this further extension of the model would provide a very attractive exogenous impetus to institutional change within Europe s cultural core, the causal chain running from these “foreign” developments to the weakness of public authority to internal disorder to low productivity to differential expertise to labor services. See for instance Bloch, M., Feudal Society (Chicago, 1961), esp. pp. 35–38Google Scholar; also Fossier, La terre et les hommes, p. 477 ff.; and compare North and Thomas, “The Rise and Fall.”

³⁴ Most of the difference between “typical” medieval and ancient husbandry reflects no more than the different regions those “types” refer to. Apart from the hard horse-collar—which in any case did not see widespread agricultural use until rather later—most medieval “inventions” were not new at all; the heavy plow, water mills, and the like were already known in Roman times. If their use did spread in the Middle Ages, it appears as an adaptation to altered circumstances, as would be created precisely by pacification.

³⁵ See for instance Raftis, J. A., The Estates of Ramsey Abbey (Toronto, 1957), pp. 184, 190nGoogle Scholar.

³⁶ See for instance Duby, Rural Economy, p. 211.

³⁷ Duby, Rural Economy, p. 197 ff.; Slicher van Bath, Agrarian History, p. 145.

³⁸ See Duby, Rural Economy, pp. 140, 272.

³⁹ Slicher van Bath argues that “their tillage … was practised in the same way as on any other mediaeval farm” (Agrarian History, p. 154); Knowles, D., The Monastic Order in England (Cambridge, 1940)Google Scholar, refers instead to “the superior agricultural methods of the monks” (p. 351). See also Roehl, R., “Plan and Reality in a Medieval Monastic Economy: The Cistercians,” in Adelson, H., ed., Studies in Medieval and Renaissance History (Lincoln, Neb., 1972), IX, esp. pp. 93–94Google Scholar.

⁴⁰ The abbots, once again, were in the van of this movement; see Knowles, The Monastic Order, p. 443, and Maitland, Domesday, p. 134 ff.

⁴ 1 There was in this period a widespread tendency for manors (or demesnes) to be rented (“farmed”) to a firmarius. In the present context, we can consider the large (multi-manor) estate as allowing three possible levels (and, correspondingly, units) of control: tile estate, the manor, and the petty tenancy. The “fanning” of manors is thus a move toward decentralization analogous to, albeit at a higher level than, the fragmentation of the demesne into tenancies. For evidence on the general lack of direction from above in twelfth-century English agriculture see for instance Miller, Ely, pp. 252–253; Knowles, D., The Religious Orders in England (Cambridge, 1948), I, 33 ff.Google Scholar, ties the growing decentralization of production to the diffusion of technical expertise.

⁴² The mobility of these medieval wage earners is noted by Slicher van Bath, Agrarian History, p. 177; see also p. 317.

⁴³ The villein had to send “unum hominem,” and the ingenious fellow who sent “unam mulierculatn” was duly amerced; see Miller, Ely, p. 270n.

⁴⁴ The increased stress on labor obligations does not contradict this view. Inflation could make labor-time more valuable than its customary money equivalent, and the lord would have reason to claim it if only to sell it for more than the traditional sum. The lord might similarly wish to increase labor obligations, rather than money rents, as a way of hedging against both inflation and possible labor shortages; but if he held no land in demesne he would be unconcerned with labor shortages, and rents in kind would protect him from inflation. For further discussion see my “The Rise and Fall,” pp. 401 ff.

⁴⁶ See n. 3 above.

⁴⁸ Perhaps the most vivid description of this wave of “improving landlords” is that of D. Knowles, The Religious Orders, I, 32–54; see also Postan, CEHE, I, 583, and Slicher van Bath, Agrarian History, p. 137. Miller, Ely, p. 252 ff., considers this the period of transition from “customary and amateur” administration to a “highly centralized” and professional one. On the lesser vigor of demesne farming on lay estates see for instance Kosminsky, E. A., “Services and Money Rents in the Thirteenth Century,” Economic History Review, V (April 1935), 32 ff. and 42Google Scholar; also Miller, Ely, pp. 73n, 177, and Pollard, A. J., “Estate Management in the Later Middle Ages: The Talbots and Whitchurch, 1383–1525,” Economic History Review, 2d series, XXV (Nov. 1972), 557Google Scholar.

⁴⁷ The quoted phrases are from Titow, J. Z., Winchester Yields (Cambridge, 1972), p. 21Google Scholar; Smith, Canterbury, p. 137; and Raftis, Ramsey Abbey, pp. 190, 200. Smith's account, pp. 133 ff. and 166 ff., is particularly rich.

⁴⁸ See Morgan, M., The English Lands of the Abbey of Bee (Oxford, 1946), p. 54Google Scholar; see also p. 79 ff. For a description of the exceptionally modern techniques in use in certain owner-exploited estates in early fourteenth-century Artois see Slicher van Bath, Agrarian History, p. 175 ff.

⁴⁹ See Duby, Rural Economy, pp. 260 ff., 319 ff.; Knowles, The Religious Orders, I, 48; Miller, Ely, p. 98; Morgan, Bee, p. 118; Raftis, Ramsey Abbey, p. 268 ff.; Smith, Canterbury, p. 143. While after the Black Death population appears to have kept declining until the end of the century, most noticeably in the 1360's and 1370's, the initial onslaught of the plague was of course by far the most severe; see for instance Helleiner, K. F. in CEHE, TV (Cambridge, 1967), 13Google Scholar.

⁵⁰ For further discussion see my “The Rise and Fall,” p. 406.

⁵¹ See for instance Helleiner, CEHE, IV, 13; Shrewsbury, J. F. D., A History of Bubonic Plague in the British Isles (Cambridge, 1970), ch. vGoogle Scholar; and Postan, M. M., “The Fifteenth Century,” Economic History Review, IX (May 1939), 160–167Google Scholar.

⁵² For a description of such an instance see Brandon, P. F., “Cereal Yields on the Sussex Estates of Battle Abbey during the Later Middle Ages,” Economic History Review, 2d series, XXV (Aug. 1972), 404 ffGoogle Scholar.

⁵³ On all this see for instance Gras, A History of Agriculture, pp. 36 ff., 157 ff.; Helleiner, CEHE, IV, 29; Slicher van Bath, Agrarian History, pp. 164 ff., 195 ff.; also my “Risk, Transaction Costs,” part VI. Enclosure for tillage or convertible (fieldgrass) husbandry is of course quite different from enclosure for permanent pasture, as the latter does not increase the labor force subject to authority, but virtually eliminates labor altogether. Conversion to pasture occurred in the later Middle Ages as well as in the (early) sixteenth century, and appears to be primarily an adaptation to the relative factor scarcities of the period! The notion that men may be driven from the land because they are too scarce is paradoxical but valid; one recalls the New England town which in recent memory lost its major source of employment as the local textile mill was dismantled and shipped to Brazil, where labor was cheaper.

⁵⁴ See for instance Gras, A History of Agriculture, pp. 181 ff., 208 ff.; Helleiner, CEHE, IV, 52–53, 67, 73; Slicher van Bath, Agrarian History, p. 221 ff.

⁵⁵ An independent peasant owner would voluntarily turn over his land and labor to the innovator if the latter offered him a combination of rent and wages in excess of his current earnings, and convinced him that his land would not suffer under the new methods of cultivation. If the innovator's technique is truly superior, the financial offer is easily arranged; but the peasant's conservatism, or diffidence, may still stand in the way of agreement. Expropriation removes that obstacle with wonderful efficiency, and speeds the improvement of husbandry accordingly.

⁵⁶ The source of differential expertise may thus have changed from “innovation”to “renewal,” suggesting that the nineteenth century may be more closely analogousto the tenth and preceding centuries than to the eighteenth, sixteenth, and thirteenth. The role of continued “innovation” seems far more important in industry, which with the “industrial revolution” also passed from independent small-scale production by skilled artisans to supervised labor by largely “unskilled” workers.

⁵⁷ The independent farmers around Amherst, for example, produce much ordinary tobacco; but shade-grown tobacco, which requires much greater care, is overwhelmingly produced by the corporate subsidiary of a well-known conglomerate.

⁵⁸ See Kula, W., Théorie économique du système féodal (Paris, 1970)CrossRefGoogle Scholar.

⁵⁹ Ibid., p. 33.

⁶⁰ Ibid., pp. 46, 85, 89. As Roman slave-owners had known, the tendency to extract all the available surplus reduces the surplus that is produced at all, and thus redounds to the master's loss.

⁶¹ Ibid., p. 51.

⁶² See for instance Postan, CEHE, I, 609; similarly North and Tromas, “The Rise and Fall,” p. 798.

⁶³ Both these items would swell the lord's income if he turned his workers into rent-paying tenants without altering their welfare. More generally, the party who wishes to abandon rents for dependent labor would bear the full cost of the resulting agricultural inefficiency plus whatever premium was necessary to induce the other party to accept the change. The present hypothesis is that the lord wished to exercise authority, and thus had to pay for it (at least in comparison to what he could have collected otherwise); conversely, the hypothesis that dependent labor was sought by the peasants as a means of shifting risk to the lord would require the peasants' incomes to drop by enough to absorb the high transaction costs and compensate the lord for accepting some of the risk (see my “Risk, Transaction Costs,” part II).

⁶⁴ One can explain anything by postulating changes in preferences; on methodological grounds, it must be considered an explanation of last resort.

⁶⁵ Compare North and Thomas, “The Rise and Fall,” p. 790n, and my “The Rise and Fall,” p. 394.

⁶⁶ Unlike authority for technical improvement, moreover, authority for servility is perfectly compatible with complete technical homogeneity between demesne and peasant tenures, including specifically a lack of differences in crops (which would in turn reduce the cost of scattering the demesne strips among the peasants; see above, n. 28).

⁶⁷ See above, n. 40.

⁶⁸ See Kula, Théorie économique, pp. 111–112.

⁶⁹ One might wonder whether the belief that attitudes are particularly afiected by one's daily work is in fact justified; if it is not, the medieval lords could have obtained the desired servility, at no agricultural cost, merely by multiplying purely symbolic acts. I would be skeptical of such an argument; and in any case one need not be “right” to be “rational.” Pareto-efficiency, after all, is always defined in the context of the available technological knowledge, and the perceived technological possibilities, in human engineering as in any other, are the ones that matter.