1 Wrong, D., Population and society 2nd rev. edn. (New York, 1981), 25Google Scholar. See also, Davis, Kingsley, Human society (New York, 1949), 562–8.Google Scholar

2 Fuchs, Victor, Women's quest for economic equality (Cambridge, Mass., 1988).Google Scholar

3 Lopez, A. D., ‘The sex-mortality difference in developed countries’, in Lopez, A. D. and Ruzika, L. T., eds., Sex differences in mortality (Canberra, 1983), 1–23.Google Scholar

4 McKeown, Thomas, The modern rise of population (London, 1976)Google Scholar, and The origins of human disease (Oxford, 1988). Both books reduce all human mortality history to relative welfare in the form of nutritional advantages.Google Scholar

5 Preston, Samuel, Mortality patterns in national populations (New York, 1976)Google Scholar. Preston's statistical study of mortality patterns makes a strong case for the superior importance of public health techniques designed to minimize the spread of infectious disease.

6 Not all forms of disease are equally sensitive to nutrition levels. For a classification of diseases according to how nutrition levels affect host susceptibility, see ‘The relationship of nutrition, disease and social conditions: a graphical presentation’, Journal of Interdisciplinary History 14 (1983), 503–6.CrossRefGoogle Scholar

7 Dasgupta, Partha, ‘Well-being, foundations and the extent of its realization in poor countries’, Economic Journal 100, 400, supplement 1990, 1–32CrossRefGoogle Scholar. Dasgupta's approach to relative welfare is derived from earlier work done by Isaiah Berlin in his classic essay, ‘Two concepts of liberty’, in Four essays on liberty (Oxford, 1959).Google Scholar

8 Most, Carl and Johansson, S. Ryan, ‘Income and mortality: evidence from modern Japan’, Population and Development Review 12 (1986), 415–40Google Scholar. Until the nineteenth century in Western Europe and the early twentieth century in Japan, high-income, urbanized populations often had lower life expectancy levels than low-income, rural populations, because the biological influences to density on exposure levels were more important than any material advantages high incomes could buy.

9 The concept of maximum potential life expectancy is frequently used by gerontological biologists even though they cannot measure it. It generally refers to how long an individual or species could live under optimal environmental conditions. For practical purposes under modern conditions we can choose a value of approximately 85 for both males and females. See, Olshansky, Jay, ‘Using biological information to forecast mortality’, in Johansson, S. Ryan, ed., ‘Aging and dying. Biology, demography and human longevity’ (manuscript version)Google Scholar. See also, Olshansky, Jay, Carnes, B. and Cassel, C., ‘In search of Methuselah: estimating the upper limit to human life expectancy’, Science, 10 1990.CrossRefGoogle Scholar

10 Johansson, S. Ryan and Mosk, Carl, ‘Exposure, resistance and life expectancy: disease and death during the economic development of Japan 1900–1960’, Population Studies (1987), 207–35.CrossRefGoogle ScholarPubMed

11 This cultural approach to the foundations of health is explained in detail in Johansson, S. Ryan, ‘Cultural software, institutional hardware, and health information processing in social systems’, forthcoming in Caldwell, John, Findley, Sally et al. , eds., What we know about health transition: the proceedings of an international workshop, Canberra, May 1989. (Health Transition Centre, Australian National University, 1991).Google Scholar

12 For a discussion of individuals as ‘unencumbered’ versus ‘situated’ decision makers, see Benhabib, Seyla and Cornell, Drucilla, ‘Introduction. Beyond the politics of gender’, in Benhabib, Seyla and Cornell, Drucilla, eds., Feminism as critique (Cambridge, 1987), 1–30.Google Scholar

13 Waldron, Ingrid, ‘What do we know about the causes of sex differences in mortality? A review of the literature’, Population Bulletin of the United Nations 18 (1985), 59–76Google Scholar and Waldron, Ingrid, ‘Sex differences in human mortality: the role of genetic factors’, Social Science Medicine 17 (1983), 321–33.CrossRefGoogle ScholarPubMed

14 The denial of complexity and the segregation of history from theory in demography is similar to that which has characterized modern economics. See David, Paul, ‘Path-dependence: putting the past into the future of economies’ (manuscript version 07 1988). Forthcoming in the Journal of Economic Literature.Google Scholar

15 Graunt, John, Natural and political observations mentioned in a following index and made upon the bills of mortality (originally published in 1662)CrossRefGoogle Scholar. See the version reprinted by Gregg International Publishers in 1973.

16 Pearson, E. S., The history of statistics in the 17th and 18th centuries (New York, 1934).Google Scholar

17 Daston, Lorraine, Classical probability in the Enlightenment (Princeton, 1988), 130.Google Scholar

18 Pearson, , The history of statistics, 190–201Google Scholar. The life tables actually used by educated eighteenth-century men to estimate their survival chances made no allowance for sex, class or geographical location. Thus Edward Gibbon used George Buffon's life table to learn that at the age of 54 he had an even change of living 15 more years. See Carnochan, W. B., Gibbon's solitude (Stanford, 1987)Google Scholar. Although Buffon had created a generic human life table, he was aware that women seemed to have a small longevity advantage over men. But he regarded it as insignificant and explained it by appealing to the differential softness of women which retarded the ‘velocity’ of the force of mortality.

19 Pearson, , The history of statistics, 201.Google Scholar

20 Wrigley, E. A., Populations and history (London, 1969), 8.Google Scholar

21 For a long history of the origins of eighteenth-century health literature, see de Ropp, Robert, Man against aging (new York, 1960).Google Scholar

22 Digby, Anne, ‘The affluent patient’, ch. 4, in The patient's dilemma: medical relationships in Georgian and Victorian Britain, forthcoming from Cambridge University PressGoogle Scholar. For a general review of attitudes to sickness and health in the eighteenth century, see Porter, Roy and Porter, Dorothy, In sickness and health (Oxford, 1988).Google Scholar

23 ‘Risk factors’ are generally defined as ‘the various determinants of disease’. See Peterson, D. and Thomas, D., Fundamentals of epidemiology (Lexington, Mass., 1978), 24Google Scholar. In the popular mind they have now become associated with risks which are under total or partial individual control, and thus more linked to negative than positive rights.

24 Mosk, and Johansson, , ‘Income and mortality: evidence from modern Japan’, 415–20.Google Scholar

25 See Smith, Daniel S., ‘Differential mortality in the United States before 1900’ Journal of Interdisciplinary History 13 (1983), 735–59CrossRefGoogle ScholarPubMed. In the United States there were no marked class differences in life expectancy levels as late as 1900. In England the upper classes had gained a decided mortality advantage somewhat earlier. See Wohl, A., Endangered lives. Public health in Victorian Britain (London, 1987).Google Scholar

26 The government of Sweden was actually the first to publish fairly accurate statistics on the age-specific death rates of ordinary (largely rural) men and women. At the end of the eighteenth century Swedish females had a life expectancy at birth of 35.7 years versus 33.2 years for men. From age 1 to age 90 the life expectancy advantage of females remained consistently less than 2 years, even during the childbearing years. An English physician interested in the study of longevity reprinted some Swedish data, but did not offer an explanation for the small longevity advantage of women. See DrJameson, Thomas, in Essays on the changes of the human body at its different ages (London, 1811), 284.Google Scholar

27 For a review of the history of English demographic statistics see Glass, D. V., Numbering the people (Farnborough, Hants., 1973).Google Scholar

28 At present there is no standardized demographic definition of ‘excess female mortality’. In this essay, unless otherwise specified, it refers to an age-specific death rate that is absolutely higher for females than the equivalent rate for same age males.

29 Farr, William (1837) ‘Vital statistics of health, sickness and disease and death’. Farr's original essay is reprinted in Wall, Richard, ed., Mortality in mid 19th century Britain (Farnborough, Hants., 1974), 569.Google Scholar

30 See Gallagher, C. and Laquer, T., eds., The making of the modern body (Berkeley, 1987)Google Scholar. Most of the essays in this volume deal with the nineteenth-century physiology of gender, which was increasingly used to ‘demonstrate’ the pervasive biological inferiority of women.

31 Although nineteenth-century British cause of death data was the best for its time and place it was far from accurate or complete. For a review of its shortcomings see Szreter, Simon, ‘The importance of social intervention in Britain's mortality decline c. 1850–1914’, Social History of Medicine 1 (1988) 1–38CrossRefGoogle Scholar. There is no reason to suppose that it was less accurate for women than for men.

32 For a detailed analysis of Fair's use of vital statistics to explain nineteenth-century excess female mortality, see Johansson, S. Ryan, ‘Sex and death in Victorian England’, in Vicinus, Martha, ed., A widening sphere (Bloomington, 1977) 163–180.Google Scholar

33 For a review of the literature on corsets produced by nineteenth-century physicians see Pearsall, R., The worm in the bud (London, 1969), 115–17Google Scholar. Extremely tight-lacing practises on a constant basis could cause deformities of the chest and spine, and damage the liver and uterus.

34 Farr did find the resources to undertake costly research to which he assigned a high value. One of his most extensive projects involved elaborate calculations which tried to prove that the incidence of cholera was a function of height above sea level. See Smith, F. B., The people's health, 1830–1910 (London, 1979), 235.Google Scholar

35 Vicinus, Martha, ‘Introduction’ in Vicinus, Martha, ed., Suffer and be still. Women in the Victorian age (Bloomington, 1972), x.Google Scholar

36 Beddoes, Thomas, Hygeia, or essays moral and medical on the causes affecting the personal state of our middling and affluent classes, 3 vols. (Bristol and London, 1802), vol. 1, 12–20Google Scholar. For the importance of Beddoes in early nineteenth-century medical thought, see Porter, Roy, ‘Reforming the patient in the age of reform: Thomas Beddoes and medical practice’ (1989). In manuscript.Google Scholar

37 Fox, Joseph, ‘On the vital statistics of the Society of Friends’, Journal of the Statistical Society of London (06, 1859), 220.Google Scholar

38 This possibility was suggested by Susan Klepp (in correspondence).

39 For empirical research supporting this assertion see the articles published annually in Stress and Immunity, 1978 to the presentGoogle Scholar. See also Levine, S., Weiner, S. and Coe, C., ‘The psychoneuronendocrinology of stress’, in Levine, S. and Brush, F., eds., Psychoendocrinology (New York, 1989)Google Scholar. See also Locke, S. and Hornig-Rohen, M., Mind, immunity and behavioral immunology (New York, 1983)Google Scholar. For an epidemiological approach, see Cassel, John, ‘The contribution of the social environment to the host resistance’, American Journal of Epidemiology 104 (1976), 107–23Google Scholar. For the long run mortality effects of stress in early life see Eitinger, L. and Stro, A., Mortality and morbidity after excessive stress. A follow-up investigation of Norwegian concentration camp survivors (Oslo and New York, 1973).Google Scholar

40 Smith-Rosenberg, C., ‘The hysterical women: sex roles and role conflict in 19th century America’, in Sir Barry, G. et al. , eds. Health and wealth (London, 1963), 25–33.Google Scholar

41 Digby, A., ‘Woman's biological straitjacket’, in Mendus, S. and Rendall, J., eds., Images of women. The Victorian legacy (forthcoming).Google Scholar

42 Dr Guy, W., ‘On the duration of life among the English gentry’, Journal of the Statistical Society of London 9 (1846), 37–49.CrossRefGoogle Scholar

43 For a review of nineteenth-century maternal mortality rates in England and Western Europe see Schofield, Roger, ‘Did the mothers really die? Three centuries of maternal mortality in the world we have lost’, in Bonfield, L., Smith, R. and Wrightson, K., eds., The world we have gained (Oxford, 1986), 231–60.Google Scholar

44 Historisk Statistik for Sverige, vol. 1, Befolking (Statistika Centralbyran), 11–18Google Scholar. See also Johansson, S. Ryan, ‘Deferred infanticide: excess female mortality in childhood’, in Hausfater, G. and Hardy, S., eds., Infanticide: comparative and evolutionary perspectives (New York, 1984), 463–85.Google Scholar

45 Shorter, Edward, A history of women's bodies (New York, 1982) 240–1.Google Scholar

46 Johansson, , ‘Sex and death’, 169.Google Scholar

47 Wall, Richard, ‘Inferring differential neglect of females from mortality data’, Annales de demographic historique (1981), 119–40CrossRefGoogle Scholar. See also Roberts, Robert, The classic slum: Salford in the first quarter of the century (Harmondsworth, Middlesex, 1971).Google Scholar

48 Shorter, , Women's bodies, 228–30.Google Scholar

49 Condran, G. and Crimmons, E., ‘A description and evaluation of the mortality data in federal censuses: 1850–1900’, Historical Methods 12 (1979), 12.Google Scholar

50 Ginsberg, C. and Swedlund, A., ‘Sex-specific mortality and economic opportunity: Massachusetts, 1860–1899’, Continuity and Change 1 (1986), 415–46.CrossRefGoogle Scholar

51 Johansson, , ‘Deferred infanticide’, 479–84.Google Scholar

52 Snell, Keith, Annals of the labouring poor (Cambridge, 1988).Google Scholar

53 Sociologists were interested in demographic history only as part of modernization theory. Modernization theory is based on the idea that there are stark and permanent contrasts between the modern industrial/urban and pre-modern, agrarian, rural worlds. The contrasts drawn, however, are often not based on any real familiarity with the dynamic nature of the preindustrial countryside, because most sociologists were not trained as historians. If they thought of women as having mortality disadvantages these were attributed solely to pregnancy and childbirth.

54 By the 1920 the founders of the Johns Hopkins School of Hygiene and Public Health were convinced that science knew little or nothing about the rules of right living; but research concerning the harmful effects of tobacco smoking was actively discouraged as being irrelevant. See Fee, Elizabeth, Disease and discovery. A history of the Johns Hopkins School of Hygiene and Public Health 1916–1939 (Baltimore, 1989), 124–5Google Scholar. Right through the 1970s research on class differentials (conducted by sociologists committed to social reform) rejected the study of individual level risk factors as moralistic and designed to blame the socially disadvantaged for their demographic disadvantages. See Kosu, John, Antronovsky, Aaron and Zola, Irving, Poverty and health. A sociological analysis (Cambridge, Mass., 1969), 328.Google Scholar

55 Thompson, W., Population problems 3rd edn. (New York, 1942), 227–8.Google Scholar

56 Madigan, F. and Vance, R., ‘Differential mortality: a research design’, Social Forces 35 (1957), 193–5CrossRefGoogle Scholar. Russell, C. and Russell, W., ‘The life span of men and women’, Guy's Hospital Gazette 86 (1972), 30–6Google Scholar. Berkson, J. ‘Mortality and marital status; reflections on the derivation of etiology from statistics’, American Journal of Public Health 52 (1961), 15–18Google Scholar. Kitagawa, E. and Kitagawa, P., Differential mortality in the United States (Cambridge, Mass., 1983)Google Scholar. Gove, W. R., ‘Sex, marital status and mortality’, American Journal of Sociology 79 (1973), 45–67CrossRefGoogle ScholarPubMed. For the data in Table 3, see p. 59.

57 Montague, A., The natural superiority of women (London, 1970)Google Scholar. Montague's book was first published in 1952. In that edition the possibility of excess female mortality at any age was not even considered. By 1970 Montague made a brief reference to the possibility that excess female mortality ‘may’ exist in India (p. 79).

58 Sagan, Leonard, Health of nations (New York, 1986)Google Scholar. In Chapter 8, Sagan shows that conservatives are still using biological explanations to show that nothing can be done to reduce mortality differentials between professional and working-class men.

59 Madigan, F., ‘Are sex mortality differentials biologically caused?’, Millbank Memorial Fund Quarterly 35 (1957) 202–23.CrossRefGoogle ScholarPubMed

60 Preston, Samuel, Older male mortality and cigarette smoking (Westport, Conn, 1970).Google Scholar

61 Stolnitz, George, ‘A century of international mortality trends’, Population Studies 10 (1976), 17–24.CrossRefGoogle Scholar

62 Preston, Samuel, Mortality patterns in national populations (New York, 1976), Ch. 6.Google Scholar

63 The full title is The endangered sex: the neglect of female children in rural north India (Ithaca, 1982).Google Scholar

64 Miller, Barbara, ‘Female infanticide and child neglect in rural north India’, in Hughes, N. Schleper, ed., Child treatment and child survival. Society and the value of children (Dordrecht, Holland, 1988).Google Scholar

65 Premi, M. K., ‘The demographic situation in India’, Papers of the East-West Population Institute 80 (02, 1982)Google Scholar. Despite the more complete counting of Indian females in the censuses, male sex ratios have continued to increase until very recently. This indicates that excess female mortality was getting worse all through the first eight decades of the twentieth century. The implications for how many females' lives are actually lost each year is astonishing. Because India's population is so large, every year millions of females' lives disappear because of infanticide and relative neglect. See Sen, Amartya, ‘Women's survival as a development problem’, Bulletin of the American Academy of Arts and Sciences 43 (1989), 14–29.CrossRefGoogle Scholar

66 Johansson, , ‘Deferred infanticide’, 483–4.Google Scholar

67 Yankauer, A., ‘An approach to the cultural base of infant mortality in India’, Population Review 3 (1959) 39–51Google Scholar. The most recent survey of the cultural limitations on the household level choices made by women in the Third World is Daisy Dwyer and Bruce, Judith, A home divided. Women and income in the third world (Stanford, 1989).Google Scholar

68 Gulat, L., Profiles in female poverty: a study of five poor women in Kerala (Delhi, 1972).Google Scholar

69 Wadley, Susan, ‘Female life chances in rural India’, Cultural Survival Quarterly 13 (1989), 35–9Google Scholar. Wadley shows that social pressures force economically advantaged families to practise more, not less, discriminatory neglect towards daughters.

70 Western economists often assume that as income rises, or women become employed, excess female mortality will automatically disappear. This approach fails to take account of the cultural constraints on individual-level behaviour, an example being Schultz, T. Paul, ‘Women's work and their status; rural Indian evidence of labour market and environment effects on sex differences in childhood mortality’, in Anker, Richard, Buvinic, Mayra and Youssef, Nadia, eds., Women. Roles and population trends in the third world (London, 1982), 201–30.Google Scholar

71 Premi, , ‘The demographic situation in India’, 122.Google Scholar

72 Weller, R. and Bouvier, L., Population, demography and policy (New York, 1981).Google Scholar

73 This is a problem common to all the social sciences. See Ragin, Charles, The comparative method', moving beyond qualitative and quantitative strategies (Berkeley, 1987).Google Scholar

74 Waldron, Ingrid, ‘What do we know about the causes of sex differences in mortality?’, Population Bulletin of the United Nations 18 (1985), 59–76.Google Scholar

75 Waldron, Ingrid, ‘Sex differences in human mortality: the role of genetic factors’, Social Science Medicine 17 (1983), 321–33.CrossRefGoogle ScholarPubMed

76 ‘Mortality patterns and life expectancy of Seventh-day Adventists in the Netherlands’, International Journal of Epidemiology 12 (1983), 455–9.CrossRefGoogle Scholar

77 Liberson, Stanley, Making it count. The improvement of social research and theory (Berkeley, 1985)Google Scholar. Liberson points out that most social science research is covertly affected by selectivity biases of one kind or another. In the study of Dutch Adventists no data was supplied on occupation, income, or prior health screening. In addition women outnumber men three to two, and what this implies for marital status was ignored.

78 Verbrugge, Lois, ‘Gender and health: an update on hypotheses and evidence’, Journal of Health and Social Behavior 26 (1985), 156–82.CrossRefGoogle ScholarPubMed

79 Verbrugge, Lois, ‘Longer life but worsening health? Trends in the health and mortality of middle-aged and older persons’, Millbank Quarterly 62 (1984), 475–519Google ScholarPubMed. Recently, a health campaign has been started to get men to have annual examinations for prostate cancer. This cancer eventually affects one in ten men, but it can be cured if detected very early. Most cases are not detected early enough, because of widespread embarrassment among men about having a rectal examination. See Anstett, Patricia, ‘Prostate exam too important to skip because of embarrassment’, San Jose Mercury News, 19 09 1990, 2C.Google Scholar

80 Graham, Hilary, Women, health and the family (London, 1982)Google Scholar. In England there appears to be much more undiagnosed depression among women than men, although women have the normal life expectancy advantage (p.75). More and more studies of specific subpopulations with specific disease conditions do point to the continuing damage done by high levels of anxiety and depression among women. A recent study dealt with women who died from, rather than survived, second heart attacks. See Blakeslee, Sandra, ‘Study links emotions to second heart attacks’, New York Times, 20 09 1990, B7.Google Scholar

81 Nathanson, C., ‘Sex differences in mortality’, Annual Review of Sociology 10 (1984), 191–213.CrossRefGoogle ScholarPubMed

82 For a review of the value of programming metaphors for how cultural forces standardize (but not homogenize) human behaviour patterns, see Johansson, S. Ryan, ‘Does culture matter: the computer paradigm and the roles of cultural information in social systems’, Historical Methods 21 (1988), 160–89.Google Scholar

83 Loeb, Martin B., ‘Social role and sexual identity in adolescent males: a study of culturally provided deprivation’, in Spindler, George, ed., Education and culture: anthropological approaches (New York, 1967), 284–300.Google Scholar

84 Sexton, Patricia, The feminized male. Classrooms, white collars and the decline of manliness (New York, 1969).Google Scholar

85 Nathanson, , ‘Sex differences in mortality’, 126–7Google Scholar. For recent data on class-specific mortality ratios among white American men aged 25–64 years old, see Duleep, Harriet, ‘Measuring socio-economic mortality differences over time’, Demography 26 (1989), 345–51CrossRefGoogle Scholar. Duleep shows that by 1973–1978 the highest income American males had mortality ratios 30 per cent below average; the lowest income males had mortality ratios 60 to 80 per cent above average (p. 347).

86 Sagan, , Health of nations, 162.Google Scholar

87 A $2.00 a day cigarette habit costs $730.00 a year or $7,300 over a decade. Since working-class men earn less than professional men, the real costs of smoking are higher. This indicates a great willingness to pay for the psychic benefits of smoking.

88 McCord, C. and Freeman, H., ‘Excess mortality in Harlem’, The New England Journal of Medicine 322 (18 01 1990), 173–7.CrossRefGoogle ScholarPubMed

89 Kosa, John, Zola, I. and Antonovsky, A., ‘Health and poverty reconsidered’, in Kosa, John, Antonovsky, A. and Zola, I., Poverty, health and sociological analysis (Cambridge, Mass., 1969), 321–34.Google Scholar

90 Sontag, Susan, Illness as metaphor (New York, 1977)Google Scholar. Sontag has also discussed how blaming the victim arguments have been successfully countered by those involved in the campaigns to help those stricken with AIDS. See ‘AIDS and its metaphors’, The New York Review of Books (27 10 1988), 88–9.Google Scholar

91 Goldberg, Herb, The hazards of being male. Surviving the myth of masculine privilege (New York, 1976).Google Scholar

92 Benhabib, and Cornell, , Feminism as critique, 12–14.Google Scholar