I am indebted, above all, to Cormac ó Gráda for his patient encouragement and advice while this article was being written. Others from whose suggestions I have benefited are Robert C. Allen, Louis Cain, Charles Calomiris, James Dunlevy, Charles Feinstein, Paul Hohenberg, Jonathan Hughes, John Lyons, and John Walton.

¹ Deane, Phyllis, The First Industrial Revolution (2nd edn., Cambridge, 1979), pp. 268–69.Google Scholar

² See, however, Mokyr, Joel and ó Gráda, Cormac, “Poor and Getting Poorer? Irish Living Standards Before the Famine,” Economic History Review (forthcoming).Google Scholar

³ Ashton, T. S., “The Standard of Life of Workers in England, 1790–1830,” in Hayek, F. A., ed., Capitalism and the Historians (Chicago, 1954), pp. 123–55.Google Scholar

⁴ Lindert, Peter H. and Williamson, Jeffrey G., “English Workers ' Living Standards During the Industrial Revolution: A New Look,” Economic History Review, 2nd series, 36 (02 1983), pp. 1–25.Google Scholar The essay is reprinted in Mokyr, Joel, ed., The Economics of the Industrial Revolution (Totowa, 1984), pp. 177–205.Google Scholar

⁵ Crafts, N.F.R., “English Workers's Real Wages During the Industrial Revolution: Some Remaining Problems,” this Journal, 45 (03 1985), pp. 139–44.Google Scholar See also Lindert, Peter and Williamson, J. G., “English Workers' Real Wages: Reply to Crafts,” this Journal, 45 (03 1985), pp. 145–53.Google Scholar

⁶ Feinstein, Charles, “Capital Accumulation and the Industrial Revolution,” in Floud, R. and N. McCloskey, Donald, eds., The Economic History of Britain since 1700 (Cambridge, 1981), vol. 1, p. 136.Google ScholarCrafts, N.R.F., British Economic Growth During the Industrial Revolution (Oxford, 1985), p. 95.Google Scholar

⁷ Harley, C. Knick, “British Industrialization before 1841: Evidence of Slower Growth during the Industrial Revolution,” this Journal, 42 (06 1982), pp. 267–90;Google Scholar Crafts, British Economic Growth, chap. 2; G. Williamson, Jeffrey, Did British Capitalism Breed Inequality? (London, 1985), chap. 11.Google Scholar

⁸ Musson, A. E., The Growth of British Industry (New York, 1978), p. 149.Google Scholar

⁹ See H. Lindert, Peter and Williamson, J. G., “Reinterpreting Britain's Social Tables, 1688–1913,” Explorations in Economic History, 20 (01 1983), pp. 100–2;CrossRefGoogle ScholarWilliamson, Jeffrey G., “Earnings Inequality in Nineteenth Century Britain,” this Journal, 40 (09 1980), pp. 457–75; Williamson, Did British Capitalism, chap. 4.Google Scholar

¹⁰ O'Brien, Patrick and Engerman, Stanley L., “Changes in Income and its Distribution during the Industrial Revolution,” in Floud and McCloskey, eds., The Economic History of Britain, vol. 1, pp. 174–75.Google Scholar

¹¹ It is easy to show the effect of the revised growth rates on O'Brien and Engerman's calculations. Suppose we accept their initial figure that the bottom 83 percent of income receivers received 41.3 percent of total income in 1850, which is consistent with Williamson's estimates. If this entire group had not enjoyed a rise in living standards in the preceding fifty years, their share in 1800 (assuming a 0.7 percent growth rate) would have been about 58 percent of income (and not 75 percent as O'Brien and Engerman conclude on the basis of a more rapid growth rate). The 58 percent is still too high: Lindert and Williamson estimate that the bottom 90 percent earned about 54.6 percent of income in 1801/03, which would mean that the bottom 83 percent received about 51 percent. But the rate of growth of the real income of the bottom 83 percent implied by this computation is low: it comes to about 0.3 percent a year, less than half of aggregate income growth. If we were to look at the bottom two-thirds or 60 percent, the rate of growth would likely shrink even more. See also Mokyr, Joel, “The Industrial Revolution and the New Economic History,” in Mokyr, , ed., The Economics of the Industrial Revolution, pp. 40–1.Google Scholar

¹² J. Hobsbawm, Eric, “The British Standard of Living, 1790–1850,” reprinted in J. Taylor, Arthur, ed., The Standard of Living (London, 1975), pp. 76–77;Google Scholar Taylor, “Editor's Introduction,” in ibid pp. xxx–xxxi; Deane, , First Industrial Revolution, pp. 267–68;Google ScholarCrafts, , British Economic Growth, p. 98.Google Scholar

¹³ Great Britain, Parliamentary Papers (1898), vol. 85, “Custom Tariffs of the United Kingdom”Google ScholarMitchell, Brian and Deane, Phyllis, Abstract of British Historical Statistics (Cambridge, 1971);Google ScholarMathias, Peter, The First Industrial Nation (2nd edn., London, 1983), p. 200.Google Scholar

¹⁴ The series used below were adjusted to remove Irish imports and corrected to take into account new population estimates. See Data Appendix.

¹⁵ Davis, Ralph, The Industrial Revolution and British Overseas Trade (Leicester, 1979).Google Scholar

¹⁶ The least-square trends (expressed in percentages) estimated are:

¹⁷ Taylor, , “Editors's Introduction,” p. xxxi.Google Scholar

¹⁸ Deane, , First Industrial Revolution, p. 268;Google ScholarCrafts, , British Economic Growth, p. 98;Google ScholarBurnett, , Cost of Living, p. 180.Google Scholar

¹⁹ A recent analysis of household expenditures collected originally by David Davies and Frederick Eden in the 1790s reveals that tea and sugar products absorbed over 11 percent of total expenditures of English laboring families. Tobacco was even more popular. Even poorhouse inmates often used their scant earnings to purchase sugar, tea, and tobacco. See Shammas, Carole, “The Eighteenth Century English Diet and Economic Change,” Explorations in Economic History, 21 (07 1984), pp. 254–69CrossRefGoogle ScholarPubMed; Burnett, , Cost of Living, pp. 261–64. In the eighteenth century the tobacco habit had already become widespread, “touching all regions and classes.”Google Scholar See C. Nash, Robert, “The English and Scottish Tobacco Trades in the Seventeenth and Eighteenth Centuries: Legal and Illegal Trade,” Economic History Review, 2nd series, 35 (08 1982), p. 357. The lack of growth of tobacco consumption between 1730 and 1800 indicated by Nash's figures is consistent with stagnant living standards in this period.CrossRefGoogle Scholar

²⁰ For arguments doubting the value of these statistics because of smuggling, see Deane, , First Industrial Revolution, p. 286Google Scholar; Mathias, , First Industrial Nation, p. 201.Google Scholar

²¹ Cole, W. A., “Trends in Eighteenth Century Smuggling,” Economic History Review, 2nd series, 10 (08 1958), pp. 395–410;Google Scholar, H. C. and Mui, L. H., “Trends in Eighteenth Century Smuggling Reconsidered,” Economic History Review, 2nd series, 28 (02 1975), pp. 28–49. It can be readily seen why an increase in demand changes the proportion of goods smuggled: if the supply curves of legal and illegal imports have different elasticities, any change in demand will change the ratio between the quantities entered.Google Scholar

²² Wrigley, E. A. and Schofield, R. S., The Population History of England 1541–1871 (Cambridge, 1981), p. 529.Google Scholar

²³ B. Wilson, George, Alcohol and the Nation (London, 1940), pp. 101, 332.Google Scholar

²⁴ For England and Wales the leap occurs in 1827, when consumption jumps from 3,684 thousand gallons in the previous year to 7,407 thousand gallon. In Scotland the leap occurs in 1825, when recorded consumption rises from 2,303 thousand to 4,350 thousand gallons in one year. Wilson, , Alcohol and the Nation, p. 340; Mitchell and Deane, Abstract, p. 257.Google Scholar

²⁵ The relative prices of coffee and tea cannot explain the switch. If the relative price ratio (inclusive of duties) for 1818/20 is set to 100, the ratio averaged 100.31 in 1836/40 and 133.8 in 1841/45. Yet it was precisely in the years before 1840 that coffee consumption increased most rapidly. In their introduction to the chapter on living standards, Mitchell and Deane speak of a “well-known decline in the popularity of coffee” in the second half of the nineteenth century. See Mitchell, and Deane, , Abstract, p. 342.Google Scholar

²⁶ The rising trend in coffee consumption is also apparent in Davis's statistics. Unfortunately an error must have crept into his estimates for coffee imports and reexports for 1814/16, which he reports to be 648,000 cwt. and 940,000 cwt. The problem is not just that the difference is negative (possibly most reexports came from stockpiles accumulated during the wars). The numbers are simply too high: retained imports in 1814/16 were about 7.1 million lbs. (63,400 cwt.) and even at their peak in 1854/56 only reached 37.4 million lbs. (334,000 cwt.).

²⁷ In the mid-1850s, when coffee consumption was at its highest level, the value of retained tea imports exceeded that of coffee almost sixfold. Coffee was still a middle-class drink, enjoyed by relatively few people, whereas tea was a drink shared by all classes. See Barnett, John, Plenty and Want: A Social History of Diet in England from 1815 to the Present Day (London, 1966), p. 10.Google Scholar

²⁸ Morris, T. N., “Management and Preservation of Food,” in Singer, Charles et al., eds., A History of Technology (Oxford, 1958), vol. 5, pp. 37–38.Google Scholar

²⁹ See Feinstein, Charles, Statistical Tables of National Income, Expenditure, and Output of the U.K., 1855–1965 (Cambridge, 1976), table T4. For a detailed analysis of the Irish aspect of the standard-of-living debate, see Mokyr and ó Gráda, “Poor and Getting Poorer?”Google Scholar

³⁰ Before 1889, self-employment data are bundled with gross trading profits. To separate them the following procedure was carried out: let E be earnings from self-employment, G gross trading profits, and T = E + G their total. Before 1889 all we have is T, which we wish to decompose into E and G. Define k = E/G. I estimate the following regression equation for 1889–1913: This equation was projected backward to give a value for the years 1855 to 1889, and then ê was computed from.

³¹ The population as a whole may be used as the denominator if the reference group is more or less a constant proportion of the population.

³² Note that Q _MIN can be positive or negative. If it is negative, the good is not consumed until income reaches some threshold level.

³³ A comparison with Ireland, in which income was far below British levels, underscores this finding. In 1840/45 Irish legal imports of tobacco were 0.65 per capita, compared with 0.94 for Britain. If one allows for more smuggling into Ireland, a more limited level of monetization, and some local growing of tobacco, then at vastly different income levels, tobacco consumption per capita was of comparable magnitude (see Mokyr and ó Gráda, “Poor and Getting Poorer?”).

³⁴ See Thomas, Mark, “An Input-Output Approach to the British Economy, 1890–1914,” (Ph.D. dissertation, Oxford University, 1984), p. 375. Thomas dismisses the data on tobacco and alcohol consumption on the ground of underreporting.Google Scholar

³⁵ See, for instance, Parnell, Henry, On Financial Reform (London, 1831), chap. 4.Google Scholar

³⁶ Great Britain, Parliamentary Papers, (1844), vol. 12, “Report from the Select Commission on the Tobacco Trade” pp. 16, 31, 157.Google Scholar

³⁷ Williams, Neville, Contraband Cargoes: Seven Centuries of Smuggling (London, 1959), pp. 159–62, 173–75;Google ScholarN. Shore, Henry (Teignmouth, Lord), Smuggling Days and Smuggling Ways (London, 1892), p. 103.Google Scholar

³⁸ This assumption is supported by evidence that smuggling was carried out by small enterprises and that entry was impossible to block. See Shore, , Smuggling Ways, p. 206;Google Scholar and the discussion in Nash, , “Tobacco Trades,” p. 368 and references cited there.Google Scholar

³⁹ The tobacco series may also contain a bias favoring the pessimist conclusion. Because the duties were charged by weight, importers learned to enter the goods dried and stripped of the less valuable stalk. This could mean that each pound of tobacco represented more smoking material, and thus fewer pounds were necessary to satisfy given levels of demand. If, however, the improved quality was reflected in higher prices—as it should have been, since the improved quality was a result of more foreign value added and not technological progress—the higher recorded price would offset the bias. A higher price, given the quantity consumed, implies higher computed income, as equations 3 and 4 show. Given some plausible estimates on the income and own-price elasticities, the bias is not likely to exceed 5 percent (see Mokyr and ó Gráda, “Poor and Getting Poorer?” for more details).

⁴⁰ See Bythell, Duncan, The Sweated Trades: Outwork in Nineteenth-Century Britain (New York, 1978), pp. 149–50.Google Scholar

⁴¹ Burnett, John, Plenty and Want, p. 15.Google Scholar

⁴² In general, in a deflationary economy where all prices are falling, but the price of A is falling more slowly than the price of B, the bias will depend on whether the share of A in the average budget is rising or falling. If the index weights are set to the budget shares in the beginning of the period (Laspeyres index), and the share of A rises over time, the index overestimates the rate of deflation and therefore overestimates the improvement in the standard of living.

⁴³ Lindert and Williamson use the five budgets reported in Burnett, John, A History of the Cost of Living (Harmondsworth, 1969), pp. 260–63,Google Scholar which date from 1795, two from 1810, 1841, and one undated. The other entries in the urban cost-of-living index come from R. S. Neale's study of Bath, which also places a heavy weight on agricultural goods because the budgets refer to poor people. See Lindert and Williamson, “English Workers” (unpublished data, appendix B). Neale, R. S., “The Standard of Living, 1780–1844: A Regional and Class Study,” in Arthur, J. Taylor, ed., The Standard of Living in Britain in the Industrial Revolution (London, 1975), pp. 154–78.Google Scholar

⁴⁴ The “best guess” price computed by Lindert and Williamson ends in 1850, so this conclusion is based on the Rousseaux price index. According to R. S. Tucker's consumers price index, 1851 was the second cheapest year, after 1852. See Tucker, R. S., “Real Wages of Artisans in London, 1729–1935,” reprinted in Taylor, , ed., The Standard of Living, p. 30.Google Scholar

⁴⁵ M. W. Flinn, who concurs with the optimist findings, has argued for beginning the comparison in the years around 1812 instead of around 1819. Little in the discussion below depends on the starting point of the comparison. Whether 1810, 1815, or 1819 is chosen as the initial point, the increase in real wages remains above 1.5 percent a year. See Flinn, M. W., “English Workers' Living Standards During the Industrial Revolution: A Comment,” Economic History Review, 2nd series, 37 (02 1984), pp. 88–92.Google Scholar

⁴⁶ See Helling, Gertrud, Nahrungsmittel—Produktion und Weltaussenhandel seit Anfang des 19. Jahrhunderts (Berlin, 1977), for new estimates of consumption of wheat and meat products between 1800 and 1850 which show almost no improvements before the mid 1840s.Google Scholar

⁴⁷ Floud, Roderick, “New Dimensions of the Industrial Revolution” (unpublished manuscript, prepared for the Social Science History Association meeting, St. Louis, 10 1986).Google Scholar

⁴⁸ Wrigley, and Schofield, , Population History, p. 528.Google Scholar

⁴⁹ Feinstein, Charles, “Capital Formation in Great Britain,” in Mathias, Peter and Postan, M. M., eds., The Cambridge Economic History of Europe (Cambridge, 1978), vol. 7, p. 42 fn. 33.Google Scholar

⁵⁰ A dramatic illustration of the relative wages of workers in the modern and traditional sectors is provided by John Lyons, who estimates that the ratio of wages of adult factory weavers to handloom weavers went from 0.55 in 1815 to 1.87 in 1840/41 (see Lyons, John, “Family Response to Economic Decline: Handloom Weavers in Early Nineteenth-Century Lancashire” (unpublished manuscript), Miami University of Ohio, 06 1987).Google Scholar

⁵¹ The proportion of Ireland in the population of the United Kingdom, which was 31 percent in 1841, had fallen to 21.5 percent by 1855 and steadily declined over the rest of the nineteenth century, reaching about 11 percent by 1900. Moreover, the post-Famine population of Ireland, after the elimination of the poorest laborer and cottier classes, was more similar to the British population than the pre-Famine population.

⁵² The original source contains at least one serious error. For the year 1836 the tea series shows an unusually large bump. The reported quantity imported and retained that year was 49 million pounds, compared with 36.6 million in 1835 and 30.6 million in 1837 (p. 204). This figure is reproduced in Mitchell, and Deane, , Abstract of Historical Statistics, p. 356. Yet a glance at the source clearly indicates an error, as can be verified by dividing total duty receipts by the mean tariff, which yields only 44.5 million pounds.Google Scholar