HISTORICAL DEMOGRAPHY

It is conventional to draw a distinction between historical demography and demographic or population history. Historical demography, strictly defined, is the application of the array of conventional demographic methods to data sets from the past that are sufficiently accurate for analysis. Such data sets may take the form of vital records and censuses, but most frequently, particularly if produced before the nineteenth century (i.e., in la période préstatistique), would not have been created for the purposes of demographic enquiry. Parish registers, militia or tax lists, testamentary records, and genealogies have been the most prominent among the great variety of documentary sources used by historical demographers. Demographic history may subsume historical demography as a field of enquiry, but is more wide-ranging in its subject matter, being just as concerned with charting the impact of demographic processes on society and economy as on measuring and accounting for demographic change per se. For instance, demographic historians would be more interested in investigating the effects of massive demographic losses, such as the effects of the plague outbreaks in mid-fourteenth century Europe on the later medieval economy and values, or the consequences for New World civilizations of the introduction of Old World diseases into the Americas. Historical demographers would be more interested in tackling the technical problems of measuring and assessing the accuracy of estimates of the resulting mortalities associated with such catastrophic episodes or phases.

While the above definitional distinctions may seem clear cut, in practice the contrasts between the approaches and their practitioners can be quite muted. However, this discussion will not undertake a review of the history of demographic thought or techniques that form another set of considerations of demographic practice in the past. Most of the pioneering demographers who have been influential since the seventeenth century such as John Graunt, William Petty, Richard Cantillon, Johann Süssmilch, Adolphe Quételet, William Farr, Jacques Bertillon, Wilhelm Lexis, and Alfred Lotka were engaged in the development of technical means through which they could better understand the demography of their own times rather than developing a set of procedures for the study of specific demographic pasts. It will nonetheless be necessary to see how historical demography became integrated into the social sciences more generally and why a self conscious historical demography emerged in the quarter century after World War II.

Historical Development and the Role of Family Reconstitution

Historical demography secured a formal status first in France at the Institut National d'Études Démographiques (INED) where Louis Henry had begun research after World War II on contemporary fertility and fecundity. He was handicapped in his investigations of these matters since by the mid-twentieth century those states that collected the most reliable vital statistics possessed populations that were controlling their fertility and those that had what he termed "natural fertility" did not for the most part have well organized and accurate systems of vital registration. He was therefore drawn to records from the deeper European past.

First, Henry exploited genealogical sources of the Genevan bourgeoisie from the sixteenth to the seventeenth centuries and among other findings revealed that these families were already controlling their fertility using some form of birth control by the late seventeenth century. Subsequently in 1958, he made a truly path-breaking move when he used the events recorded in the parish register of the Normandy parish of Crulai to reconstruct the lives of the individual families resident in that community in the eighteenth century. The technique that he developed came to be known as family reconstitution and established a means of using data sets that had accumulated in the parochial registers of the Christian Church to calculate for the first time detailed and accurate measures of fertility, mortality, and nuptiality for the centuries that preceded state-based systems of census taking and vital registration.

Prior to Henry's innovation historical demography in the era that lacked censuses and vital registration had no obvious means of measuring demographic stocks and flows so essential for the calculation of crude and age-specific rates. Henry's method of reconstitution made it possible to devise a set of rules to determine the period of time during which a particular family might be regarded as under observation. The technique was quickly adopted and modified for work on English parish registers, which exist in large numbers from the late 1530s. E. A. Wrigley, an economic historian and geographer at Cambridge University, completed the first of these English studies on the Devon parish of Colyton in 1966. The findings from this study attracted much attention because it seemed that the parishioners of Colyton were limiting their fertility within marriage in the late seventeenth century and were also suppressing overall fertility by raising female marriage ages significantly.

In the subsequent three decades a large number of reconstitution studies from various European countries were completed using parish registers. The largest national samples still derive from France and England, but there are significant totals from Germany and a growing number from Spain and Scandinavia. One pervasive theme in these studies concerns the analysis of marital fertility. Levels of marital fertility proved to vary substantially even though the communities in question all displayed the characteristics of "natural fertility" (i.e., they showed no tendency toward parity dependent control or "stopping" behavior). For instance, marital fertility in Belgian Flanders was 40 percent higher than that of England in the eighteenth century, although the two regions are separated by only a few miles across the English Channel. Likewise marital fertility was almost 50 percent higher in Bavaria than it was in East Friesland in the eighteenth and nineteenth centuries. By enabling a comparison of the intervals between marriage and first birth with those between first and second birth, and comparison of inter-birth intervals following the birth of infants that died within the first year of life with those following births that survived at least through the first year of life, family reconstitution made it possible to conclude that the principal determinant of such fertility variations was the incidence and duration of breast feeding.

Family reconstitution also made it possible to scrutinize the findings of a seminal paper by John Hajnal on the distinctiveness of European marriage patterns. In the early phase of historical demography's emergence as a recognizable sub-discipline, the 1965 paper drew attention–principally on the basis of evidence in nineteenth century northern and western European censuses–to a marriage pattern characterized by late female marriage and a high proportion of women remaining permanently single. Such a pattern was apparently absent from all other major world regions. Family reconstitution allowed marriage ages to be calculated by linking individuals from their baptism to first marriage and showed that the geography of marriage sketched out by Hajnal had a deeper chronological presence and was detectable at least from the early seventeenth century and had not emerged as a result of social and economic changes downstream from or associated with rapid urbanization and industrialization after 1750.

Beyond Family Reconstitution and The Testing of Malthusian and Demographic Transition Theory

Family reconstitution provided remarkably detailed information about European populations in the période préstatistique. However, it was a technique that was not without certain shortcomings. It was extremely time-consuming to perform and required very well maintained parish registers to yield reliable results. Consequently even in relatively thoroughly researched settings such as England there are still fewer than 40 parishes out of some 10,000 that have been investigated demographically using this method; in France, a major project using family reconstitution overseen by Henry was based upon a sample of just one percent of 40,000 French rural parishes. The rules of family reconstitution result in the bulk of the demographic data accumulating around individuals who do not migrate and hence there is always the suspicion that the results might be biased toward the immobile, particularly in societies such as England with high rates of movement among parishes. This flaw also means that the demographic characteristics of highly mobile urban populations are far less well researched than those of rural and small town settings. But perhaps the most serious difficulty arises from the fact that family reconstitution does not make it possible to compute aggregate measures such as crude birth and death rates or indices of reproduction or natural increase. Searching for a solution to this problem loomed large in the research of the Cambridge Group for the History of Population and Social Structure in the 1970s and 1980s.

The Cambridge Group was the first center exclusively devoted to historical demographic research, notwithstanding the resources devoted to this field in Paris at INED and later in the École des Hautes Études en Sciences Sociales. In Cambridge the intellectual enquiry was principally energized by the desire to understand why the population in England had grown so much more rapidly over the eighteenth century than anywhere else in Europe. Family reconstitution, which had promised so much, could not provide a clear-cut answer to this highly significant question. Fundamentally a means was needed to provide information about population stocks that could be used to complement the data on flows that were provided by the events recorded in parish registers. A technique to accomplish this was devised in the late 1970s, initially termed "back projection" by James Oeppen, building upon pioneering work by Ronald D. Lee. The technique has been developed further and is known in the early twenty-first century as generalised inverse projection (GIP). The data requirements for the technique are not particularly demanding since only annual totals of births and deaths are needed, stretching backwards from a census of proven reliability–especially in accurately recording the death of persons at the very oldest ages. A valuable attribute of GIP is that all the estimates of demographic variables are constrained to be mutually consistent. GIP generates estimates of population totals and age structures at any earlier date, of fertility in the form of gross reproduction rates, estimates of expectation of life at birth and related sets of age-specific rates, and net migration rates. The technique has made it possible to generate long-run fertility and mortality series for England, Denmark, Scania, and Tuscany.

European countries vary considerably in the extent to which birth and death series may be assembled. For instance, in many Catholic countries the burials of children who died before they were of the age to receive communion were often not entered in the registers in the early decades of registration, making death series problematic. Such series have made it possible to test in specific contexts the applicability of the Malthusian model concerning the relationship between fertility and mortality and measures of per capita income. The English case has received most attention and as a result of the outputs from GIP it is possible to trace the size of the English population between the sixteenth and the nineteenth century and to compare it with a measure of real wages over the same period. Periods of rapid population growth were associated with significant declines in real wages. In fact, between 1541 and 1801 population growth rates up to 0.5 percent per annum did not reduce living standards, but once that rate was exceeded living standards fell. It is noteworthy that in the English case this economic-demographic relationship was disappearing just at the moment that T. R. Malthus published his famous Essay in 1798.

Not all of Malthus's predictions stand up to empirical testing. Real wages in England rose significantly between 1650 and 1750, but expectation of life at birth fell quite markedly. Fertility, driven by nuptiality changes, appeared to move much more consistently with real wages in the manner postulated by Malthus, particularly in his more optimistic second edition of the Essay published in 1803. In 1700 the English intrinsic population growth rate was very close to zero, but by the early decades of the nineteenth century the rate had risen to approximately1.5 per cent, and over three-quarters of that growth could be attributed to a fall in the age of female marriage and a rising proportion ever married. Changes in marital and extra-marital fertility had little influence on the rate of population growth.

As a result of investigations of the kind carried out on the English demographic past historical demography moved to center stage in the debate over why the Industrial Revolution occurred and what its impact on demographic behavior had been. While this was a matter much discussed previously, it was not until historical demography had begun to generate reliable demographic data sets that economists acknowledged that the issues could be more clearly specified and historical economic-demographic relationships modeled formally.

Parish-register based demographic enquiry was only one strand in the demographic investigations of the Cambridge Group in the period from 1965 to 1985. Peter Laslett, the group's co-founder with Wrigley, pursued a linked enquiry into household and family demography and was concerned to construct patterns of household formation and population turnover. He came to realize that an essential correlate of late marriage in north-west European settings was the propensity of young adults to leave their natal hearths in their early-to mid-teens and to circulate as servants in the households of persons to whom they were generally unrelated, before marriage. At marriage they would for the most part establish nuclear households in communities where neither bride nor groom had been born. This feature was used by Hajnal to draw out a major contrast in the marital and household formation dynamics of historic and near contemporary societies distinguished by what he termed "north-west European household formation" rules and those in which "joint household formation" rules prevailed. The latter were distinguished frequently by early marriage of women who moved directly from their household of birth to that of marriage and in which sons tended to marry and co-reside patrivirilocally until household fission took place at a later point in time. Such research helped to justify a mode of enquiry in which demographic research was undertaken in conjunction with simultaneous investigation of family and social structure, ensuring strong disciplinary connections with historical sociology and anthropology.

As the quantity of research accumulated it became apparent by the 1980s that eighteenth and nineteenth century European demographic patterns were also noteworthy for their geographical variability. England and some other parts of Britain had rapid demographic expansion after 1750, but population growth in France was much more constrained, with fertility drifting down as life expectancy moved upward to sustain near zero-growth conditions. In contrast in Sweden, where mortality was far more volatile from year to year, fertility changed little, but demographic growth in the early nineteenth century resulted from a significant rise in life expectancy. The Swedish case is of particular interest since it appears to meet far more effectively the image of the pattern of long-term demographic change associated with classic demographic transition theory. Furthermore, because Swedish demographic data from the mid-eighteenth century had long been available, Sweden was erroneously used as a paradigm case. The innovations in the use of early data sets have, however, changed researchers' sense of the demographic landscape of what was for long deemed stage I of classic transition theory. This stage is assumed in pre-modern societies to be distinguished by high mortality and high natural fertility. The latter was viewed as varying little both regionally and through time.

Classic transition theory also emphasized the role played by industrialization and associated urbanization and its resulting improvements in well-being and medical science. These socioeconomic changes were seen as instrumental in shifting mortality away from a high plateau to lower and more stable levels and leaving fertility at pre-modern levels so that rapid demographic growth ensued, before fertility was adjusted downwards. The European Fertility Project, set up by Ansley Coale at Princeton University in 1963, investigated the fall in marital fertility in Europe from the late nineteenth century. Using evidence from early national census and civil registration offices, it confirmed that overall fertility was highly variable, reflecting major regional contrasts in nuptiality and marital fertility. Parishregister based research has subsequently shown that this feature extended back into the sixteenth and seventeenth centuries. Furthermore, the European Fertility Project also showed that there was a very limited correlation between the levels of economic development and living standards in European countries on the one hand and the resort to fertility control by their populations on the other. Such evidence from historical demographers working on the pre-transition and transition periods suggested that there was little to be gained by treating these epochs as distinct demographic systems and constituted a realization which drew historical demography more firmly within the larger fold of demographic practice.

Recent Developments and Future Preoccupations

The 1970s and 1980s marked a "golden age" of historical demographic research. While major issues that concern the characteristics of pre-transitional fertility and nuptiality and Malthusian theory still have pride of place in the sub-discipline, the subsequent years have seen a broadening of research interests and, particularly, research contexts. The West European center of gravity of early demographic research with its concentration upon nuptiality and Malthusian notions as a framework for understanding demographic dynamics has been challenged by a growth in research on non-European, particularly Asian, demographic regimes in the past and exploitation of household registers and genealogies as key demographic sources in such settings which, unless Christianized, lack parish registers. One theme, above all, has emerged suggesting that notwithstanding the prevalence in these areas of very early female marriage, population growth rates were generally no more rapid than those found in areas dominated by the European marriage regime. Extended breastfeeding, delayed starts to reproduction in marriage through spousal separation, abstinence, and abortion have been identified as means by which births were spaced out in China and Japan to produce total marital fertility rates that frequently fell below those found in Western Europe. Infanticide and child neglect often further constrained population growth. Researchers in these matters are now inclined to challenge a notion, prominent in the 1980s, that West European demographic growth rates were lower and more likely to have facilitated high savings levels, thereby facilitating longer term economic growth than would have been possible in areas of "joint household formation." A continued interest in further researching this theme is likely to dominate much future enquiry linking historical demography and the history of economic growth.

Use of population registers, both in East Asia and in some parts of Western Europe (in Sweden from 1750, in Netherlands, Belgium, and parts of Italy from the early-or mid-nineteenth century), is enabling a more sophisticated type of demographic research using event histories that make it possible to investigate demographic behavior both at the household and the individual level in a highly robust statistical fashion. However, the restricted geographical and chronological contexts within which this research method is feasible will likely limit the value of such research unless more effective means of creating longitudinal data sets can be achieved. Creation of such data sets would require linking censuses to the vital registration data that exist in large quantities for many societies both within and outside Europe from the beginning of the nineteenth century.

Research on mortality by historical demographers historically received far less attention than did investigations of nuptiality and fertility. Two highly distinctive positions dominated thinking on this topic: From the 1970s many argued that mortality was largely autonomous, being determined by the waxing and waning of epidemic disease and by climatic change, and was in no sense determined by human agency. Another argument claimed that mortality was largely determined by nutritional factors and only with the improvements in food supplies contingent upon rising agricultural productivity after 1750 did mortality decline as a result of growing resistance to infectious disease. British physician Thomas McKeown succeeded in promoting this latter view, particularly in medical circles.

Both positions regarding mortality in the past have been increasingly challenged as a result of growing realization that the inter-and intra-continental movements of population and the emergence of large metropolitan centers within emerging international trading systems had a major influence on levels of exposure and acquired resistance to infections–which in turn could lead to major changes in mortality levels, particularly among infants, children, and their mothers. Such influences were most likely responsible for the failure of model life tables based upon late-nineteenth and twentieth century population data to capture age-specific mortality patterns in historical populations. Those processes may also have been responsible for the existence of extended periods in the past when infant and child mortality could be seen to have worsened while the life chances of adults were static or improving.

Furthermore, McKeown's dismissal of the role of human intervention has been challenged by historians of public health who would argue that substantial declines in mortality arose from political interventions within urbanised societies in the late-nineteenth century that brought benefits to the poor as well as to the well nourished. New research on high and low status sections of European societies from the late Middle Ages has revealed very small differences in mortality levels and trends, especially after 1700 when adult life expectancies appear to have moved upward for all social status groups. There is growing evidence that adult female life expectancy in a substantial number of European countries has been moving upward in an unbroken fashion since about 1700 for all income groups.

Profitable research linking demographic patterns with anthropometric and paleodemographic investigation of skeletal remains and with data on human heights has been undertaken and there is considerable potential for more work of this kind over longer sweeps of time and in a variety of geographical contexts. The rising interest in adult mortality, particularly in declining mortality at the oldest ages, which is now regarded as the most dynamic demographic variable in many societies, has indirectly impinged on historical demography. Historical demographers are searching for more accurate measurements of change in that component as well as seeking to better understand the factors that may have influenced adult longevity. They are making use of the techniques of event-history analysis. This growth area of historical investigation is undoubtedly a refection of interest in contemporary demographic developments and once again reveals how greatly the orientation of research on the demographic past is determined by contemporary demographic concerns.

See also: Ancient World, Demography of; Cities, Demographic History of; Demography, History of; Family Reconstitution; Hayami, Akira; Health Transition; Henry, Louis; Household Composition; Laslett, Peter; Malthus, Thomas Robert; Paleodemography; Peopling of the Continents; Population Thought, History of; World Population Growth.

bibliography

Bengtsson, Tommy, and M. Lindstrom. 2000. "Childhood Misery and Disease in Later Life: The Effects on Mortality in Old Age of Hazards Experienced in Early Life, Southern Sweden, 1760–1894." Population Studies 54: 263–277.

Fleury, Michel, and Louis Henry. 1965. Nouveau manuel de dépouillement et d'exploitation de l'état civil ancien. Paris: Institut National d'Études Démographiques.

Fogel, Robert W. 1994. "The Relevance of Malthus for the Study of Mortality Today: Long-run Influences on Health, Mortality, Labour-force Participation and Population Growth." In Population, Economic Development and the Environment, ed. Kerstin Lindahl-Kiessling and Hans Landberg. pp. 231–284. Oxford and New York: Oxford University Press.

Gautier, Étienne, and Louis Henry. 1958. La population de Crulai, paroisse normande: Étude historique. Institut National d'Études Démographiques, Travaux et Documents, Cahier No.33. Paris: Presses Universitaires de France.

Gutmann, M. P., and G. Alter. 1993. "Family Reconstitution and Event-History Analysis." In Old and New Methods in Historical Demography, ed. David Reher and R. Schofield. pp. 159–181. Oxford and New York: Oxford University Press.

Hajnal, John. 1965. "European Marriage Patterns in Perspective." In Population in History: Essays in Historical Demography, ed. David V. Glass and D. E. C. Eversley. pp. 101–145. London: Arnold, and Chicago: Aldine.

——. 1982. "Two Kinds of Pre-industrial Household Formation System." Population and Development Review 8: 449–484.

Haines, Michael R., and Richard H. Steckel, eds. 2000. A Population History of North America. Cambridge and New York: Cambridge University Press.

Henry, Louis. 1956. Anciennes familles genevoises: Etude démographique XVIe-XXe siècle. Travaux et Documents, Cahier No 26. Paris: Institut National d'Etudes Démographiques.

Johansen, H. C. 2002. Danish Population History 1600–1939. Odense: University Press of Southern Denmark.

Knodel, John E. 1988. Demographic Behaviour in the Past: A Study of Fourteen German Villages in the Eighteenth and Nineteenth Centuries. Cambridge, Eng., and New York: Cambridge University Press.

Landers, John. 1993. "From Colyton to Waterloo: Mortality, Politics and Economics in Historical Demography." In Rethinking Social History: English Society, 1570–1920, and Its Interpretation, ed. Adrian Wilson. pp. 97–127. Manchester and New York: Manchester University Press.

Laslett, Peter, and Richard Wall, eds. 1972. Household and Family in Past Time: Comparative Studies in the Size and Structure of the Domestic Group over the Last Three Decades in England, France, Serbia, Japan, and Colonial North America, with Further Materials from Western Europe. London and New York: Cambridge University Press.

Lee, James Z., and Wang Feng. 1999. One Quarter of Humanity: Malthusian Mythology and Chinese Realities. Cambridge, MA, and London: Harvard University Press.

Lee, Ronald D. 1974. "Estimating Series of Vital Rates and Age Structure from Baptisms and Burials: A New Technique with Application to Pre-industrial England." Population Studies 28:495–512.

Lee, Ronald D., and M. Anderson. 2002. "Malthus in State Space: Macro Economic-demographic Relations in English History, 1540 to 1870." Journal of Population Economics 15: 195–220.

Livi-Bacci, Massimo. 1991. Population and Nutrition: An Essay on European Demographic History. Cambridge, Eng., and New York: Cambridge University Press.

McKeown, Thomas. 1976. The Modern Rise of Population. London: Arnold.

Oeppen, James. 1993. "Back Projection and Inverse Projection: Members of a Wider Class of Constrained Projection Models." Population Studies 47: 245–267.

Saito, Osamu. 1996. "Historical Demography: Achievements and Prospects." Population Studies 50: 537–53.

Wrigley, E. A. 1966. "Family Reconstitution." In An Introduction to English Historical Demography from the Sixteenth to the Nineteenth Century, ed. E. A. Wrigley. pp. 96–159. London: Weidenfeld and Nicholson; New York: Basic Books.

Wrigley, E. A., R. Davies, J. Oeppen, and R. S. Schofield. 1997. English Population History for Family Reconstitution 1580–1837. Cambridge, Eng., and New York: Cambridge University Press.

Wrigley, E. A., and R. S. Schofield. 1981. The Population History of England 1541–1871: A Reconstruction. Cambridge, MA: Harvard University Press.

——. 1989. The Population History of England 1541–1871: A Reconstruction. 1st paperback edition with new introduction. Cambridge, Eng., and New York: Cambridge University Press.

Richard M. Smith