How much tobacco a planter or field hand could or did make in colonial Virginia varied from place to place, year to year, man to man. Tobacco productivity was a direct function of numerous factors including climate, tobacco variety, soil quality, crop stategy, labor management, and “the strength, stamina, and industriousness of individual workers.”[ref]Robert Anthony Wheeler, “Lancaster County, Virginia, 1650‑1750: The Evolution of a Southern Tidewater Community,” diss., Brown University, 1972, 56; Edmund S. Morgan, American Slavery, American Freedom: The Ordeal of Colonial Virginia (New York: Norton, 1975) 142-3n33; Russell R. Menard, “The Tobacco Industry in the Chesapeake Colonies, 1617-1730: An Interpretation,” Research in Economic History 5 (1980): 145; Lorena S. Walsh, “Slave Life, Slave Society, and Tobacco Production in the Tidewater Chesapeake, 1620-1820,” Cultivation and Culture: Labor and the Shaping of Slave Life in the Americas, eds. Ira Berlin and Philip D. Morgan (Charlottesville: University Press of Virginia, 1993) 174.[/ref] What we would like to isolate here is the indirect impact changing tobacco prices had on tobacco productivity.
From numerous contemporary statements it is quite clear that the price planters received for their tobacco was very important to them. Unfortunately, planters did not know how much they would get for their crops in the fall when they planted it in the spring. All they knew for sure was how much they had received for the previous year’s crop and earlier crops.[ref]Planters who consigned their tobacco to English merchants, while certainly aware of past and present local prices, often did not know for a couple of years the actual price their tobacco sold for in European markets. See John Spencer Bassett, “The Relation between the Virginia Planter and the London Merchant,” Annual Report of the American Historical Association for the Year 1901, Vol. 1 (Washington: GPO, 1902) 570. Consignment, which grew markedly during the war years of the 1690s although never surpassing direct sales well into the eighteenth century, only slightly complicates the analysis of the relationship between tobacco prices and tobacco productivity. Although consigning planters undoubtedly based their production decisions on far more than past and present farm prices, by taking on more of the costs and risks of merchants they simply tried to tap into the potential profits of the tobacco trade. But consigners hardly escaped the vagaries of tobacco prices on their side of the Atlantic which in a competitive market captured many of those costs, risks, and profits. Indeed, one might suggest that the growth of more risky consignment over local sales and the rise of the merchant-planter was a response to the relative poverty of the 1690s. On consignment in general, see Jacob M. Price, Perry of London: A Family and a Firm on the Seaborne Frontier, 1615-1753 (Cambridge: Harvard University Press, 1992) 31-2.[/ref]
Studies of modern farmers have shown that they do not act like stockjobbers on the Chicago commodities exchange, responding readily and rapidly to every swing in crop prices. In part, this is inherent in the very nature of farming as business; it is simply very difficult and expensive for a farmer to change crop strategies after the crop is in the ground. Crop strategies have to be worked out well in advance of any solid knowledge of the price the farmer will receive for his crop. In devising their crop strategies, farmers thus are forced to rely on educated guesses of future crop prices. In effect, they devise rule-of-thumb models that take into account not just current crop prices, but also past prices and recent price trends. Thus, in analyzing the elasticity of agricultural supply to crop prices, these studies emphasize the need to experiment with different models of prices with the goal of improving the fit of predicted to actual data.[ref]Wassily Leontief, “Comment,” Studies in Income and Wealth, Vol. 11 (New York: National Bureau of Economic Research, 1949) 443; T. M. Brown, “Habit Persistence and Lags in Consumer Behaviour,” Econometrica 20 (1952): 355?; Milton Friedman, “Supplementary Comment,” Journal of Political Economy 66 (1958): 549; Marc Nerlove, Distributed Lags and Demand Analysis for Agricultural and Other Commodities (Washington: GPO, 1958) 1, 7-8; Hossein Askari and John Thomas Cummings, Agricultural Supply Response: A Survey of the Econometric Evidence (New York: Praeger, 1976) 25-51; G. S. Maddala, Introduction to Econometrics (New York: Macmillan, 1988) 337-87.[/ref] In addition, risk-averse planters may have also based their production decisions on minimum prices they received for their crop in the recent past, accommodating themselves to planting and tending enough plants to assure themselves against the risk of the market.[ref]On various safety-first models, see Jean-Marc Boussard and Michel Petit, “Representation of Farmers’ Behavior Under Uncertainty with a Focus-Loss Constraint,” Journal of Farm Economics 49 (1967): 869-80; Gavin Wright and Howard Kunreuther, “Cotton, Corn and Risk in the Nineteenth Century,” Journal of Economic History 35 (1975): 526-51; James A. Roumasset, Jean-Marc Boussard, and Inderjit Singh, eds., Risk, Uncertainty and Agricultural Development (College, Laguna, Philippines: Southeast Asian Regional Center for Graduate Study and Research in Agriculture, 1979); Mary Douglas, Risk Acceptability According to the Social Sciences (New York: Sage, 1985) 73-82; Frank Ellis, Peasant Economics: Farm Households and Agrarian Development (Cambridge: Cambridge University Press, 1988) 80-101. Unfortunately, the general concern of these studies involves explaining crop mix among different farmers in two-crop models and does not suggest ways of modeling prices in time series analysis of productivity. See Jere R. Behrman, “Agricultural Supply,” The New Palgrave Economic Development, eds. John Eatwell et al. (New York: Norton, 1989) 39-40. More promising models of prices have been developed in studies of times series analysis of consumption and savings behavior in the face of uncertainty and under the influence of habit. Most studies, whether stressing the role of uncertainty (and ideas of a “permanent” or “expected” income) or habit persistence, following the precedent of T. M. Brown and Milton Friedman, usually proxy the impact of prices by a Koyck or other such distributed lag, as do studies of agricultural supply. See T. M. Brown, “Habit Persitence” 355-71; Milton Friedman, A Theory of the Consumption Function (Princeton: Princeton University Press, 1957); Robert Ferber, “Consumer Economics, A Survey,” Journal of Economic Literature 11 (1973): 1303-42. On the difficulties of distinguishing between these “permanent income” and “habit persistence” models despite their very different assumptions about economic behavior, see L. M. Koyck, Distributed Lags and Investment Analysis (Amsterdam: North, 1954) 8; Friedman, Theory 14n, 97; Nerlove, Distributed Lags 1-7, 20-2; L. R. Klein, “The Friedman-Becker Illusion,” Journal of Political Economy 66 (1958): 539-45; Friedman, “Supplementary Comment” 547-9; Balvir Singh and Aman Ullah, “The Consumption Function: The Permanent Income Versus the Habit Persistence Hypothesis,” Review of Economics and Statistics 58 (1976): 96-103. However, other economists following the lead of James S. Duesenberry and Franco Modigliani, stress discontinuous, irreversible, ratchet-like behavior with consumption habits developed during “peak” income years. See James S. Duesenberry, Income, Saving and the Theory of Consumer Behavior (1949; Cambridge: Harvard University Press, 1967); Franco Modigliani, “Fluctuations in the Saving-Income Ratio: A Problem in Economic Forecasting,” Studies in Income and Wealth, Vol. 11 (New York: National Bureau of Economic Research, 1949) 369-441; Tom E. Davis, “The Consumption Function as a Tool for Prediction,” Review of Economics and Statistics 34 (1952): 270-7; Donald Wittman, “The Peak Income Hypothesis: An Econometric Reinvestigation,” Review of Economics and Statistics 65 (1983): 358-60. Although no one as far as I have found has applied such peak income models to studies of supply behavior under conditions of uncertainty or the influence of habit, the idea that effort as a result of uncertainty or habit might be a function, not of peak maximum, but “valley minimum” income seems rather inviting.[/ref]
The only plausible series of tobacco prices we have for developing such models of prices in the seventeenth-century Virginia is Russell R. Menard’s series drawn from prices occasionally mentioned in Maryland probate records. Although Menard’s farm price series is hardly ideal for a study of Virginia‑-and more data on regional price differences, in particular prices on sweetscented tobacco, would certainly be welcome–it is the best measure of the current market value of tobacco that we have and are likely to develop in the near future.[ref]See Russell R. Menard, “Farm Prices of Maryland Tobacco, 1659-1710,” Maryland Historical Magazine 68 (1973): 80, 82. Indeed, an analysis of new land acquisition in the sweetscented region of Virginia shows a stronger correlation with Menard’s price series than with sweetscented prices, although this may reflect the weakness of the available data on sweetscented prices. See Bruce Chandler Baird, Jr., “New Land Acquisition in the Colonial Chesapeake, 1660-1706: A Test of the Malthusian and Staples Hypotheses,” M.A. thesis, College of William and Mary, 1990, 94-6.[/ref] (For a refutation of arguments against using Menard’s tobacco price series for studies of the Chesapeake as a whole, see Appendix V.)
Frankly, any behavioral analysis of the tobacco economy has to move beyond Virginia to include the entire Chesapeake. We would not be able to get far in our statistical analysis if we relied solely on the extant evidence from seventeenth-century Virginia. For example, thanks to the work of Menard and others in Maryland records we have quite workable annual data on servant prices and tithable population for substantial parts of the seventeenth and early eighteenth centuries.[ref]Menard, “Tobacco Industry” 157-61, 163-6. As with our discussion of tobacco prices, there does not appear any a priori reason for rejecting the applicability of such data to Virginia. It was with good reason that social historians usurped the traditional historiographical focus on Virginia to incorporate the entire Chesapeake region, for the similarities in conditions between Maryland and Virginia far outweighed the differences and broadening the data pool gave that much more access to stingy surviving records. The regional interpretation, of course, had often been used by earlier scholars like Wesley Frank Craven, Avery O. Craven, Lewis C. Gray, Arthur P. Middleton, Aubrey C. Land, and Jacob M. Price, but had never fully overcome traditional political divisions. See, e.g., Wesley Frank Craven, The Southern Colonies in the Seventeenth Century 1607‑1689 (Baton Rouge: Louisiana State University Press, 1949) 208-9; Arthur Pierce Middleton, Tobacco Coast: A Maritime History of Chesapeake Bay in the Colonial Era (Newport News, VA: Mariners’ Museum, 1953) viii, 355-8.[/ref] There seems little reason to believe that the responses of Maryland planters to tobacco prices differed significantly from those of Virginia planters.
We also need to think in terms of the Chesapeake when it comes to estimates of our dependent variable‑-tobacco productivity‑-because the only suitable series of data for estimating total annual tobacco production are the records of tobacco imports into London and/or England (and, after 1707, Great Britain) which do not separate out Maryland and Virginia.[ref]Menard, “Tobacco Industry” 113; John J. McCusker and Russell R. Menard, The Economy of British America, 1607-1789 (Chapel Hill: University of North Carolina Press, 1985) 120.[/ref] Import data provide a proxy not for the current year’s tobacco production, but the prior year’s production harvested the previous fall and delivered to England in the spring or summer. The import series measures all tobacco delivered in years alternately beginning September 28 and December 24, but since the bulk of the tobacco crop was delivered to England between January and September the annual series captures only the prior year’s production.[ref]U.S. Bureau of the Census, Historical Statistics of the United States: Colonial Times to 1970, Bicentennial Edition, 2 vols. (Washington: GPO, 1975) 2: 1189-91. Since, according to Menard, there is little evidence that the price of tobacco varied over the course of the planting year from January to November, there is no problem in assigning Menard’s annual price to the entire production year. See Russell R. Menard, “Farm Prices of Maryland Tobacco, 1659-1710,” Maryland Historical Magazine 68 (1973): 82-3; Russell R. Menard, “A Note on Chesapeake Tobacco Prices, 1618-1660,” Virginia Magazine of History and Biography 84 (1976): 401. For the price series for the years 1618-60, Menard assigns prices to the “marketing year” beginning on December 1 “when the crop was ready for sale and when ships from England began to arrive in the colonies.” I presume he followed the same practice in the later years but even if he followed a calendar year there would still be no problem since the planting year nested within the marketing year. William Byrd I and William Fitzhugh in the late seventeenth century often spoke of a “forward” market in the fall and a “latter” market in the spring. However, they also suggested that differences between the two markets were highly unpredictable, based on the amount and timing of shipping in relation to available tobacco. See, e.g., William Byrd, The Correspondence of The Three William Byrds of Westover, Virginia 1684‑1776, ed. Marion Tinling, 2 vols. (Charlottesville: University Press of Virginia, 1977) 1: 66; William Fitzhugh, William Fitzhugh and his Chesapeake World 1676-1701: The Fitzhugh Letters and Other Documents, ed. Richard Beale Davis (Chapel Hill: University of North Carolina Press, 1963) 204. See also John C. Rainbolt, From Prescription To Persuasion: Manipulation of Eighteenth [Seventeenth] Century Virginia Economy (Port Washington, NY: Kennikat, 1974) 45‑6. Although distinct forward and latter markets might naturally arise due to the nature of the local market as prices adjusted themselves to supply and demand between the time the first and the last ships arrived, there is some evidence that the distinctions between the markets rested on other factors. Throughout the seventeenth century, ships arrived and departed at all times of year. However, from 1625 at least until at least the 1690s, the principal ships followed a traditional pattern of leaving England in August and September in order to arrive in the Chesapeake by November when the bulk of the crop was ready and to avoid the winter westerlies, then departing in the winter and spring before the summer when the teredo worm attacked ship hulls and fevers became epidemic. Traditionally another peak of shipping, coming to transport any remaining tobacco not picked up by earlier ships, left England in December, arrived in February or March, and departed later in the spring. See Philip Alexander Bruce, Economic History of Virginia in the Seventeenth Century, 2 vols. (1895; New York: Peter Smith, 1935) 1: 622-3; Susan E. Hillier, “The Trade of the Virginia Colony 1606 to 1660,” diss., University of Liverpool, 1971, 96-9, 279-80, 419; Ian K. Steele, The English Atlantic 1675-1740: An Exploration of Communication and Community (New York: Oxford University Press, 1986) 42, 492. While the early fleet came predominantly from London where ships were dependent on seasonal easterly wind in the Channel, the latter fleet may have come predominantly from the outports, although the evidence is not clear on this last point. In the controversies over single or multiple convoys in the 1690s, merchants from the outports, having smaller ships and a shorter voyage, expressed a preference to sail in January. See Hillier, “Trade of the Virginia Colony” 100-1; Steele, English Atlantic 43, 292. During the war years 1690-1715, tobacco convoys, forced by wartime contingencies, arrived in the Chesapeake concentrated in spring but also with another peak in October, while they left between June and October. Yet, during the Peace of Ryswick, shipping from the outports followed a pattern of autumn sailings from England and spring sailings from the Chesapeake. See V. J. Wyckoff, “Ships and Shipping of Seventeenth Century Maryland,” Maryland Historical Magazine 34 (1939): 352-4; Steele, English Atlantic 43-4, 291-2, 330n13. In the eighteenth century, shipping shifted to spring arrival and late summer clearances. See Steele, English Atlantic 42-5.[/ref] However, the increasing potential for carryover of previous years’ crops by the second half of the seventeenth century could potentially distort estimates of annual production.[ref]On carryover, see L. C. Gray, “The Market Surplus Problems of Colonial Tobacco,” Agricultural History 2 (1928): 4-5; Rainbolt, From Prescription to Persuasion 63. See further discussion below of the impact of a shortage of shipping.[/ref]
The import data are highly imperfect for our purposes. Import totals exclude “tobacco shipped directly to foreign ports, lost in transit, smuggled into England, consumed domestically, and traded to other regions of British America,” and include “tobacco grown in other colonies other than Virginia and Maryland.”[ref]Menard, “Tobacco Industry” 113. On delivery to countries other than England (including Ireland and the Continent) before the demise of the Dutch trade in the 1660s that makes earlier import numbers suspect, see George Louis Beer, The Origins of the British Colonial System, 1578-1660 (1908; Gloucester, MA: Peter Smith, 1959) 169, 191-2, 208-15; Hillier, “Trade of the Virginia Colony” 248-52, 298-330, 341-2, 351-7; Jan Kupp, “Dutch Notarial Acts Relating to the Tobacco trade of Virginia, 1608‑1653,” William and Mary Quarterly 3rd ser. 30 (1973): 654; Rainbolt, From Prescription to Persuasion 56. On later violations of the Navigation Acts, see George Louis Beer, The Old Colonial System, 1660-1754, 2 vols. (1913; Gloucester, MA: Peter Smith, 1958) 1: 95-104; C. M. MacInnes, The Early English Tobacco Trade (London: Kegan Paul, 1926) 147-50; Gray, “Market Surplus Problems” 9; N. C. P. Tyack, “The Trade Relations of Bristol with Virginia during the 17th Century,” Master’s Thesis, Bristol U, 1930, 23; Morgan, American Slavery, American Freedom 198‑203. On losses at sea due to storms, privateers, and pirates, see Fitzhugh, William Fitzhugh 310, 313; Beer, Old Colonial System 1: 120-7; MacInnes, Early English Tobacco Trade 60-2, 140-1; Tyack, “Trade Relations with Bristol” 16-21; Middleton, Tobacco Coast 102-4, 289-350; Hillier, “Trade of the Virginia Colony” 144-56. On loss due to improper curing and prizing, see Middleton, Tobacco Coast 102-3. On loss of moisture in transit, see Middleton, Tobacco Coast 103-4; Menard, “Tobacco Industry” 113. On general deterioration in transit, see Hillier, “Trade of the Virginia Colony” 257-60. On smuggling into England (including misrepresentation of imports, illegal drawbacks, bribery, relanding, hawking by sailors, individual shipboard purchases), see The Case of the Planters of Tobacco in Virginia (London, 1733) 4-8; Bruce, Economic History of Virginia 1: 362-3n, 452-3; Beer, Old Colonial System 1: 148-9; MacInnes, Early English Tobacco Trade 53, 56-8; Alfred Rive, “The Consumption of Tobacco since 1600,” Economic History 1 (1926): 66-9; Alfred Rive, “A Short History of Tobacco Smuggling,” Economic History 1 (1929): 554-69; Lewis Cecil Gray, History of Agriculture in the Southern United States to 1860, 2 vols. (1933; Gloucester, MA: Peter Smith, 1958) 1: 251-2; Tyack, “Trade Relations with Bristol” 16, 25-7; Jerome E. Brooks, The Mighty Leaf: Tobacco Through the Centuries (Boston: Little, 1952) 51-2, 56, 89, 100, 103, 112, 156-7, 168-80; Middleton, Tobacco Coast 118-21; Jacob M. Price, “The Tobacco Trade and the Treasury, 1685-1733: British Mercantilism in Its Fiscal Aspects,” diss., Harvard U, 1954; T. C. Barker, “Smuggling in the Eighteenth Century: The Evidence of the Scottish Tobacco Trade,” Virginia Magazine of History and Biography 62 (1954): 387-99; Neville Williams, “England’s Tobacco Trade in the Reign of Charles I,” Virginia Magazine of History and Biography 65 (1957): 410; W. A. Cole, “Trends in Eighteenth-Century Smuggling,” Economic History Review 2nd ser. 10 (1958): 395-409; Darrett B. Rutman, The Morning of America, 1603‑1789 (Boston: Houghton, 1971) 63; Hillier, “Trade of the Virginia Colony” 224-6; Rainbolt, From Prescription to Persuasion 128-9; Paul G. E. Clemens, The Atlantic Economy and Colonial Maryland’s Eastern Shore: From Tobacco to Grain (Ithaca: Cornell University Press, 1980) 36-7, 114, 116. On intercolonial trade and smuggling, see Hillier, “Trade of the Virginia Colony” 334-75; Clemens, Atlantic Economy 88, 171-2. Surprisingly, domestic consumption of tobacco both in the Chesapeake and other colonies has drawn little attention from historians, but archaeological evidence and some commentaries by visitors suggest that Virginians of both sexes smoked pipes incessantly from a very early age. See, e.g., Durand de Dauphiné, A Hugenot Exile in Virginia, ed. Gilbert Chinard (New York: Press of the Pioneers, 1934) 111, 118.[/ref] The series is furthermore incomplete and entirely lacking in estimates for the years 1641-1668.[ref]Stanley Gray and V. J. Wyckoff, “The International Tobacco Trade in the Seventeenth Century,” Southern Economic Journal 7 (1940): 1-26; John R. Pagan, “Growth of the Tobacco Trade between London and Virginia, 1614-40,” Guildhall Studies in London History 3 (1979): 248-62; Menard, “Tobacco Industry” 113, 157-62.[/ref] The total also does not include tobacco left to rot in Virginia as a result of a shortage of shipping or excessive freight rates, laments frequently made by contemporaries especially in times of war or threat of war. As William Byrd I, writing in October 1690 to William Blathwayt observed: “Here is a considerable quantity of tobacco left in the country for want of ships to carry itt out, which must much prejudice Their Majesty’s revenue att home, as well as lessen itt here, which suffers allso by the great quantity’s of tobacco now prest into hogsheads, by reason fraight is now att such excessive rates.”[ref]Byrd, Correspondence of the Three William Byrds 1: 140. See also Byrd, Correspondence of the Three William Byrds 1: 19, 28, 97, 107, 159; Fitzhugh, William Fitzhugh 204, 214, 220, 239, 342; H. R. McIlwaine, ed., Journals of the House of Burgessesof Virginia, 1659/60-1693 (Richmond: n.p., 1914) 323; John Spencer Bassett, “Introduction,” The Writings of ‘Colonal William Byrd of Westover in Virginia Esqr.’, ed. John Spencer Bassett (New York: Doubleday, 1901) xxvi, xxxvi; Bassett, “The Relation between the Virginia Planter and the London Merchant” 564-6; MacInnes, Early English Tobacco Trade 141; Gray, “Market Surplus Problems” 5n27; Brooks, Mighty Leaf 112; Richard L. Morton, Colonial Virginia, 2 vols., Chapel Hill: University of North Carolina Press, 1960) 1: 327, 339; Richard Beale Davis, William Fitzhugh and his Chesapeake World 1676-1701: The Fitzhugh Letters and Other Documents, by William Fitzhugh (Chapel Hill: University of North Carolina Press, 1963) 22; Hillier, “Trade of the Virginia Colony” 296; Rainbolt, From Prescription to Persuasion 156; Morgan, American Slavery, American Freedom 240-1; Menard, “Tobacco Industry” 146-8; Gloria L. Main, Tobacco Colony: Life in Early Maryland, 1650‑1720 (Princeton: Princeton University Press, 1982) 21-3; Steele, English Atlantic 41-56; Price, Perry of London 31, 49, 52‑3.[/ref] The years before the Navigation Acts are highly suspect due to the prevalence of shipping direct to the Continent, while the years before 1696 are, according to Jacob M. Price, “not as satisfactory” as those after.[ref]U. S. Bureau of the Census, Historical Statistics 1162.[/ref] Recently, Robert C. Nash, drawing upon a close correlation between the impost on tobacco and customs totals, has significantly modified estimates for the years 1686-88.[ref]Robert C. Nash, “The English and Scottish Tobacco Trades in the Seventeenth and Eighteenth Centuries: Legal and Illegal Trade,” Economic History Review 35 (1982): 356.[/ref]
Nevertheless, historians have found a striking similarity in estimates of changing labor productivity determined using either aggregate or individual data from probate records, particularly during the late seventeenth century when the Navigation acts were more strictly enforced. They suggest that the import data provide “a reasonable approximation of the size of the crop.”[ref]Darrett B. Rutman and Anita H. Rutman, A Place in Time: Explicatus (New York: W. W. Norton, 1984) 9. See also Terry L. Anderson and Robert Paul Thomas, “Economic Growth in the Seventeenth-Century Chesapeake,” Explorations in Economic History 15 (1978): 385. The most recent and thorough study of the legal and illegal tobacco trade downplays the prevalence of smuggling in the late seventeenth and early eighteenth century. See Nash, “English and Scottish Tobacco Trades” 354-72. See also Lawrence A. Harper, The English Navigation Laws: A Seventeenth-Century Experiment in Social Engineering (New York: Columbia University Press, 1939) 246-71; McCusker and Menard, Economy of British America 77. Laments about rotting tobacco were based on the traditional belief that the stored crop would not survive the heat of a Chesapeake summer but by the 1660s, while planters may have still maintained a strong preference for selling while fresh and heavy, curing and packing techniques were improved enough that they were not above carrying over the crop for the next contingent of ships if freight was high or prices poor. See Gray, “Market Surplus Problems” 4-5; Rainbolt, From Prescription to Persuasion 63; Morgan, American Slavery, American Freedom 192; Steele, English Atlantic 42, 329n3. On traditional beliefs, see Bassett, “Relation between the Virginia Planter and the London Merchant” 566; Morgan, American Slavery, American Freedom 174; Main, Tobacco Colony 37. The increasing use of convoys should have helped increase available shipping during war years and offset losses at sea. See V. J. Wyckoff, “Ships and Shipping” 352; Middleton, Tobacco Coast 289-309; Morton, Colonial Virginia 1: 193‑4; Rainbolt, From Prescription to Persuasion 157.[/ref] However, in order to adjust for the impact of the availability of shipping on production estimates, one can include both a check variable on the prevalence of war in any particular year and an index of sailor’s wages as measures of both the dangers of shipping and general competition among shippers, all of which might have decreased available shipping to the Chesapeake.
Next we need to convert tobacco production into tobacco productivity per worker. Increases in tobacco production may have arisen from an increase in effort of the current tobacco labor force and/or shifts of laborers from other occupations and non-laborers into tobacco production. However, the only measure of the labor force we have–the total number of tithables–allows no such sophistication. The tithables–the basis for the head tax in Virginia and Maryland–represent in effect the legally defined potential tobacco labor force (black adults and white males of working age) by excluding those individuals whose age, sex, and race traditionally precluded their working in the tobacco fields. Thus the number of tithables includes those who might have but did not labor in tobacco (such as those employed in non-agricultural occupations) but does not include those who might sometimes have worked in the tobacco field (such as white women and young children).[ref]See, e.g., Lois Green Carr and Lorena S. Walsh, “The Planter’s Wife: The Experience of White Women in Seventeenth-Century Maryland,” William and Mary Quarterly 3rd ser. 34 (1977): 561-3; Clemens, Atlantic Economy 86; Main, Tobacco Colony 108-12, 175-82; McCusker and Menard, Economy of British America 236-7.[/ref] If as suggested by several contemporaries, the poverty resulting from falling tobacco prices forced tradesmen and housewives to begin working in the tobacco fields rather than their traditional occupations, we would not be able to distinguish this response from a simple increase in effort of the current tobacco labor force. We might expect an increase in both types of response to poverty, but it is possible that if the responses move in opposite directions and the shift in occupations proves significant, the impact of falling tobacco prices on tobacco productivity of the current tobacco labor force could be obscured.[ref]Anderson and Thomas, “Economic Growth” 384; Rutman and Rutman, A Place in Time: Explicatus 11.[/ref]
Although numerous lists of tithables exist for various counties for a few scattered years in the seventeenth century, creating an annual series for the entire Chesapeake is not straightforward. Overall, though, I believe that Russell R. Menard has reconstructed a plausible annual series.[ref]See Menard, “Tobacco Economy” 116-23, 157-61, 163-6.[/ref] Thus we will define tobacco productivity in terms of the ratio of tobacco imports into England (lagged by one year to reflect the delay between production and importation) to the number of tithables in the Chesapeake in any one year.
Besides tobacco prices, there are many other variables that could have had an impact on tobacco productivity quite independent of changing tobacco prices. Among the many potential independent variables affecting tobacco productivity, I will concentrate on those for which we have sufficient data. These include proxies for the quality of land, quality of labor, and changes in technology and organization independent of these other changes.
For quality of land, we have fairly good data on new land acquisition in Virginia in the late seventeenth century (which, given the lack of any comparable data for Maryland will have to serve as a proxy for land acquisition in the entire Chesapeake) but it is not at all clear whether such new land increased or decreased productivity.[ref]For the data, see Baird, “New Land Acquisition”. See further discussion of new land acquisition in Appendix III.[/ref] Some contemporaries, like the Burgesses in 1638, and some historians have suggested that new land promoted greater productivity and, indeed, constantly acquiring new, more fertile land was the only way that planters could offset lower tobacco prices. Edmund Plowden noted in 1648 that new land produced a pound per plant while worn land produced five to six plants per pound.[ref]Avery Odelle Craven, Soil Exhaustion as a Factor in the Agricultural History of Virginia and Maryland, 1606-1860 (Urbana: University of Illinois, 1926); Ulrich Bonnell Phillips, Life and Labor in the Old South (1929; Boston: Little, 1946) 28; Brooks, Mighty Leaf 96; Melvin Herndon, Tobacco in Colonial Virginia: ‘The Sovereign Remedy’ (Williamsburg, VA; Virginia 350th Anniversary Celebration Corporation, 1957) 7-9; Morton, Colonial Virginia 1: 133; 2: 456; Rainbolt, From Prescription to Persuasion 56; Main, Tobacco Colony 32.[/ref]
However, some historians have noted that by the late seventeenth century the only available land was more isolated from the main shipping traffic, further from navigable waters (thus tending to promote greater self-sufficiency and reduced tobacco production) and less fertile, and thus new land acquisition would have led to a decline in productivity.[ref]Davis, William Fitzhugh 20-1; Lorena Seebach Walsh, “Charles County, Maryland, 1658-1705: A Study of Chesapeake Social and Political Structure,” diss., Michigan State U, 1977, 405-20; Menard, “Tobacco Industry” 153; Main, Tobacco Colony 58, 262. See also Appendix III for discussion of demand for land.[/ref] We can model the impact of soil fertility with a measure of recently acquired land (which should have a positive coefficient if new land was more fertile) and isolation with a measure of total cumulative land acquisition (which should have a negative coefficient if market access decreased with the geographical size of the colony). Since it is not certain how long it took for new land to be placed in production after patenting, we will have to experiment with various time lags.[ref]For various estimates, see W. Craven, Southern Colonies 209; Main, Tobacco Colony 32. Economic historians analyzing cotton production in the antebellum era face many of the same definitional problems as those analyzing tobacco production in the colonial era. On the definition of an appropriate lag for the impact of land sales on production, see Gavin Wright, “An Econometric Study of Cotton Production and Trade, 1830-1860,” Review of Economics and Statistics 53 (1971): 111-2, 115.[/ref]
For measures of labor quality, the best proxy we have is the number of unindentured servants registered in any particular year, for which we have totals for several counties in Maryland and Virginia. Assuming that such servants, usually much younger and less skilled than indentured servants, were inferior in their productivity to both indentured servants and individual planters, one might suppose that an increase in the unindentured numbers independent of any change in the total number of tithables would lead to a decrease in productivity. (See further Appendix II.) Other historians suggest that freedmen were significantly more productive than servants in general, whether due to greater freedom which spurred effort or a tendency when unsupervised to sacrifice quality to quantity.[ref]Wheeler, “Lancaster County” 60-1; Edmund S. Morgan, “The First American Boom: Virginia 1618 to 1630,” William and Mary Quarterly 3rd ser. 28 (1971): 196.[/ref] Assuming that each county total captures most of the unindentured servants brought into that particular county, that taken in aggregate these annual totals reflect similar trends across the Chesapeake, that losses due to seasoning remained fairly constant over the time period considered, and that these servants served for approximately five years, we have a fairly good estimate of the changing size of the unindentured portion of the labor force.[ref]See discussion of demand for labor in Appendix II.[/ref]
One factor that is difficult to access is the impact of an increasing proportion of slaves in the work force. We simply do not have good annual aggregate numbers of slave imports in the years before 1698.[ref]See Walter Minchinton, Celia King, and Peter Waite, eds., Virginia Slave-Trade Statistics 1698-1775 (Richmond: Virginia State Library, 1984) ix.[/ref] Carr, Menard, and Walsh believe that the shift from servant to slave labor raised productivity levels for a time, whether due to the increased proportion of prime males in the work force or the ability of masters to extract more labor from slaves.[ref]Lois Green Carr and Lorena S. Walsh, “Economic Diversification and Labor Organization in the Chesapeake, 1650-1820,” Work and Labor in Early America, ed. Stephen Innes (Chapel Hill: University of North Carolina Press, 1988) 154-8; Lois Green Carr and Russell R. Menard, “Land, Labor, and Economies of Scale in Early Maryland: Some Limits to Growth in the Chesapeake System of Husbandry,” Journal of Economic History 49 (1989): 415; Lorena S. Walsh, “Plantation Management in the Chesapeake, 1620-1820,” Journal of Economic History 49 (1989): 394; Walsh, “Slave Life” 174-80.[/ref]
Labor quality might also have been affected by increasing tithable population. Although there is little hard evidence, one might suppose that increasing numbers of workers promoted a division of labor, but it is unclear what effect diversification would have had on tobacco productivity if specialized tobacco workers picked up the slack of artisans and craftsmen.[ref]On the theoretical link between population and diversification, see Ronald Eugene Grim, “The Absence of Towns in Seventeenth-Century Virginia: The Emergence of Service Centers in York County,” diss., University of Maryland, 1977. However, McCusker and Menard suggest that greater self-sufficiency tended to obviate the deepening of the domestic market made possible by greater population density. See McCusker and Menard, Economy of British America 128.[/ref] Furthermore, increases in the size of the work force could actually have allowed greater specialization in tobacco as a greater proportion of the labor force was freed from subsistence production. It is also possible that increased competition might have spurred greater increases in productivity above and beyond any impact from changing tobacco prices.
On technological or organizational changes in the seventeenth century unrelated to tobacco prices, there is little hard evidence. Historians have regularly suggested that most of whatever productivity gains due to technique occurred in the first half of the seventeenth century with agricultural practices fairly fixed from the mid-seventeenth century through the eighteenth century, the same for big and small planter alike (although the eighteenth-century saw significant economies of scale in combining grain and tobacco production for export).[ref]Thomas J. Wertenbaker, Patrician and Plebeian in Virginia, The Shaping of Colonial Virginia (1910; New York: Russell, 1958) 184‑5; Gray, History of Agriculture 1: 215-8; Morgan, American Slavery, American Freedom 142n33, 302; David W. Galenson and Russell R. Menard, “Approaches to the Analysis of Economic Growth in Colonial British America,” Historical Methods 13 (1980): 9; Main, Tobacco Colony 38; Russell R. Menard, Lois Green Carr, and Lorena S. Walsh, “A Small Planter’s Profits: The Cole Estate and the Growth of the Early Chesapeake Economy,” William and Mary Quarterly 3rd ser. 40 (1983): 192; Darrett B. Rutman and Anita H. Rutman, A Place in Time: Middlesex County, Virginia 1650-1750 (New York: Norton, 1984) 41-3; Carr and Menard, “Land, Labor, and Economies of Scale” 407-18; Walsh, “Plantation Management” 394; Walsh, “Slave Life” 173-4.[/ref] Lacking any better information, we can generally assume that any such change would be in a positive direction and might best be captured by a secular time trend.
There are, of course, numerous other variables which could have affected tobacco productivity. Unfortunately they do not readily lend themselves to quantitative measurement. For the impact of climate and other “acts of God”, we have little to go on apart from occasional comments in the historical record which suggest great variance from year to year in tobacco production due to natural forces.[ref]See, e.g., Fitzhugh, William Fitzhugh 126.[/ref] The impact of climate is further supported by general knowledge of the nature of agriculture in any time and place, but especially applicable to tobacco culture. As Menard notes:
Tobacco is a plant of extraordinary delicacy, and output and quality often fell sharply when the weather failed to cooperate. During especially bad seasons, production fell by as much as two-thirds or more, and years in which 20-30 percent of the crop was destroyed proved common.[ref]Menard, “Tobacco Industry” 128. See also Middleton, Tobacco Coast 102; Morgan, “First American Boom” 178n42; Morgan, American Slavery, American Freedom 110n11, 302; Anderson and Thomas, “Economic Growth” 384-5; Menard, Carr, and Walsh, “A Small Planter’s Profits” 181-2; Rutman and Rutman, A Place in Time: Explicatus 17.[/ref]
We might also include here the impact of such “manmade disasters” as Bacon’s Rebellion of 1676 and the tobacco cutting riots of 1682 which were reported to have led to the loss or destruction of much of the tobacco crop.[ref]Rainbolt, From Prescription to Persuasion 110-22.[/ref] Lacking any objective standard of the proportional effect of such natural or manmade forces, all we can do is model the effect for any particular year by a check variable in the years when the impact was said to occur to determine whether tobacco productivity in that year was significantly above or below expectations based on the pattern established in more normal years. However, to avoid overloading the model with independent variables (which create problems when we are working with as few observations as we have), we will restrict any analysis of the impact of climate to an observation of residuals of unexplained variance to see if certain years suggest significantly higher or lower productivity than predicted.
Finally, before launching into the actual multiple regression analysis, one should remember that while none of the data employed here are perfect, we will have to at some point simply let statistical analysis show us whether they are good enough to confirm or reject the hypotheses presented.
Test Results
The best-fit equation for tobacco productivity for the years 1669-1703 is presented in Table 1. (For a brief explanation of the multiple regression analysis employed here, see Appendix IV.) The statistical results show that, contrary to both the view of comtemporaries and historians, that planters responded to falling average tobacco prices and minimum tobacco prices by increasing tobacco productivity, and, perhaps more importantly, they reversed their behavior in response to rising prices by decreasing productivity. Figure 3 graphically reveals the closeness of the match between predicted and actual measures of tobacco productivity.
The regression analysis reveals that, for the period 1669-1703 inclusive, ceteris paribus, a fall in the five-year average price of tobacco (PA5) of a tenth of a penny (say from 1.0 to 0.9 pence sterling) led to an increase in tobacco production per tithable of 387 pounds while a fall in the four-year minimum price (PMIN4) of the same amount led to an increase of 73 pounds. Thus, due to the fall in tobacco prices, if all other things were held constant, planters, servants, and slaves in 1687 when prices were relatively low would have produced 1330 pounds per tithable more than they did in 1669 when prices were relatively high.[ref]1256 pounds due to fall in PA5 from 1.21 to 0.883 pence and 74 pounds due to fall in PMIN4 from 0.90 to 0.80 pence. Tobacco productivity actually rose only 445 pounds per tithable from 866 to 1311 pounds, primarily due to the influence of the downward time trend which reduced tobacco productivity by 677 pounds per tithable over the period.[/ref] On the other hand, with rising prices during the Peace of Ryswick, Virginians and Marylanders in 1701 would have produced 198 pounds less tobacco per tihable than they had in 1687.[ref]161 pounds due to rise in PA5 from 0.883 to 0.925 pence and 37 pounds due to rise in PMIN4 from 0.80 to 0.85 pence. Actual productivity fell by 450 pounds per tithable from 1311 to 861 pounds, due again primarily to the influence of the downward time trend which reduced tobacco productivity by 527 pounds per tithable over the period.[/ref] In addition, those years in which tobacco prices fell from the previous year, regardless of the amount or rate of fall (as measured by CHECK), averaged 44 pounds of tobacco per tithable greater than years in which there was a rise or no change in tobacco prices. Table 2 shows the robustness of the model for the years 1669-1703 to various other proxies of average tobacco prices (PAVG) employing alternate lags and weights of previous tobacco prices.
Increased land acquisition (LAND), at least in the short run, led to an increase in tobacco productivity of 3.4 pounds per tithable for every thousand acres acquired. With total tithables circa 1690 of about 30000 and assuming that new land employed a minimum of one tithable for every twenty acres, this suggests an increase in productivity of 2040 pounds for each worker in new soil. It is likely that our proxy for land quality captures other unintended factors in tobacco production decisions which parallel demand for land because of the weakening effect of including acreage estimates further back than the current year. More expected, the regression (as a result of the negative coefficient on SUM) tends to confirm the belief that by the late seventeenth century, newly patented land was of lower quality or more remote.
Our proxy for labor quality based on unindentured servant registration (UNIND) also seems to capture unintended effects which parallel tobacco production decisions. Based on total tithables circa 1690 of 30000 and assuming that the sample counties for servant registration captured about a fifth of the total registration (based on the number of tithables in the sample counties compared to total tithables in Maryland and Virginia), the regression suggests that each unindentured servant registered produced 3165 pounds more than the average tithable, clearly an unrealistic expectation. However, the strong positive coefficient does suggest both that unindentured servants were not inferior to other laborers and that freedmen were not more likely to emphasize quantity over quality when compared to servants. Furthermore, (as suggested by the strong positive coefficient on POP) an independent increase in productivity was derivable from the general increase in the number of tithables. If productivity fell with the increasing geographical expansion of the colony, the same can not be said of the increasing demographic expansion. This finding tends to downplay any negative impact of diversification on tobacco productivity and tends to suggest that, if anything, increasing population led to greater specialization and/or competition.
There was a statistically significant time trend (as captured by YEAR) of falling tobacco productivity of 38 pounds per year which could better capture secular declines in land and labor quality, but could also reflect a gradual shift over time to higher quality over quantity production, increasing diversification or self-sufficiency of the labor force over time, deflation in commodity prices, or other secular changes in the staple or domestic economy independent of changing tobacco prices. However, the time trend does challenge the view of historians that the rise in productivity in the late seventeenth century resulted from a shift to a quantity over quality strategy or that there were no secular declines in productivity due to soil exhaustion. To the degree that the negative time trend was due to secular deflation (meaning the money received from tobacco was more valuable over time in terms of the commodities it could purchase), this would tend to reinforce the finding that rising tobacco prices led to a fall in tobacco productivity.[ref]For evidence of secular deflation, see Menard, “Tobacco Industry” 143-4, 153; Warren M. Billings, John E. Selby, and Thad W. Tate, Colonial Virginia: A History (White Plains, NY: KTO, 1986) 121. However, although much effort has gone into developing a commodity price index for the seventeenth-century Chesapeake, the series available do not seem to significantly improve upon undeflated price and income proxies. On developing and employing a consumer price index for the Chesapeake, see Carville V. Earle, The Evolution of a Tidewater Settlement System: All Hallow’s Parish, Maryland, 1650-1783 (Chicago: University of Chicago, Dept. of Geography, 1975) 15-6; Lois Green Carr and Lorena S. Walsh, “Inventories and the Analysis of Wealth and Consumption Patterns in St. Mary’s County, Maryland, 1658-1777,” Historical Methods 13 (1980): 96-101; Menard, Carr, and Walsh, “A Small Planter’s Profits” 176n10.[/ref]
The analysis shows that estimates of tobacco production were significantly restricted by the availability of shipping but due more to competition for and risk of shipping (SHPWG) than wars per se (WAR). Indeed the presence of wars by itself promoted even greater shipment of tobacco from the Chesapeake and the total impact of war was thus ambiguous.[ref]Some historians have suggested that the threat of war led to greater speculation in tobacco.[/ref] During the Dutch war 1672-4 the two opposing tendencies cancelled each other out. However, the lack of shipping during the years 1689-96 reduced imports per tithable by 511 pounds from the years immediately preceding (1679-88) and following (1697-1701).
There was no clear pattern between contemporary observations of short or abundant crops and those predicted by the model, which interestingly suggests that climate and man-made disasters were not as important as expected in the variance in aggregate productivity although they most certainly impacted the yields of individual planters.
[cite]