Stated by Postan in a conversation with J. Ambrose Raftis.

Harvey, “Introduction,” p. 14.

At harvest time, certain tenancies in some locales owed multiple works or workers per day.

For some views as to why direct cultivation of the demesne became widespread in these centuries, see Harvey, “English Inflation”; Miller, “Farming”; and Raftis, Ramsey Estates, especially pp. 76–91 and 96–128.

The manors analyzed in this article lay in three counties, which were employed by Koskminsky in his study: Bedford, Cambridge, and Huntingdon. He notes that this estimate is not necessarily representative for the rest of England. See his Studies, pp. 89–94.

On the development of manorial farm accounts, see Harvey, “Introduction.” On this matter and for a discussion of auditing procedures, also see Oschinsky, Walter of Henley, especially pp. 213–60.

The reeve received free rent and sometimes some perks for this job. The demesne might also be managed by a “baliff,” who worked for wages. As a rule, reeves managed the demesnes on Ramsey Estates before the Black Death.

Account rolls were drawn up at the end of the fiscal year. For example, a fiscal account roll year of 29 September,1323 to 28 September, 1324 was produced in September of 1324. A crop was harvested in the summer of 1324 but there was no possibility of processing the entire crop fast enough to record its output in the fiscal year of 1323/24. Instead, this crop was threshed and winnowed over the course of 1324/25 and appeared in the accounts for this fiscal year. On this medieval accounting practice, see Titow, p. 35. The labor inputs, therefore, expended on the harvest recorded in the fiscal year of 1323/24, with the exception of threshing and winnowing, appeared in the fiscal accounting year of 1322/23. Interpolation of labor inputs can be safely employed. Farmers did not change their labor inputs per acre in the arable sector significantly from one year to the next; nor did sown acreage vary much from one year to the next year. Changes were made gradually, over time, and farmers did not change their practices dramatically within two years. Changes in labor inputs found in consecutive years, extant for the post-1350 period, were usually an insignificant 10 percent a year. Even for fiscal years spaced widely apart in the pre-1350 period, labor inputs, as shown in Appendix Table 1, are still fairly consistent between years. Of the account rolls that differed in their labor inputs per acre significantly from other years, four were from the exceptionally poor harvest year of 1297/98. Unusually higher inputs per acre in this year resulted from many more tenants attending the reaping and plowing boonworks. This was likely a charitable gesture on the landlord's part, as large amounts of food were distributed to workers at these gatherings.

The date at which these individual demesnes were leased varied from the late fourteenth to the early fifteenth centuries.

Oschinsky, p. 421. On the possible dating of the Husbandry to c.1300, see ibid., p. 201. For instance, in Ramsey Estates' account rolls, the grange section, which recorded the output and the yearly expenditures of crops, very carefully distinguished the amount of each cereal or legume type threshed by pieceworkers and tenants. Full-time demesne employees rarely threshed until after the Black Death.

Ibid., p. 445.

Loveboons or boonworks were works required in addition to the regular work week. They were usually performed at times of peak demand for labor, such as harvesting. Although strictly defined in number and type, they were theoretically performed only when requested by the landlord. That is, they were somewhat arbitrary in nature. On the Estates, such loveboons were standardized in number and type and incorporated into the regular labor rent system. On the standardization of loveboons, see also, Jones, “Harvest Customs … the Hay Harvest,” p. 106.

“Seneschaucy,” p. 269 in Walter of Henley. Oschinsky dates this treatise to c.1276. See, Oschinsky, “Introduction to the Seneschaucy,” p. 75.

Ibid., pp. 279. See also the “Husbandry,” p. 435 in Walter of Henley. “Customary tenancies” were those landholdings that owed a specified amount of labor on the demesne or the cash value of this work for their rent. Those leasing freehold property on the manor also often provided a slight amount of work on the demesne as part of their rent, usually in the form of boonwork(s).

On the impact of economic epistemologies on accounting practices, see Poovey and Miller, “Accounting”; and Hoskins and Maeve, “Writing.”

For an outline of the Abbey's foundation, see Raftis, pp. 1–6, and for its agrarian holdings in 1086 and subsequent purchases of land in the thirteenth century, ibid, table 1, p. 20 and table 20, p. 110.

Because opera sections were often attached on a separate membrane to the end of the roll before the Black Death, they could be easily detached and lost. The number of such sections that survive for Ramsey Estates, based on my comparative reading of three other large estates' account rolls, again, seems to be exceptional. Without reading all extant account rolls in England, though, it is hard to say anything more definitive. See Appendix 2.

By using manors located within one estate, it was also possible to distinguish between estate-wide managerial decisions and those carried out on a local level, on each manor, which seemed to reflect local farming practices. This issue is paramount in distinguishing which demesne productivity rates are applicable to the vast tenant farmers' sector. Kathleen Biddick's work (Biddick with Bijleveld, “Agrarian Productivity,” pp. 95–123.) has highlighted the import of approaching the question of land productivity within the framework of central estate management. The hypothesis that medieval English land productivity rates were low was predicated largely upon the evidence from Titow's Winchester Yields. Biddick attributes low yields per acre on many Winchester manors to an orientation to pastoralism on these manors. Equally as important were the extraordinary low capitalization rates on these demesnes, which were caused by the periodic stripping of Winchester Estates' working capital by the crown. J. Ambrose Raftis's The Estates of Ramsey Abbey, which examined Ramsey Estates' central management practices over the centuries in quantitative detail, has proven to be an invaluable aid in untangling the local and central decision-making process.

See, Raftis, Estates, throughout, on the nature of the farming system on Ramsey Estates' manors.

This practice is found for both wage and tenant labor. Tenants who worked on a manor other than their own had their works entered on the opera section of their home manor to which they owed the labor rents; no notation that such “foreign” tenant work was performed was made on the account roll of the nonhome manor. The reeve, the demesne manager, only needed to provide a report of those labor inputs for which he was financially responsible. So, the reeve on the home manor noted the payment of labor rents from customary tenants with holdings on his manor but the reeve on the manor where these “foreign” tenants paid their labor rent, did not need to report their inputs. As long as all account rolls and opera sections are extant for the same years for each manor in such an administrative unit, it is possible to accurately calculate labor inputs. Even when this is not the case, interpolation can still supply relatively precise estimates.

Because in all these cases, each manor consisted of a single village, the terms manor and village are used interchangeably, whenever referring to characteristics or activities that were likely common to both the seigniorial and tenant sector.

The villages of Abbots Ripton, Elton, Upwood, Warboys, and Wistow were located in Huntingdonshire. Gravely and Knapwell lay in Cambridgeshire, and the village of Shillington was in Bedfordshire. For a map of these and the other manors of Ramsey Estates, see Raftis, Estates, p. 20. For a description of these manors, see ibid., throughout.

The negative qualities of clay soils have been a bit overstated in the literature. It should be borne in mind that although such soils required more work in some types of jobs, particularly to ensure adequate irrigation, they were far more fertile than the chalky and sandy soils found often in the eastern regions of England. The “bread basket” regions of both Canada and the United States are dominated by clay soils.

Necessary also to this analysis was the reconstruction of the labor rent system on the estates, as well as the socioeconomic, cultural, and environmental conditions of each village in this study. The latter permitted an understanding of the rationale for farmers' economic decisions. These findings will be explored elsewhere.

I have located only few early modern farm accounts that might permit a breakdown of days worked in the two sectors: arable and pastoral. Likely the most viable approach would be to estimate the output per worker for both sectors with these records. In using these accounts, some allowance must be made for labor inputs of the farmer and his family, because they are not recorded in the same fashion as that of wage labor. For a discussion of early modern farm accounts and their methodological issues, see Collins, E. J. T. “Historical Farm Records”; and also Turner, Beckett, and Afton, Farm Production.

Mate, “Medieval Agrarian Practices,” pp. 22–31. Campbell, (“Arable Productivity”) like Mate, concludes that “the man who owned a demesne was probably less important that the man who ran it … [according to] … prevailing local husbandry practices,” pp. 387, 397–98.

Campbell includes only the inputs of full-time workers and payments made to part-time work on weeding, and implies that other part-time work was slight. I assumed that each weeder was paid one penny a day. See, “Arable Productivity,” table 5, pp. 38–39. I reconfigured Martham's demesne acreage to statute acres. For the size of the customary acre at Martham, see Campbell, “Arable Productivity,” p. 385.

Fox, “Exploitation,” 544–45.

Thorton, “Determinants,” table 7.7, pp. 205. On fallow, see ibid., p. 186.

Parker and Klein estimate about 30 hours of work per sown acre on wheat crops in the southern United States. See, “Productivity Growth,” p. 527, table 1 on p. 532, and p. 570.

Days per acre were calculated from Allen's estimates of work per acre required on different size farms. Allen assumes 250 days of work a year per full-time worker. See, “Growth,” table 11, p. 131, and table 13, p. 140.

See Timmer, “Turnip.”

Timmer argued that the greater labor inputs required in turnip production off-set any possible increase in land and thus labor productivity in c.1800 England. For this hypothesis, see “Turnip.”

Tenants ad opera (at labor rent) could pay some off their rent in cash rather than labor. Such payments are recorded in the account rolls as “sales” of each labor rent back to the tenant.

BM 34885; PRO 882/13, 882/19; Elton Manorial Records, pp. 204–29.

Karakacili, “Peasants,” tables 3.1 and 3.2, pp. 87–190.

Campbell bases his recent estimates on levels of English medieval national grain production on the assumption that land productivity rates of the seigniorial and peasant sector were comparable. See his, English Seigniorial, p. 396.

On the role of furlongs in distinguishing arable qualities, see, for example, Hall, Open Fields; and Hogan, “Clays.” On post–Black Death demesne leasing rates that indicate divergence of soil quality, see Karakacili, “Peasants,” pp. 190–91.

On the differences in crop choices and labor inputs found between small and large farms in developing countries, see, Berry and Cline, Agrarian Structure.

The estimates of tenants' plow numbers are taken from the number of tenants' plows recorded as present at the plowing boonworks in the account rolls. This number fluctuate widely from year to year or even in the same year. For example, in the village of Upwood in 1297/98, the first boonwork employed 26 plows and the second used only 18 tenant plows. See, PRO 885/11 and 885/14. As food was given to all workers at these loveboons, as a cost-saving measure, the reeve probably issued a request for the minimum number required to complete plowing needs. For this reason, the highest number of plows likely illustrates the minimum number of plow-teams held by tenants in each village; it is possible that tenants held an even greater number of plow-teams. So, tenants in Elton had one plow-team for every 43 or 52 acres but the landlord employed only one plow-team for every 105 acres. In Upwood, the demesne had one plow-team for every 110 acres but tenants could draw upon one for every 35 or 46 acres. The landlord was not undercapitalized in this area, though. A medieval agricultural treatise, the Seneschaucy, states that one plow-team should be able to plow about 180 acres in one year, p. 264. Two estimates were given for the ratio of tenants' plow-teams to tenants' acreage because of the occasional shift in the categorization of freehold and customary land. The first estimate deducts only freehold holdings of one hide, following the argument of Kosminsky that holdings of this size tended to operate as a sub-manor. On these issues, see Kosminsky, Studies, pp. 80–84, 78, 88.

Wrigley, “Energy,” pp. 325–26; see also his Continuity, pp. 38–39.

Langdon, Horses, pp. 254, 268, 253.

It should be noted that output per worker in the village of Shillington was probably much higher than this amount would seem to indicate. The only usable account roll that survives from before the Black Death was for a year characterized by a relatively poor harvest (1323) and therefore this estimate represents the lower levels of labor productivity here.

The causes that underlie these differences offer insight into the elements shaping medieval labor rates and the potential for change in later centuries. A discussion of these matters will be presented elsewhere.

Based on my direct measurement of arable workers' output, using a few early modern farm accounts, it appears that output per worker likely rose, at least slightly, in later centuries.

Holderness, “Prices,” p. 174.

For a summary of some of these issues, see Afton and Turner, “Agrarian Occupations,” pp. 1956–63.

It should be noted that because the bulk of data are post-1300, they cannot indicate whether labor rates rose or fell from thirteenth-century levels.

For a summary of some criticisms and defenses of Boserup's work, see, for instance, Hunt, “Boserup Revisited”; and Eder, “Agricultural Intensification.”

For instance, by 1400 in Norfolk, wheat yields declined by a mean of 20 percent from average bushels an acre found between1275 to 1349. Barley yields fell in the 24 years after 1349 by 6.5 percent as compared to the mean from the previous 74 years and then rose again in the next quarter century to equal their highest pre–Black Death level. See Campbell's, English Seigniorial Agriculture, table 6.8, p. 180.

In post–Black Death villages where labor inputs decreased, they did so by a slight amount and not in proportion to the huge population decline. Karakacili, “Peasants,” pp. 176–241.

Decent aggregate data emerge by the mid-twentieth century on the factors of production employed in developing countries' agricultural sectors. I also use these early data to avoid the later green revolution, which increased land productivity dramatically in some regions.

Constructed using information in Colin Clark and Haswell, table 17, p. 82. To compare these numbers with the output of medieval workers, hours worked per acre were converted to days per acre, assuming eight hours per work day and 300 days of work per full-time worker. If ten hours a day are used rather than eight, labor productivity rises by 20 percent. The conversion to wheat bushels made using weights given by Clark, which are based on Food and Agricultural Organization of the United Nations data from 1960. See, ibid., pp. 197–96.

Ibid., p. 90, with the assumption that the metric ton was employed. The long ton would mean that each worker produced 134 wheat equivalent bushels a year.

Tax's study of Guatemalan peasants cited in ibid, p. 118. Agricultural productivity made using procedure described in notes 62 and 63.

Made using information found in Buck, Land Utilization, table 14, p. 224, and table 15, p. 304, and his Statistical Supplement, table 9, p. 209. Buck gives man-days per acre but does not give the length of a working day. Conversion to wheat bushels was made as described previously.

Clark and Haswell, Economics, pp. 86 and 88. On the comparative advantages of farming in temperate versus tropical zones, see Gorou, Tropical World.

Sen, “Problems,” pp. 533–45. Kuznets argued that mid-twentieth-century agrarian labor productivity rates in less developed countries were only one-fourth to one-third the levels found in wealthier nations during their similar pre-industrialization stage. See his, “Underdeveloped Countries,” p. 179.

The estimates of population density requirements in early-fourteenth-century England were made with the assumption that arable land totaled 10 million acres in 1300. Campbell states that by 1315, English arable acreage was perhaps as high as 10.5 million acres. See, English Seigniorial Agriculture, pp. 386–90. Estimates of population numbers for 1950 and arable acreage for China (1947), India (1948), and Japan (1950) are in Food and Agricultural Organization of the United Nations, Yearbook table 1, p. 5 and table 2, p. 2.

Campbell, “Arable Productivity,” pp. 397–98.

E. A. Kosminsky argued that this was the norm as opposed to a village consisting of a single manor. Kosminsky, Studies, pp. 73–77.

Data employed in these calculations were taken from table 2A.5, found in Collins, “Rural and Agricultural Change,” pp.132–33.

My thanks to Campbell for permitting me access to his forthcoming book, Lay Lordship.

Account rolls dating from after the Black Death on the Ramsey Estates differ from their earlier structure somewhat in the type of information which they contain and the manner of its presentation.

More rarely on the pre–Black Death Estates, the value of works might also be categorized by job type rather than by season.

Part-time workers made up a very small percentage of the workforce on Ramsey Estates. Usually information on such wage labor was quite detailed. Once in a rare while, the account roll simply gives the type of part-time worker and the sum total of his or her wages. This tends to occur in two instances: workers hired for a relatively short time (a few weeks or so) and workers hired to cart grain at harvest-time for one day or so. When the worker was employed for a week or more and food was given as part of the wage packet, quite often the number of weeks of their employment can be gleaned from one of the grain expenditure sections. Alternatively, because the type of workers and their employment patterns tended to stay pretty constant from year to year, often a look through the account rolls of years that proceed or follow will uncover an entry that offers the number of weeks or days of employment for such regular part-time workers. Finally, the wages of these workers can also be used to extrapolate the length of their employment.

Jones, “Harvest Customs … Hay Harvest,” p. 106.

Such works sold on account were often likely “false sales” to the reeve, who had failed to use these works or to collect their monetary value from tenants.

The “Seneschaucy,” p. 283 did indicate that carters might cart manure or marl or grain during the year. Usually only one carter was employed full-time on each Ramsey Estates manor and he was frequently away from the manor making deliveries; indeed, extra help for such deliveries was often hired throughout the year. Largely for this reason, labor inputs from the full-time carter were not included. Given the large numbers of days worked on the demesne, even adding 50 or 100 days from this worker on the arable, though, would not significantly change labor productivity rates. The pastoral workers could not abandon their animals for any length of time; this sector also needed to hire part-time workers for periods of peak activity, such as lambing. Medieval agricultural treatises suggest dairymaids could help with the winnowing of the crop. On Ramsey Estates, before the Black Death, the crop was usually winnowed by paid part-time workers and less commonly by tenants as part of their labor rents. As for the woodward, quite often during the harvest season he would leave his post to help with the harvest. Such practices are listed in the harvest expense section. During this time, a part-time worker was hired to work as the woodward. The woodward's inputs, in such an instance, are included in the category of harvest workers. For descriptions of the range of jobs possibly performed by full-time employees, see ibid, pp. 281–89.

For example, the 1344/45 works account for Elton states that tenants performed 137 plowing works (driving the plows) because the plowmen were occupied with other jobs on the manor. See Elton Manorial Records, p. 336.

On full-time wage workers, see Karakacili, “Peasants,” pp. 75–77.

Account roll entries usually just give the rate of pay and number of employment weeks for those workers hired by the week and do not indicate the number of days in a week that they worked. A receipt for Bigging, dating from 1378/79, indicates that almost all the harvesters, paid by the week, actually worked only 2.5 or three days in each week. Similarly, in the manor of Cuxhom, held by these estates of Merton College, an indenture for 1351 survives which indicates that the maximum of four days worked per week by each harvester. See, Manorial Records, pp. 155–56.

In Knapwell, part-time workers were not hired at all in 1323/24, and 1347/48; in 1325/26 and 1329/30, days worked by a part-time employee made up only 1 percent of the total inputs in the arable sector. PRO 768/30, 768/31, 768/32, and 768/35. On most demesnes, part-time arable workers usually supplied about 6 percent or less of the total inputs. For example, in 1323/24 in Elton, part-time arable wage workers contributed 6.6 percent of the total inputs in the arable sector.

For incentive reasons, a likely scenario would find a day worker who was paid four pence (d.) a day thresh four bushels of wheat in one day, should the piece rate pay 8d. for eight bushels. This situation is exemplified in the farm accounts of Robert Loder, a seventeenth-century English farmer. He notes in 1614 that “I payd 12 d. a quarter for pese threshing which was to much by 5 d. in a quarter at least, for they earned 13 d. a day when the dayes were shortest.” In other words, his “tasker” produced 8.5 bushels of peas in one day of threshing and he felt that he had overpaid him. Loder had expected only about 5 bushels to be produced, because the implied day wage was seven pence (d.), and the tasker received 1.5 d. for each threshed bushel. See, Robert Loder's Farm Accounts, p. 80. The output of his day workers in threshing, however, varied greatly. In 1613, they nearly doubled their output from the previous year to over 15 bushels of barley, even though the barley was the same size and dryness. Loder observed “thus men can work yf they list & soe they can loyter,” ibid., p. 25. Gregory Clark has also employed relative wage levels in threshing piecework rates to try to establish efficiency rates. An examination of the differences in relative medieval threshing wages within villages over time and between villages in different regions in the same year suggest that the value of the output rather than the efficiency of labor played the key role in setting relative wage levels. See, Karakacili, “Peasants,” pp. 44–45.

Making this point, see Campbell, “Uses,” pp. 1–2; and Slicher Van Bath, Agrarian History, p. 299. Collins, though, argues that changes in the hand-tools in the pre-industrial period might increase efficiency for some jobs. See his “Sickle to Combine.”

The extents do not discuss winnowing works but the account rolls supply a bit of information on the efficiency of tenants in this work. At Wistow in 1350/51, tenants winnowed 24 bushels per work (each arable work could not exceed one day) but the next year, more than twice this amount (over 48 bushels) per work. BM 39897, 39898. The type of grain or legume possibly affected the output but this information is given only in the later year. The number of days worked by winnowers are relatively few and changes here would not have a significant effect on labor productivity. An English source, c.1800, states 30 bushels could be winnowed in one day: Volume 4: Midlands, General View of the Agriculture of the county of Warwick, p. 322. Leo Rogin cites two widely differing estimates for winnowing wheat with a sheet and sacking it: 12.5 to 15 bushels in 10 or 12 hours as opposed to 33 bushels in one day. See Rogin, Introduction, pp. 176–77. If any researcher has information on winnowing productivity, I would very much appreciate any such references.

Postan notes that such tenants appear to, at times, receive wages of some sort (cash or food) from the landlord. This managerial practice of paying akermen both free rent and cash was also described in the surveys for Ramsey Estates' manors of Ringstead and Shillington. In the case of the latter manor, the landlord was trying to stop this practice. Cartularium Monasterii, vol. 1, pp. 408 and 474.

Land on the manor designated as akermanlands could also be subsumed into the regular labor rent system, leased, or held by another type of demesne official (non-akerman) or employee. The account rolls, on the Estates, always clearly indicate such situations and therefore there is little danger of adding too many akermen to the mix.

For a discussion of the Ramsey Estates' evidence concerning akermen, see Karakacili, “Peasants,” pp. 61–67.

For those rare occasions when the clerk failed to record the number of individuals attending such precaria, interpolation using other such entries was employed.

Ramsey Estates was not unique in this regard. On the widespread practice of one labor rent requiring a maximum of a half-day of work, see, Bennett, Life, pp. 104–05.

For example, in Shillington, a ditching work could be defined by piecework or by a length of time. A full day of ditching was counted as two works. If the tenant was given a meal, then one day of work counted only for one work. Cartularium Monasterii, vol. 1, pp. 460–61.

The amount of food to be received for exceeding the norm of a half-day of work is often carefully described in the surveys, but the account rolls suggest that the practice before the Black Death—at least, if the amount of food distributed at the harvesting boonworks are any indication—was to give tenants far more than the minimum allowance of food. Thus, food expenditures on tenants cannot be employed to estimate the number of full-days they worked before 1348.

For example, threshing inputs constituted 56 percent of all tenants' arable works in Shillington in 1323–24 and 43 percent in Wistow in 1335/36. Threshing works were less significant in Warboys, tending to fall in the 30-percent range, consuming 32 percent of tenants' labor rents in 1342/43, 25 percent two years later, and 36 percent in 1347/48. The number of threshing works employed would shift according to the size of the demesne harvest: a very good harvest meant that more threshing works would be used and vice versa. PRO 741/20, BM 39895, 39803, 39805, 34900, 39809.

Cartularium Monasterii, vol. 2, pp. 38 and 44. See also, Cartularium Monasterii, vol. 1, pp. 460 and 408. The value of labor rents varied, according to the fiscal “season” in which they were performed. The value of a “winter” work, the season during which threshing was generally performed, was usually 0.5 pence. The evidence that these were half-days was often explicitly stated in the extents. Postan noted that the valuation of 0.5 pence for certain labor rents in the surveys indicated that they must have been one half day of work, given that the wage norm for one day of work in the winter was one penny in the thirteenth century. See, Postan, Famulus, p. 42, fn. 4. For a more lengthy discussion of threshing and the tenant labor rent system, see Karakacili, “Peasants,” pp. 42–54 and 270–76.

For years in which an opera section does not survive, an Estimate 1 can still be made for threshing works. Interpolation of the number of threshing works can be made by using the information found in the account rolls on the average number of bushels produced per threshing work. This should be considered in conjunction with the information contained in the grange on the types and totals of bushels threshed by tenants for their works. The inputs of paid piecework threshers are also treated differently in Estimate 1. Rather than estimating days worked based on the likely minimum efficiency of labor as in Estimate 2, a calculation of one day of work is entered for any output up to eight bushels of any crop type. Again, the purpose of this procedure was to achieve a lower-bound estimate of labor productivity.

The bushels obtained from the same number of sheaves could vary from year to year, depending on the quality of the harvest, but evidence suggests that sheaves were standardized to a certain size. The constancy in the number of sheaves to be threshed for one work on most Ramsey Estates manors implies standardization. On this point also, see Jones, “Harvest Customs … Corn Harvest,” especially pp. 15–16. He notes that some methods would be more precise than others.

Ramsey Estates' labor rent system was not unique in this respect. For example, such an output-based system for threshing was also found in Merton College. On the manor of Cuxham, held by Merton College, one threshing rent was defined as the production of four bushels of wheat/beans/peas or five of barley or six of dredge or eight of oats. See, Manorial Records, pp. 141–43.

Two threshers were hired for 4.5 days and they produced 68 bushels of wheat. See Elton Manorial Records, p. 375.

For a summary of the nineteenth-century evidence on flailing, see Rikoon, Threshing, p. 609; and Rogin, Introduction, pp. 178–80. See also, Bidwell and Falconer, p. 205. Collins (“Sickle,” p. 9) provides an estimate of threshing productivity of wheat in weight; assuming a statute bushel weight of 57 pounds, the output was about six bushels per day. Gregory Clark (“Labour Productivity,” p. 233) estimates 4.2 bushels of wheat threshed per day from 1561 to 1850, based on the ratio of daily wage to piece rates and the relative wages in threshing different grains. On this method see the text and note 87.

The inputs of paid threshers are also treated in this fashion in Estimate 2. This methodology offers researchers who lack opera sections a manner in which to calculate the inputs of tenants in threshing. Days worked in threshing can be gauged from the amounts given in the grange as threshed by tenants.

Three manors in this study had surviving opera sections for all their extant account rolls. At least two and often three opera sections survive for manors that had more than one complete account roll year available. See Appendix 1.

If it is feasible, for reasons of accuracy, it is best to reconstruct the entire labor-rent system on each estate before making estimates from each manor.

PRO 885/30; 885/18; BM 39798; 39803; 34803; 34804.

See note 17.

On those few manors, such as Lawshall, where a substantial amount of plowing was to be performed by tenants, the account rolls do carefully record the amount of plowing owed, sold, and performed by tenants. For example, in one year, Lawshall tenants owed a total of 168 plowings. They performed 32 at the fall sowing of the wheat, 42 at the seeding of the spring crops, 25 acres in the summer (probably on the fallow), and paid cash not to perform 69 plowings. They also record the plowings that tenants performed as boonworks. See, PRO 1001/09. Plowings are also found in the opera section of the account rolls, but these were a rare occurrence and few in number. Thus, in Elton in 1297/98, tenants ploughed for a total of 123 works. But these works were deducted from the regular number owed for the week and no mention is made of any Friday plowing works owed as part of the rent. See Elton Manorial Records, p. 79. On several manors, including Gravely and Knapwell, there was a special section in the account rolls that documented the sale of plowings to tenants. That is, tenants paid a money rent and did not perform this job. For a full discussion of the Ramsey Estates' evidence on tenants' plowing found in the surveys and account rolls, see, Karakacili, “Peasants,” pp. 277–95.

Farmer found for about 150 manors from different estates, tenants plowed only 8 percent of the demesne arable; see his “Famuli,” pp. 212.

Hypothetical plowing labor inputs were maximized by calculating them for all holdings on each manor that owed such plowing, both for tenancies paying their rent with labor as well as for those paying off their labor rents with cash (said to be ad censum in the account rolls). The surveys consistently state that tenants who do not own a plow were either not required to perform this task, or they were to help their neighbors plow, or they could join their animals together with those of a neighbor to plow the required one-half acre or one-quarter acre. I allowed for only 10 percent of tenancies (excluding smallholdings) not to own a full plow. I also assume that tenants owned small plow teams of four animals or less, and so employed only one man. There is some debate over the number of men used on tenant plow teams. A deep plowing on heavy soil or a first plowing, might also require more than one man per team. If two men were the norm, then the inputs should be doubled. Such plowing jobs were generally scheduled from 29 September to 1 August, with three weeks off for holidays. The surveys also make it clear that plowing could not be performed, sometimes, on Fridays, because of frost or poor weather conditions. I estimate 34 Fridays on which tenants might plow, deducting the three Fridays allowed for holidays and a possible seven Fridays on which work could not done because of unfavorable weather. Plowing a half-acre only required a maximum of a half-day, so that each holding then owed 17 full days of plowing work. These estimates can be found in Karakacili, “Peasants,” table C, pp. 296–97.

This follows Gregory Clark who calculated 300 days of work per “full-time” worker for the thirteenth to nineteenth centuries, based on the argument that this was a fairly typical work year in the 1800s.

Should another researcher wish to make a comparison with the rates in this article to those of workers who were employed 250 days a year, the output per medieval worker should be adjusted downwards about 17 percent. At 275 days a year, the difference would be 8 percent in the output.

Small numbers of sheaves from a previous or new harvest are occasionally recorded in the grange; these bushels were not included in the total output. I do include the labor inputs employed in processing such sheaves, because I was interested in tracking percentages of labor inputs in the three categories of jobs. As these labor inputs were relatively few, and so lowered labor productivity by a minute amount, I retained them in the final calculation of labor inputs. A few unthreshed sheaves for the harvest year are likewise found sometimes, usually for legumes. Unless an estimate of their output in bushels was given in the grange, one was not made and added to the total, since the output of sheaves might vary from year to year. I did not want to run the risk of overestimating the output and thus driving up labor productivity.

In the printed edition of Elton's account rolls, this account was misdated as 1324/25. The eighth year of the Abbot Simon Eye, who assumed office in 1316, was 1323/24 (the first year of office must be counted in dating the roll). See, Elton Manorial Records, p. 263.

It is unlikely that the reeve and akermen generally worked the entire day for each work owed. Such a number of work days would have posed some difficulties in the management of their own farms. The maximum number of days worked were added for these workers in order to promote the policy of achieving a lower-bound estimate of labor productivity.