Historical Demography Research Paper

1. A Social Science History

Historical demography is a social science history. It is distinguished from other branches of history by its explicit use of concepts and tools derived from a discipline in social sciences, i.e., demography in this case. Its goal is to establish historical facts by applying such conceptual frameworks to historical sources, and its methodological novelty lies in quantification. The research results of historical demography as such, have often questioned the conventional pictures of past populations, thereby throwing doubts about assumptions many theories rest upon. The position of historical demography in social and behavioral sciences thus resembles that of New Economic History, but it should be realized that there are a few notable differences in emphasis between these two social science history fields. While the application of theoretically constructed econometric models is given prime importance in New Economic History, emphasis in historical demography has always been on measurement. Attempts to formulate models to explain historical facts have never been put aside, but more effort has been directed to the development of finer measures and the identification of patterns, temporal and cross-cultural, by using those quantitative measures.

2. Sources, Methods, And Impacts

The innovation in methodology often goes with the discovery of new data sources. In historical demography too, its birth was closely tied with the French scholars’ application of family reconstitution method, a form of nominal linkage, to registers of baptism, burial, and marriage kept in Christian parish churches. Parish register demography, as one may call this type of research, is an attempt to reconstruct a demographic profile of the past based on a set of family histories reconstituted according to the rules set out by the demographer Louis Henry who turned to historical material to identify what he called ‘natural fertility,’ that is, fertility before the deliberate effort to reduce family size came to be practised.

Information that historical demography started supplying since then is a kind that they had always thought would only be available from modern vital records or sample surveys. Thanks to laborious works by historical demographers, demographers are now able to describe past populations in much the same vocabulary as for the present. However, there are a few drawbacks. In this kind of research, analysis ought to be restricted to persons whose burial dates are known, or who are known to have survived until their reproductive period was over. Since more people tended to leave the community with advancing age, it is very difficult to compute death probabilities across all the age groups, although a large enough sample, such as the one constructed by E. A. Wrigley and his associates at the Cambridge Group for the History of Population and Social Structure, does enable researchers to estimate age-specific patterns of mortality even from family reconstitution (Wrigley et al. 1997).

Furthermore, demographic research based on historical micro data has enabled the researchers to work with ‘cohort’ measures. Most demographic concepts, such as life table and marital fertility, are designed to portray the life events of a group of people who were born or married in the same period (termed ‘cohort’), but usually it is not possible for present-day demographers to single-out people born in a particular period and to trace them to the moment they die. Instead of using cohort measures, they have to rely on ‘period’ rates to account for demographic events in a specified period, in which the behavior of one cohort is bound to be mixed with that of other cohorts. In this sense, historical micro data have given demographers an opportunity to work with methodologically ideal populations.

For historians, on the other hand, historical demography has offered them a different set of merits. First, population becomes tractable as ‘population change’ and can now be decomposed into demographic components such as births, deaths, and migrations. This in turn enables them to relate those demographic phenomena to changes in economic circumstances. For example, one of the results of the earliest family reconstitution studies in France showed us how changes in grain prices were correlated with those in births and deaths. Similarly, one may examine interrelationships between time series of demographic rates and real wages in detail.

Another impact historical demography has had is being felt in social history. Since parish register demography reconstructs families and their life events that occurred in familial contexts, it has offered social historians invariable insights into the demographic domain of family life. Marriage, for example, is no longer a subject tied only with legal and ecclesiastical concerns. Who married whom, at what age, with what age gap between spouses—all these questions can now be answered. In addition, information about prenuptial pregnancy, gender differentials in infant and child mortality, and other related topics, can shed light on the intimacy of family life in the past.

Also related, but somewhat different approaches, have been adopted by another group of social historians. Statistical studies of the family and household, launched by Peter Laslett at the Cambridge Group in the 1960s, have been conducted in close association with research in parish register demography (Laslett 1983, Laslett and Wall 1972). The research along this line strongly suggests that types of family system, whether or not a new household was formed on marriage, whether the presence of live-in servants in the household was frequent or not, the mean age at marriage, and the frequency of celibacy, were all closely interrelated.

These links with social history have led to the burgeoning of family and household studies in history. Topics studied thus far include servanthood, gender inequality within the household, care for the elderly, widows and orphans, and the succession of headship and property. There are also issues that have enabled historians to get further into the inner domain of the individual. Abortion, infanticide, child neglect, and other means of family limitation embedded in traditional reproductive culture, are such issues.

However, source material being used in historical demography and its related research fields is not confined to the parish registers. In fact, there is a serious drawback in the use of parish registers as a source material. Since only people who registered baptism, marriage, and burial in a single community were recorded, it is impossible to reconstruct families who eventually left the community. Also impossible is to know how many people existed in that community at a particular point in time. All this implies that aggregative studies are extremely difficult to conduct, implying that the size of a settlement population, its birth and death rates, and in- and out-migration ratios can hardly be computed. Also difficult to conduct, for the reasons already mentioned, are mortality studies.

It is not surprising, therefore, that in many west European countries other micro data are being used. Especially useful are listings of households and population, most of which were one-off ‘censuses’ taken for various purposes. In other cultural regions, there are places where census-type records were regularly taken. In traditional Japan, for example, registers of village population were compiled by officials every year while in some parts of imperial China similar records were taken triennially. Such repetitive census-type data are useful not only for household analysis, but also for demographic studies since probabilities of entry or disappearance can be calculated at a regular interval, which, as monographs such as Smith (1977) and Lee and Campbell (1997) illustrate, allow researchers to draw demographic interpretations. One of the few shortcomings of such population registers is that it is very difficult to estimate the infant mortality rate since those born after the previous compilation, but died before the next, were not recorded at all.

3. Two Dimensions: Micro And Macro

Two broad areas of issues have been occupying a central place in the historiography of historical demography. One is the Malthusian models of positive and preventive checks, and the other the thesis of demographic transition. The former concerns premodern settings. The latter is about the transition processes from the pre-modern to the modern age, as Demeny (1972) put it: ‘In traditional societies, fertility and mortality are high. In modern societies fertility and mortality are low. In between there is the demographic transition. In both fields of research, however, it is useful to distinguish micro-behavioral from macro-aggregative dimensions.

3.1 Micro Demography

As noted in Sect. 2, parish register demography was born as a fertility-oriented research field. Malthus’s notion of positive check posits that overpopulation tended to be checked by a rise in mortality, which postulates that reproductive behavior of couples was governed by the invariable ‘passion between the sexes.’ The classic theory of demographic transition assumes that pre-transition populations were under ‘strong pressure on their members to reproduce,’ and hence that marital fertility was uncontrolled and its level close to the physiological maximum. Both Louis Henry, who coined the term ‘natural fertility,’ and Ansley Coale, who developed a technique to measure a deviation from the natural, uncontrolled fertility pattern in the course of his own project on the European fertility decline, assumed that if the couple has no target size of family and is not concerned with the existing number of children (termed ‘parity’), marital fertility is not controlled and hence its level must be invariably high. In other words, the decision made by a couple to end childbearing (termed ‘stopping’) is assumed to mark the onset of fertility limitation. Much effort, therefore, has been made to measure fertility levels and to identify patterns by examining age, parity, birth interval, and other factors associated with individual couples’ reproductive behavior (see Coale and Watkins 1986).

However, according to a large body of empirical studies so far conducted the original assumptions about reproductive behavior prior to the demographic transition are no longer tenable (see Cleland and Wilson 1987, Friedlander et al. 1999). First, pretransition of marital fertility was far more variable than previously imagined. Few populations exhibited levels comparable to that exemplified by the Hutterites, a small group of anabaptist settlers in north America, whose total marital fertility rate (TMFR, defined as the number of children a married woman would have within her reproductive period) stood as high as 11. In most cases, in fact, it was kept at moderate levels with their TMFR ranging from 6 to 9.

Second, it is confirmed that stopping behavior was never common among pre-modern populations. It should be noted, however, that some elite subpopulations, such as Catholic aristocrats in France and Italy and bourgeois families in Calvinist Geneva, exhibited a distinct stopping behavior, so that they can be regarded as forerunners of modern family limitation. Third, moreover, the couple’s behavior to ‘space’ births at particular intervals, as contrasted with stopping, seems to have been more important in accounting for the moderate levels of marital fertility in traditional times. Unlike stopping, birth spacing could have been achieved either through deliberate means such as contraception, or through unconscious actions such as breastfeeding, and many research results suggest that the role played by differences in feeding practices of the infant was unmistakable. Fourth, it is now widely recognized that in many instances marital fertility was rising, rather than falling, in the period immediately prior to the onset of modern fertility decline. This was the case in the UK and other European populations as well as in twentieth-century developing countries (Knodel 1988, Wrigley 1998). One of the factors suggested as having affected this rising tendency is a reduction in the level of breastfeeding, which was often associated with early phases of urbanization. Another is a decline in the incidence of stillbirths, which has recently been suggested for eighteenth-century England. The latter is a promising area of micro-demographic study, because even from family reconstitution, it is possible to investigate into the interplay between fertility and mortality measures at very early phases of life.

The conclusion that pre-transition fertility was not uncontrolled, be it consciously or unconsciously, demolishes the conventional divide between the modern and the pre-modern regime. Furthermore, findings from micro-demographic data analysis have made it difficult to see the demographic transition in a dichotomous framework. To what extent the transition was a behavioral revolution, still remains debatable.

3.2 Macro Demography

The concept ‘reproduction’ can also be applied to a population as a whole. It means the way in which the population replaces itself through natural processes of birth and death. The overall level of such replacement may be indicated by the total number of children that an average woman will have during her childbearing period. This is called the total fertility rate (TFR). Since it does not take mortality into account, the population’s natural process should be measured in terms of both fertility and mortality such as life expectancy at birth (e ). Thus one may describe historical populations by referring to the two, TFR and e . According to Livi-Bacci (1997), for example, Paleolithic populations and the Russians in 1879 happened to roughly exhibit the same level of TFR, a little lower than 6.5 children per woman. But mortality was different. The life expectancy is estimated to have been at below 20 for the Paleolithic, whereas that for Russia was a little below 30. As a result, the difference in the rate of natural increase was considerable. It was only marginally above zero in the Paleolithic case whereas, more than 10,000 years later, the Russian population was growing at somewhat over 1 percent per annum.

This, however, should not be taken to imply that the driving force in population history was always mortality decline, with fertility remaining high throughout the long, long period before the modern age. Indeed, over time both fertility and mortality may have exhibited substantial volatility. Neolithic settlers are supposed to have been characterized by fertility and mortality higher than Paleolithic hunters and gatherers, reflecting a combination of higher population density and lower mobility in the farming population. A much recent case in point is England from 1550 to 1800. Wrigley and Schofield (1981) have documented that from the late sixteenth century, when England’s rate of natural increase stood at below 1 percent per annum, the country’s TFR and e simultaneously declined towards the late seventeenth century. From that period to the second half of the eighteenth century the two measures came back to a position of higher fertility and higher life expectancy again. This time, however, the speed of rise in fertility exceeded that in life expectancy, so that the country’s population grew faster than 1 percent.

Since information from family reconstitution studies does not add up to macro, aggregate measures, complex and sophisticated estimation techniques, such as back projection, are needed to reconstruct the history of a past population at large. Judging from a few well-grounded back projection works, such as Wrigley and Schofield (1981) for England from the sixteenth to the nineteenth century and Weir (1994) and Bonneuil (1997) for eighteenth and nineteenthcentury France, first, there is now agreement that the role of nuptiality should be stressed in accounting for how overall fertility changed in pre-transition western Europe. The age at marriage was instrumental in adjusting the actual level of fertility in relation to changing economic circumstances. This what the classic Malthusian model of the preventive check predicted, thus suggesting that such a homeostatic mechanism was at work in the long- to medium-run time frame of the European past.

Second, this does not imply that mortality crises were not important. Short-term fluctuations in mortality were frequent in various regions in Europe, most of which were subsistence crises occasioned by crop failure. The frequency of famines, therefore, could account for the long-run level of death rates, but it should also be noted that changing levels of ‘exposure’ to viruses and other micro-organs were equally— sometimes more—responsible in determining that level of mortality in the past. The supposed link between mortality and the general standard of material life was thus less direct than often imagined (see papers in Bengtsson and Saito 2000).

It is true that during the period of fertility decline, neither nuptiality, nor mortality, could offset the declining trend of marital fertility and, hence, the corresponding rise in population. This, however, should not be taken to imply that economic growth, which accompanied this demographic transition in most cases, changed the whole relationship between the demographic variables. Neither industrialization, nor urbanization, was the crucial determinant of the onset of fertility decline. Even the modern decline of mortality was not in close association with an increase in the general standard of living.

Finally, mention must be made to the role of migration. The basic demographic equation tells us that a population will change according to the addition of births and immigrants, and the subtraction of deaths and emigrants. In other words, without information about in- and out-migration it is impossible to account for any macro change. Migration, therefore, is as important a topic as nuptiality, fertility, and mortality in historical demography.

Unfortunately, however, as historical sources rarely recorded actual population movements, not much has been done on this important topic. The historiography reveals that the volumes of in- and out-migration, distance, pull and push mechanisms are issues relatively better covered thus far. Much less known is the demographic impact of such migration flows. Were death probabilities in areas migrants headed for, different from those in their native communities? Was their marriage and reproductive behavior altered by their geographical mobility? Or was even their parents’ demographic decision-making affected by the prospect that their children would eventually leave for a place of, presumably, better opportunities? It is only in cases where no effect of these was substantial, that we do not need to know anything other than the rate of net migration (immigrants minus emigrants), which tended to be rather small. More specifically, any interpretation of results from family reconstitution studies rests upon the assumption that ‘leavers’ would behave just as ‘stayers’ did. All the demographic measures derived from this method only relate to the population of stayers, i.e., those born locally who continued to live in the their native community. The mean age at marriage, for example, could have differed between these two populations. Although it turns out that the difference was minimal in the case of English family reconstitution, the possible effects of migration on marriage as well as the other demographic variables can only be determined empirically (Wrigley et al. 1997, Chap. 5).

One of the few findings we have in this regard is about rural-to-urban migration. In a traditional set- ting the urban population was distinguished by a greater exposure to diseases and, hence, higher mortality than in the countryside. Obviously, even in such a mortality-dominant urban demography, the marriage market may also have been affected by the size of in-migration as well as that of labor demand. Moreover, the rural sector should not be considered demographically homogenous. There are European case studies suggesting that ‘proto-industrial’ villages exhibited an earlier age at marriage and higher fertility rates than farming villages. To what extent this was universal is debatable. What is certain, however, is that the destination and volume of each migration stream between the rural and urban sectors could affect the overall rate of population change in a preindustrial setting (de Vries 1984).

The picture becomes less clear for the period of modern fertility transition, which in many cases coincided with extremely large volumes of migration streams, both long-and short-distance movements. It is probably safe to suppose that rural-to-urban migration in this period also meant deterioration in the disease environment, while trans-oceanic migration to land-abundant settlements led to lower age at marriage and higher fertility. What is less clear is its influence on the population sending those migrants in the course of change. In the literature of demographic transition, migration, especially town-ward migration, tends to be regarded as having facilitated the decline of fertility. On the other hand, it is often noted that rural populations did not readily respond to changing economic and social circumstances by lowering their fertility. One suggested reason why it was still beneficial for peasant families to keep their fertility relatively high is the expanding prospects for the younger generation. When it was felt that they could emigrate and form their own households elsewhere, the older generation’s immediate need for the limitation of fertility may have been reduced. In nineteenth-century Scotland, for example, people left when circumstances allowed, and the same set of factors accounting for the emigration forced stayers to postpone marriage. Once married, however, they were less pressurized to limit the family size (see Andersdon and Morse 1993, Friedlander et al. 1999).

4. The Demographic Landscape In Cross-Cultural Perspective

At a more generalized level, historical demographers have developed a notion of homeostatic demographic regime. Many societies in the past had an ability to maintain equilibrium between their resources and the demographic characteristics. It is the ability to restore the original state of equilibrium, even if an external shock forces the system away from that position. When the ways in which this homeostasis is achieved are similar for a group of populations, it is probably appropriate to call that particular mode of interrelationships between the demographic variables a ‘demographic regime.’ Given empirical evidence so far accumulated, therefore, we may see the traditional west European populations as characterized by a particular mode of demographic regime. In this regime, both fertility and mortality levels were relatively low, and the key to its adjusting mechanism was nuptiality acting to ensure the system to restore equilibrium.

Turning to the nuptiality patterns, John Hajnal in his influential articles has shown that in traditional Europe, west of the line from St Petersburg to Trieste, the mean age at marriage was relatively high and celibacy not negligible, and that all this was closely tied with the nuclear family system and its associated principle of neo-localism (Hajnal 1965, 1983). In contrast, non-European countries exhibited comparatively low ages at marriage and very low proportions never married, suggesting that different demographic regimes were operating in other parts of the world.

Take China and Japan. According to in-depth regional studies in traditional periods, the level of life expectancy at birth (e ) in northern China was 32, while it stood at 39 in central Japan (see Lee and Wang 1999 for China, Bideau et al. 1997, Chap. 8 and, Liu et al. 2001, Chap. 7 for Japan). Compared with the western European average, China’s mortality was on the high side and Japan’s was on the low side. On the other hand, TMFR in both countries fell between 6 and 6.5, unmistakably on the low side of the European standard.

The Chinese and Japanese results raise some interesting questions. First, it is clear that the much lower level of marital fertility did not reflect stopping behavior, but spacing, thus confirming much of what the work on Europe has suggested. What is not certain at this stage is whether the observed long birth interval was a mere reflection of low fecundity, or a product of deliberate actions. While the factors accounting for the former vary from breastfeeding to malnutrition, the latter raises further questions: to what extent abortion, infanticide, and child neglect were practised, which of the three was most responsible, and whether it was motivated by despair or guided by the notion of ‘planning,’ however, rudimentary it would have been. Second, the nuptiality pattern was undoubtedly different from that of western Europe. There was no neolocal principle working. Nor was the age at marriage functioning so as to adjust overall fertility to economic conditions in Asia. Rather, nuptiality appears to have been more closely related to mortality. Indeed, the female mean age at marriage was unmistakably higher in Japan than in China, reflecting the aforementioned difference in e , although both countries seem to have shared the characteristics of universal marriage.

Third, however, if attention is turned to the status of females and family systems, we realize that there were some fundamental differences between China and Japan. For example, while no particular anomalies are observed from Japanese sex-specific life tables, the Chinese studies point out that there were persistent gender inequalities in mortality: the probability of girls dying between ages 1 and 5 was about 20 percent higher than that of boys, and such gender imbalance continued up to age 45. Clearly the status of girls and women was very different in traditional Chinese society. More specifically, it is the family system that differed. Predominant in Japan was the stem family system, under which the family household was vertically structured, and customs such as adoption, marriage, and remarriage were all for the family line to perpetuate. In China, on the other hand, what was operating was the rules of joint family. Under that system the household was laterally structured and the composition tended to be perennially complex.

To conclude, historical demography as it now stands, is both interdisciplinary and comparative and has raised new questions of its own, as it has evolved over time as well as cross-culturally.

Bibliography:

1. Andersdon M, Morse D J 1993 High fertility, high emigration, low nuptiality: Adjustment processes in Scotland’s demographic experience, 1861–1914. Parts I and II. Population Studies 47: 5–25, 319–43
2. Bengtsson T, Saito O (eds.) 2000 Population and Economy: From Hunger to Modern Economic Growth. Oxford University Press, Oxford, UK
3. Bideau A, Desjardins B, Perez-Brignoli H (eds.) 1997 Infant and Child Mortality in the Past. Oxford University Press, Oxford, UK
4. Bonneuil N 1997 Transformation of the French Demographic Landscape 1806–1906. Oxford University Press, Oxford, UK
5. Cleland J, Wilson C 1987 Demand theories of the fertility transition: An iconoclastic view. Population Studies 41: 5–30
6. Coale A J, Watkins S C (eds.) 1986 The Decline of Fertility in Europe. Princeton University Press, Princeton, NJ
7. Demeny P 1972 Early fertility decline in Austria-Hungary: A lesson in demographic transition. In: Glass D V, Revelle R (eds.) Population and Social Change. Arnold, London
8. de Vries J 1984 European Urbanization 1500–1800. Methuen, London
9. Friedlander D, Okun B S, Segal S 1999 The demographic transition then and now: Processes, perspectives, and analyses. Journal of Family History 24: 493–533
10. Hajnal J 1965 European marriage patterns in perspective. In: Glass D V, Eversley D E C (eds.) Population in History. Arnold, London
11. Hajnal J 1983 Two kinds of pre-industrial household formation system. In: Wall R, Robin J, Laslett P (eds.) Family Forms in Historic Europe. Cambridge University Press, Cambridge, UK
12. Knodel J E 1988 Demographic Behavior in the Past: A Study of Fourteen German Village Populations in the Eighteenth and Nineteenth Centuries. Cambridge University Press, Cambridge, UK
13. Laslett P 1983 The World We Ha e Lost—Further Explored. Methuen, London
14. Laslett P, Wall R (eds.) 1972 Household and Family in Past Time. Cambridge University Press, Cambridge, UK
15. Lee J Z, Campbell C D 1997 Fate and Fortune in Rural China: Social Organization and Population Behavior in Lioning 1774–1873. Cambridge University Press, Cambridge, UK
16. Lee J Z, Wang Feng 1999 One Quarter of Humanity: Malthusian Mythology and Chinese Realities. Harvard University Press, Cambridge, MA
17. Liu Ts’ui-jung, Lee J, Reher D, Saito O, Wang Feng (eds.) 2001 Asian Population History. Oxford University Press, Oxford, UK
18. Livi-Bacci M 1997 A Concise History of World Population, 2nd edn. Blackwell, Cambridge, UK
19. Smith T C 1977 Nakahara: Family Farming and Population in a Japanese Village, 1717–1830. Stanford University Press, Stanford, CA
20. Weir D R 1994 New estimates of nuptiality and marital fertility in France, 1740–1911. Population Studies 48: 307–31
21. Wrigley E A 1998 Explaining the rise in marital fertility in England in the ‘long’ eighteenth century. Economic History Review 2nd ser. 51: 435–64
22. Wrigley E A, Schofield R S 1981 The Population History of England, 1541–1871: A Reconstruction. Arnold, London
23. Wrigley E A, Davies R S, Oeppen J E, Schofield R S 1997 English Population History from Family Reconstitution 1580– 1837. Cambridge University Press, Cambridge, UK