The European Population: A Historical Data Handbook for 21 European Countries from 1850-1945

This contribution describes the project which is to publish a historical data handbook on the European population in 21 European countries, covering the time period of the first demographic transition, the years 1850-1945. All eighteen Western European nations are included plus the three Eastern European countries Czechoslovakia, Hungary and Poland. The handbook covers such topics as population development; the population structure by sex, age and marital status; the regional population structure; annual vital statistics developments in the fields of population growth and migration, fertility and legitimacy, infant mortality and life expectancy, nuptiality and divortiality. Furthermore, household and family data have been collected systematically using the population censuses as a basis. The book comprises several introductory comparative chapters and 21 national chapters for each country. Extensive documentation of available data and sources concludes the volume.

Background

The origins of this handbook go back to the research project 'Family Change and Family Policies in the West', started in 1994 at the Mannheim Centre for European Social Research (MZES) in Mannheim, Germany. The project had an international perspective, covering the main Western and three Eastern European countries: Austria, Belgium, Czechoslovakia, Denmark, Finland, France, Germany, Greece, Hungary, Italy, The Netherlands, Norway, Poland, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. The project also included some non-European countries: Canada, New Zealand, and the United States of America. From the beginning, Iceland, Ireland, and Luxembourg were not included, nor were the Eastern and southeastern European countries. The project produced standardized case studies for all project countries and two additional comparative volumes: one volume on family change, and one on family policies. While the project focused mainly on family policy and used demographic, household, family, and employment structures as a baseline and framework for the national family policy profiles, it soon became clear that in the realm of demography and household and family statistics the available data could not fulfil the project requirements. The project had a very long time horizon, reaching back - at least metaphorically - to the nineteenth century. The available data did not suffice for such an ambitious endeavour: they were mostly national data, but without long time series; and the international-especially European-data only went back to the 1960s. Therefore we decided to set up our own demographic database.
The work originated with a dataset compiled by the HIWED project. This dataset extended up to 1975 and included thirteen industrialized Western European countries: Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, The Netherlands, Norway, Sweden, Switzerland, and the United Kingdom. First, this dataset was checked for errors (which were numerous) and corrected. Second, it was updated from 1970-75. Demographic data available for the time periods between censuses become more and more inaccurate the longer the period of time between censuses is. Therefore the annual vital statistics data were corrected after the censuses of 1980/1, and these corrected data were used for the period after 1970. The annual time series were updated up to the early 1990s. (The same problem occurs with the vital statistics data from the 1990s, which will not be corrected before the censuses of 2000/1). Third, not only were data checked and extrapolated; in addition, new countries were added. Completely new time series were collected for Czechoslovakia, Greece, Hungary, Iceland, Luxembourg, Poland, Portugal, and Spain. Moreover, several new indicators were formulated and calculated: the illegitimacy rate, the crude legal separation rate, and the separation rate, among others. A completely new dataset was compiled for the population structure, containing for each census a table combining age in oneyear age groups with sex and the marital status. Finally, we also collected completely new data on household and family statistics. This field of historical European statistics had been completely untouched thus far: there were national collections as standalone products (for Germany see Rothenbacher, 1997), but household and family statistics were usually only rudimentary additions to general historical-statistical works. For the first time, the main available data wereeen compiled at the national level for the 21 countries included in this data handbook.
Given the very time-consuming task of compiling annual statistics for 21 countries covering over 100 years (the number of censuses used for data collection amounts to approximately 200), regionalizing this huge amount of variables, time series, and combinatory tables would have been impossible for only one or a few persons. This is a lesson learned from the Princeton Fertility Project, which took more than a decade to complete and a large staff of people to collect, process, and analyse data at the regional level for the subject of fertility alone.
The data are thus far complete to such a degree that not only individual countries can be compared, but countries can also be grouped into clusters and compared with other individual countries or country clusters. The dataset allows, for example, a comparison of the Nordic crude birth rate with the Mediterranean crude birth rate. Furthermore, European rates can be and have been calculated. Europe can be delimited in various ways: it is possible to create rates for Western Europe, for Central Europe, or for Continental Europe alone. Calculating European rates allows for very instructive comparisons of national rates with the European (weighted) averages. In principle, variation coefficients can be calculated and therefore convergence and divergence analysed.

Contents

The 21 country chapters have a standard format. Each chapter consists of eleven text sections and an appendix with tables and figures. The eleven sections are (1) State Formation and Territory; (2) Regional Population Structure; (3) Population Growth; (4) The First Demographic Transition; (5) Fertility and Legitimacy; (6) Mortality and Life Expectancy; (7) Marriage and Divorce; (8) Age, Sex, and Civil Status; (9) Family and Household Structures; (10) The National System of Demographic Statistics; and (11) Boundary Changes. Section 10 is furthermore divided into three subsections, presenting information on the development of official statistics in the fields of a) population structure, b) vital statistics, and c) households and families. Each chapter concludes with a standardized appendix containing tables and figures.

Structure of Country Chapters

We will now outline the different sections of the country chapters. A short introductory section on state formation and territory (1) presents essential background information necessary for understanding, using, and interpreting the statistical information presented thereafter. This section describes the political history of each country in terms of state evolution, date of independence, major changes in political regimes, and essential boundary changes through secession, mergers, or territorial losses. Not only politics, but also the longterm economic development, the economic position of the country in European comparison (the 'wealth of the nation'), and important features of the economic structure are described. In addition, major features of the social structure are highlighted which may have important and explanatory influences on population and demographic developments.
The section on regional population structure (2) deals with the internal population distribution of a country. Two indicators are used to describe the regional population structure: the proportion of each region's population in per cent of total population, and the population density (in inhabitants per square kilometre) of each region. The hundred years from 1850 to 1950 reveal major shifts in population distribution which are known as urbanization and rural exodus. These data show the main settlement structures of a country and the extent to which this structure changed during modernization.
The section on population growth (3) deals with the longterm growth processes during the first demographic transition. The main result of the demographic transition in all European countries was enormous growth in the European population and the population of the individual nationstates. Nevertheless, this development was very different according to the conditions in each country. Not only the longterm growth processes and macrosettings are discussed, but also the impact of wars, economic crises, and epidemics on growth rates. This section also describes major developments in net migration.
The next section describes the first demographic transition (4) of each country. The 'theory' of the demographic transition is in principle a model describing the development of societies from a state of high population turnover to a state of low population turnover. The model furthermore states that the death rate declined first and the birth rate reacted later also with a decline. The development of individual countries shows that this was not always true, and that in several countries the birth rate declined before the death rate. Nevertheless, the model is an important heuristic device for understanding this longterm process. The section describes the main features of the national process of demographic transition, each country's pretransition level, start and speed of transition. Explanations or interpretations of the individual characteristics of the demographic transition are given wherever possible.
The section on fertility and legitimacy (5) presents the data on legitimate and illegitimate fertility and on the proportion of illegitimate births to all births (the illegitimacy rate). The disaggregation of births by legitimacy and the calculation of agestandardized birth ratios by legitimacy reveals interesting and important aspects of family organization (the importance of cohabitation), illegitimacy, and attitudes towards the legal status of children. The causes of illegitimate fertility have been manifold and differ from country to country.
The section on mortality and life expectancy (6) presents and discusses the data on the infant mortality rate. This section is closely related to the section on the first demographic transition, because the infant mortality rate strongly influenced the crude mortality rate. Therefore, in a country where the infant mortality rate was high, the crude mortality rate was high as well. The national figures are described with reference to other European countries. The singularities of the national developments are presented and possible explanations are given. A second aspect of mortality is life expectancy which provides a much broader picture of mortality. Men and women are included and mortality is calculated also for higher age groups.
The section on marriage and divorce (7) deals with the marriage patterns in a country. Indicators used to describe nuptiality and marriage behaviour are the mean age at marriage, the proportion married at age 20-24, the marriage ratio, and the celibacy rate. The typical configuration of a country concerning these indicators is presented and the country's position with reference to such typologies as the 'European Marriage Pattern' is discussed. This section also deals with the long-term growth of marital instability due to divorce and legal separation.
The section on age, sex, and civil status (8) discusses the development of the population in a more disaggregated form, looking at the development of the age structure and population changes in the marital status, all according to sex. Major developments are population ageing, which in most countries was already visible in 1900, the increase of the proportion married until the mid-1930s, the lowering of the age at marriage, and the decline in celibacy.
In the section on family and household structures (9) the presentation is solely based on the available official household and family statistics collected by the statistical offices. Historical studies using primary sources such as original population census sheets or early population registers are not reviewed here; nevertheless, important relevant results have been included for explanatory purposes.
The section on the national system of demographic statistics (10) is documentary in character. It describes the available statistics concerning the introduction of investigation, the history of data collection, and the definition of statistical concepts for the three fields of population structure, vital statistics, and households and families. Especially important is the documentation of the definitions of statistical concepts, because only knowledge about the way data are collected and processed allows for a meaningful interpretation of the empirical facts. The documentation of the definition of statistical concepts is more important for household and family statistics, which were not standardized until after 1945.
The final section on boundary changes (11) provides information on the most important boundary changes necessary for understanding and interpreting the different population sizes and the demographic time series.
After the textual presentation there is a large section with appendix tables and figures, comprising six standard tables with statistical data, one documentary table, and several figures. All tables and figures included in this appendix have been standardized as far as possible. Appendix Table 1 documents the census dates and presents for each population census the most basic statistical information: the population by sex, civil status, and three age groups (0-14, 15-64, and 64+) in absolute and relative terms. Appendix Table 2 includes the regional population distribution for the different population censuses in relative terms. The proportion of each region in per cent of the total population has been calculated. Appendix Table 3 presents a different kind of regional data: it gives the population density measured by the number of inhabitants per square kilometre for each region and population census. Appendix Table 4 comprises demographic time series, if available, for the period 1850-1945. The time series are structured in the same way for all countries. They contain information on midyear population, two different population growth rates, migration, several fertility indicators, legitimacy, and various mortality, nuptiality, and divorce measures. Appendix Table 5 presents the development of life expectancy at various ages for both sexes. The Appendix Tables 6A-6E on households and families are less highly standardized due to the varying national statistics. But, wherever possible, the tables provide information on the main household types such as oneperson, family, and institutional households (absolute and per cent) together with the respective population living in these households (Appendix Table 6A). A second table presents the distribution of households by size in absolute terms (Appendix Table 6B) and a third table in per cent distribution (Appendix Table 6C). A fourth table gives average household sizes for different household types (Appendix Table 6D). A fifth table-if available-presents information on household composition (Appendix Table 6E). These five tables are included-if possible-as standard and are supplemented by additional tables if other interesting statistics are available. Such additional data may include disaggregation of households by socioeconomic status of the household head or regional information. Appendix Table 7 documents the availability of the individual vital statistics and population census variables. Appendix 8 includes several standardized figures on population by age, sex, and marital status. These figures are based on the population censuses. The number of figures varies between countries according to the availability of population census results.
All sources and references have been combined in one bibliography at the end of the volume. This bibliography has two main sections: sources and references. Sources are all statistical titles that have actually been used for this data handbook, while references are all the literature cited in the texts. The references have been arranged in alphabetical order. The sources have been subdivided, first into sources used in the comparative Introduction, and then into sources used for each of the 21 countries. For each country the sources have been further divided into three sections: (1) sources on vital statistics, (2) sources on population structure by age, sex, and marital status, and (3) sources on population census results on households and families.

Data Presented

This data handbook presents the population, demographic, and household data, collected in a standardized and systematic way for the 21 countries of Western and Central Europe whenever possible from 1850 to 1945. Several general guidelines have shaped the work from the beginning:

1. Territorially aggregated data: results have been collected at the level of the nationstates only, and regional data have only been collected for population size and population density.
   
   
2. Complete census coverage: all the population censuses in the time period have been covered, amounting to nearly 200 individual censuses in the 21 countries.
   
   
3. Disaggregated data collection: while there was no regional disaggregation, in other respects the data have been collected with as much detail as possible on a disaggregated level. Thus, vital statistics have been collected on an annual basis and have not been aggregated into time periods (quinquennia or decennia). Furthermore, the data on age, sex, and marital status have been collected in the most disaggregated way possible. Whenever available, oneyear age groups have been chosen, and all the different types and combinations of the marital status are included (there are more detailed types of marital status in some countries than the usual four types, single, married, widowed, and divorced).
   
   
4. Historical perspective: for all countries the collection starts in the nineteenth century, for some already in the eighteenth. The series are documented from 1850 in this data handbook. Nevertheless, in some cases the data collection reaches much further back into history.

Selected Comparative and National Results

Population Structure and Density

Population density continues to differ greatly by geographic and geoeconomic position of the European countries. Already since the Middle Ages the corridor from southern England via The Netherlands and Belgium down the Rhine Valley through Switzerland to Northern Italy was the most economically developed and most densely populated region of Europe. This remained the case in the nineteenth century and was even enhanced by the population revolution; external and internal migration processes added to it. Therefore, there is a supranational system of population distribution in Europe. Thus, the centre of population gravity in Europe since the sixteenth century has shifted to continental Europe north of the Alps.
It can even be said that since the decline of the Roman Empire, in a very longterm perspective, the centre of gravity of the European population shifted from the Mediterranean to north of the Alps. But this was only a gradual movement, and until the high Middle Ages, Italy for example was one of the most densely populated countries. Furthermore, the largest cities, such as Constantinople, Venice, and Genoa, were located on the Mediterranean. From the high Middle Ages to the sixteenth century there was a further shift from the Mediterranean to northern Europe, caused by the rise of the Ottomans and the loss of North Africa, parts of Spain and Portugal, and finally of most of the Balkans to the Muslims. As a result, the Mediterranean region was split between cultures; the trading routes were threatened by the resulting instability and increasingly shifted away from the Mediterranean region to overseas, resulting in the decline of Mediterranean trade and of such trading powers as Venice and Genoa. The major discoveries of new territory in the fifteenth and sixteenth centuries by the Portuguese, Spanish, Dutch, and British finally moved the economic and population centre to continental Europe.
Starting in the sixteenth century the Southern European powers Portugal and Spain suffered from an economic decline which became very apparent in the seventeenth century, when the powers of the North, mainly the Dutch and the British, became dominant. The South declined not only in terms of economic and military power, but also in terms of population (Pounds, 1979; Catalan, 1995), as can be seen from the number of inhabitants of the European countries. Germany, although not a seapower, had the largest population in 1870 with 41 million. France came second with 35 million; population growth in France had already been low since the late eighteenth century and continued during the nineteenth. The United Kingdom came third with 28 million inhabitants. Population growth in the British Isles and Ireland was quite vigorous, and the population of the United Kingdom exceeded that of France in the decade 1900-10; by 1930 the UK had six million more inhabitants than France. Italy ranked fourth in population with 27 million in 1870. Population growth was higher than in France, but far lower than in the United Kingdom. In 1930 Italy's population was equal to that of France. While France's population increased by only six million inhabitants, Italy's population went up by 14 million. Spain's population ranked fifth in 1880 with 17 million inhabitants, but population growth was quite strong until the 1930s, with an increase of approximately seven million people. Most other European countries were small or medium-sized in terms of absolute population, though some showed strong growth. Population growth was highest in The Netherlands: 'only' 3.6 million in 1870, the population doubled in the decade 1920-30. Further strong increases occurred in the Nordic countries, especially in Denmark, Finland, and Iceland.
There is no relationship between the size of a country in terms of territory and the size of its population: a large territory does not automatically lead to a large population and vice versa. When the different territorial sizes of the European countries (in square kilometres) are compared, France is the largest, followed in rank order by Germany, Spain, Sweden, Poland, Finland, Italy, Norway, and the United Kingdom. This list shows that the countries with the highest population growth are often those with small territories and that there is only little correlation between population size and size of the territory.
Population density of the European countries is given in Table 1. Population density is defined as inhabitants per square kilometre. In 1870 the most densely populated European country was Belgium with 164 inhabitants per sq. km. The United Kingdom was second, followed by The Netherlands, Italy, France, and Germany. This picture changed only gradually up to 1930: Belgium remained the most densely populated country with 265 inhabitants per sq. km., while The Netherlands moved into second place due to their strong population growth. The United Kingdom fell to third place, followed in order by Germany, Italy, and Luxemburg.

Table 1: Population Density in Europe (inhabitants per sq. km.)

Population Growth

Population growth is made up of three variables: the number of births, the number of deaths, and the extent of migration, i.e. of net migration. There are very different possibilities of combination (leaving migration aside for the moment): high fertility can go along with high mortality, resulting in medium natural population growth. High fertility can also occur with low mortality; in this case the natural population growth would be highest. Low fertility can be combined with low mortality, in which case natural population growth would also be medium. Finally, low fertility can be combined with high mortality, in which case natural population growth would be lowest.
To these four different combinations can be added the element of net migration (the difference between immigration and emigration) which can be high or low. Thus, a country with high natural population growth could have high net migration, reducing the overall population growth substantially. By contrast, a country with low natural population growth could have low net migration, with a similar effect for overall population growth.
It is not possible to determine a priori whether these three variables are high or low in a given country, as it depends on many factors related to the social, economic, and value (religious) structure of the population. The birth rate in preindustrial agrarian countries, for example, depends very much on the type of agricultural organization and the system of inheritance. The death rate depends heavily on the amount of infant mortality, which again depends on factors related to the educational level of the population, development of medicine, and sanitary infrastructure. Migration, finally, also depends on a variety of factors, such as the legal possibility to immigrate or emigrate, the population pressure due to overpopulation, expulsion of parts of the population due to political or religious conflicts (Huguenots, Puritans, Waldensians). Of all these, relative overpopulation (or relative underpopulation in countries receiving immigrants) is probably the most important factor causing migration.
Population growth could furthermore depend on the population density in a country at the start of the demographic transition. It might be postulated that the higher the population density in a country at the pretransition stage, the lower the population growth during the first demographic transition. The underlying supposition would be that high population density will cause people to limit reproduction due to the difficulties arising from high settlement density.

Figure 1: Population growth in Europe, 1850-1945 (1850, etc. = 100)
Figure 1 presents population growth rates from annual mid-year population figures based on the year 1850 (or 1871, 1919, etc.). The five countries with the highest population growth during the period 1850-1945 are Greece, The Netherlands, Denmark, England and Wales, and Finland. But Greece has to be removed from this list, because of its large territorial gains since independence. The five countries with the lowest population growth (in declining order) are Spain, Switzerland, Luxemburg, France, and Ireland. Ireland is the only European country in which population declined absolutely from 1850 to 1945, mainly caused by emigration. These patterns obviously do not confirm the above hypothesis that population growth will be low in countries that had a low population density already before the start of the first demographic transition. As early as 1870 The Netherlands and the United Kingdom had among the highest population densities in Europe. Nevertheless, Denmark and Finland, which are in the topfive group with the highest population growth, had a remarkably low population density in 1870. Let us now look at the country group with the lowest population growth: these countries by no means belong to the group with a high population density in 1870. In France, Luxemburg, and Switzerland, population density was on a medium level, while Spain's population density was low.
Mere number of people per square kilometre therefore does not explain population developments. Rather, there must have been different 'population regimes' (Bevölkerungsweisen, Gerhard Mackenroth) in European countries. These different population regimes are probably strongly related to the different European marriage patterns, to the unequal economic structure in Europe, leading to early industrialization in one country and persistence of agriculture in another, and to the type of agricultural organization in the countries still strongly dominated by agriculture until the mid-twentieth century.
The differing population growth of course had enormous consequences for the territorial population pattern in Europe. Looking at the population of individual European countries as a proportion of the overall European population of that time, we see that the population of some countries increased relative to the total European population, while that of others decreased. Countries whose population shares increased are The Netherlands, Denmark, Finland, Greece, and the United Kingdom. Countries with severe relative losses are mainly France and Ireland, and to some degree Spain.

Mortality and Life Expectancy in Europe

The model of the first demographic transition states that the mortality decline is the decisive factor for the fertility decline, because people reduced their fertility according to their children's improved chances of survival (Schofield, Reher, and Bideau, 1991). No matter which declined first in the demographic transition, mortality or fertility, the longterm trend in mortality is similar to that of fertility. The general trend is a decline in mortality at least since the 1880s, with some national variation. There were pioneering countries, in which mortality declined some decades earlier, as well as laggards, in which mortality fell later than the average. Until the 1940s there was a clear convergence in the crude mortality rate: mortality in those countries with high mortality around 1890/1900 declined faster than in countries where mortality was already low at that time. Already in the 1840s the national differences in the mortality rate had become fairly small. Nevertheless, the structure between countries remained very stable, thus reflecting rather invariable characteristics of European societies, such as geographical characteristics, country size, and religion.

Table 2: Infant Mortality Rate in Europe (deaths under 1 year of age in a given year per 1,000 live births of that year)

This surprising progress in mortality reduction cannot be interpreted in the sense that adult people lived much longer than before. Instead, the mortality decline during the first demographic transition was mainly a decline in infant mortality. The greatest progress was made in combating deaths in early childhood, mainly during the first five years, primarily due to the introduction of vaccinations (the first one was for smallpox shortly after 1800), sanitation, and breastfeeding, among others (Table 2). The first stage in the so-called epidemiological transition was therefore the fight against early childhood diseases. The secondmost dangerous phase of human life was young adulthood, which was mainly threatened by tuberculosis. Whereas much progress in reducing childhood infectious diseases was already made during the nineteenth century, the fight against tuberculosis began to be won only in the first half of the twentieth century. In the nineteenth century, if people reached age 60 they had a good chance of living to age 70 or more. Mortality reduction for the elderly was rather small until the 1940s; the major decreases in mortality of the elderly were achieved after 1945 and mainly since the 1970s.

Marriage Patterns in Europe

In 1965, John Hajnal distinguished between the 'European Marriage Pattern' and the 'Non-European Marriage Pattern' (Hajnal, 1965), with the dividing line between the two running from St. Petersburg to Trieste. This model claims to be valid for the time period until roughly 1940. The European marriage pattern is defined by a high age at first marriage, low nuptiality, and a large share of people remaining single for life. By contrast, the non-European marriage pattern is defined by a low age at first marriage, high nuptiality, and a small share of the population remaining single. According to this definition, the whole of Eastern and southeastern Europe belongs to the non-European marriage pattern. Hajnal's data refer to the time period until roughly 1900; in the subsequent time period, Hajnal concedes modernization processes on both sides: the non-European marriage pattern becoming more like the European marriage pattern and the European more like the non-European marriage pattern. The question here is whether this hypothetical convergence of marriage patterns for the time after 1900 can really be proved (see Rothenbacher, 1998).

From Universalization of Marriage to Deinstitutionalization

Until the end of the nineteenth century, in many countries access to marriage was restricted to persons having a certain amount of assets. This regulation did not originate in the ancien régime, but was introduced in the first half of the nineteenth century to avoid the pauperization of broad segments of the population. The towns in particular feared the burdens of providing communal care for the aged and the sick caused by the growing population exceeding the positions available in the labour market. Marriage was often the precondition for access to political rights, such as the communal right to vote, and thus was highly valued. Access to marriage was restricted particularly in the southern German states until 1867, whereas Prussia was more liberal in this respect (Matz, 1980). Marriage restrictions were removed in the territory of the Federation of Northern Germany (Norddeutscher Bund) with the federal treaty (Bundesordnung) of 1867 and for the whole of Germany in 1871. After that it was no longer necessary to have a certain income in order to marry. Marriage restrictions were also found in some other continental countries such as Switzerland. In general, the legal barriers to marriage were removed only in the last third of the nineteenth century. In all German states and in many European countries, there was a slight increase in the marriage rate in the 1870s. After the legal barriers to marriage had been removed, social and economic barriers increasingly came to the forefront. Economic and social developments, such as the spread of female employment, the increase in educational participation, the growth of employment in the public sector, and longer time spent in education, were new checks on the freedom to marry. These factors could be significant for the stagnating or declining nuptiality in Western Europe up to 1914.
The general trend in all European countries over the last 150 years is thus the universalization of the right to marry for all population groups and a declining interest of the state in marriage law. The law still regulates the age at marriage and the sex of the marriage partners and bars incestuous and polygamous marriages (Glendon, 1989: 38ff.). In a longterm perspective, the importance of bourgeois marriage as a model of familial behaviour increased after 1850. Thus, the general tendency can be called universalization of marriage regardless of social status and income position.
Since the 1960s a trend in the opposite direction has appeared, in family sociology called the 'deinstitutionalization of marriage and family' (Kaufmann, 1990; see also Meyer, 1993). Based on the sociological tradition of institutional analysis, it is claimed that the 'bourgeois' (or middle-class) family as a highly institutionalized model was valid until the 1960s, and that since then a reversal of this model has occurred. Institutionalization of marriage as a longterm process should therefore mean the societal acceptance of the model of the 'bourgeois' family, understood as the Parsonian normal family consisting of parents and two children.

Figure 2: Fertility and Legitimacy, Denmark 1867-1940
This problem can be dealt with under different aspects: first, the legal aspect of family and marriage law which create explicit norms. One may ask whether in this domain there are important processes of institutionalization and subsequently processes of deinstitutionalization. Second, the question concerning social structural changes has to be raised, exerting possible effects on these processes. Third, a society cannot be seen as a homogeneous block, but must be analysed as a stratified system with classspecific family and related social behaviour.

Marriage and Legitimacy

According to Goody (1983), in a longterm perspective the conjugal family as it exists today-with the monopoly on reproduction-is a product of Christianity, fighting for centuries against non-marital and extra-marital relationships and thus creating the problem of illegitimacy (Figure 2). A perspective orientated much more in economic history argues that illegitimacy is heavily bound to the respective mode of production. Thus, surprisingly, one of the highest proportions of illegitimacy in history was found in Catholic countries (Bavaria, Austria). The influence of the Church as monocausal factor is therefore not a sufficient explanation. In agricultural societies the extent of illegitimacy was, among other things, also dependent on the regulations of agriculture. Illegitimacy was fairly high in regions with single farms and inheritance by the principal heir; the structure of the agrarian economy in latifundia or family economies may also have played a role. Other factors, especially since the nineteenth century, are liberalization and the growth of lower strata in the process of industrialization. Another chain of argumentation could be that the extent of illegitimacy mainly depends on other demographic factors. Thus, the lower the age at marriage and/or the proportion of persons married, the lower the share of illegitimate births. Another important factor is probably the legal status of the illegitimate child, which differed greatly from country to country. In some German regions with inheritance law of the principle heir and in Austria, premarital, non-marital and extra-marital children were highly welcome as labourers and proof of fertility, explaining the very high rates in these regions

(Mitterauer and Sieder, 1982; Mitterauer, 1983).

Growth of Divorces and Liberalization of Divorce Law

The dissolution of marriage was not possible until the beginning of the nineteenth century, when it was only possible in some cases, if at all (Figure 3). The grounds for divorce were very limited. Between 1920 and 1965 there were no fundamental changes in divorce legislation, in contrast to the development in divorce rates. Exceptions can be seen in a longer list of grounds for divorce in England (1937) and the divorce legislation of the National Socialists in Germany and Austria (1938). Northern European divorce law is characterized by its flexibility. In Scandinavia there is a wide range of grounds for divorce and it is possible to get a divorce after a certain phase of marital separation, indicating the existence of the principle of breakdown of a marriage and divorce by mutual consent. In the United Kingdom flexibility is provided by the common-law system, based on case law and the principle of legal precedent, which gives the judge greater autonomy in deciding. In other parts of Europe divorce law is bound much more by legal regulation. The countries with law based on the Napoleonic Code-France, Belgium, Luxemburg, and The Netherlands-can be distinguished from Germany and Austria, with the German law tradition based on the civil code (Bürgerliches Gesetzbuch (BGB)); the BGB also exerts its influence on Swiss divorce legislation. The Code Napoléon was in principle valid until the 1970s. Divorce legislation based on the BGB was not as liberal as that based on the Code Napoléon. The BGB allowed for divorce only in case of violation of marital duties, and divorce by mutual consent was accepted only in a few cases.

Figure 3: Marriages and Divorces, Denmark 1850-1939

Population Ageing

Using Figure 4, some trends in age and sex structure, fertility, nuptiality, celibacy and life expectancy are discussed, taking Denmark as an example. The direction of the tendendies revealed by the Danish case is similar for most European countries, although level and speed of differ by country. Figure 4 reveals interesting developments in the age structure of the Danish population combined with sex and the marital status over 80 years, covering the period from 1860-1940. In 1860 the age pyramid still shows the pretransitional pattern of a population with nonexistent birth control. The age structure is concave with very large proportions of the age group 0-4 and 5-9. The cohorts of young adults in their twenties were comparatively small when compared with the age groups of the 35-39 year olds. Possibly during the 1820s there was an increase in the birth rate after the end of the Napoleonic wars and the hunger crisis of 1816. A very large proportion remained unmarried for a long time in their lives and age at marriage was rather high. People only started to marry during their twenties and thirties. In 1860 men married later than women. In the higher age groups, widowhood became an essential social status for women. There were much fewer widowed men due to lower life expectancy and therefore smaller cohorts and therefore higher chances to remarry when a man became a widower.
During the fourty years from 1860 until 1901 this structure changed only little. Nevertheless, first signs of the starting fertility decline can be detected. The concave pattern of the age pyramid became more linear, i.e. the youngest age groups (0-4 and 5-9) declined in size due to the spread of birth control. Age at marriage was lowered and the proportions of nevermarried persons had declined. The effect of both was an increased proportion of the population married.
During the next fourty years, until 1940, the picture changed dramatically. The age pyramid now took on to a convex shape: the birth decline since around 1900 reduced the proportion of people under 20 years of age considerably. The cohorts of the 0-4 year olds was as large as the one of the 15-19 year olds. Age at marriage had declined further and the proportion of people married had increased. In addition, there was already a slighty improvement in life expectancy for persons of very high age, e.g. for persons over 80 years old. Far more women than men now remained single for a longer time. The proportions of women never married were now higher than in 1901.

Figure 4: Population by Age, Sex and Marital Status, Denmark 1860, 1901 and 1940 (per 10.000 of total population)

References

Catalan, J. (1995), The Development of Two European Peripheral Economies in the Long Term: Poland and Spain, 1450-1990. MZES, Working Paper, AB III/No. 13. Mannheim: Mannheim Centre for European Social Research.
Glendon, M. A. (1989), The Transformation of Family Law. State, Law, and Family in the United States and Western Europe. Chicago and London: The University of Chicago Press.
Goody, J. (1983), The Development of the Family and Marriage in Europe. Cambridge: Cambridge University Press.
Hajnal, J. (1965), 'European Marriage Patterns in Perspective'. In D.V. Glass and D.E.C. Eversley, eds. Population in History. Essays in Historical Demography. London, 101-43.
Kaufmann, F.-X. (1990), 'Familie und Modernität'. In K. Lüscher, F. Schultheis and M. Wehrspaun, eds. Die 'postmoderne' Familie. Familiale Strategien und Familienpolitik in einer Übergangszeit. Konstanz: Universitätsverlag, 391-415.
Matz, K.-J. (1980), Pauperismus und Bevölkerung: Die gesetzlichen Ehebeschränkungen in den süddeutschen Staaten während des 19. Jahrhunderts. Industrielle Welt, vol. 31. Stuttgart: Klett.
Meyer, T. (1993), 'Der Monopolverlust der Familie: Vom Teilsystem Familie zum Teilsystem privater Lebensformen'. Kölner Zeitschrift für Soziologie und Sozialpsychologie 45: 23-40.
Mitterauer, M. (1983), Ledige Mütter: Zur Geschichte illegitimer Geburten in Europa. München: Beck.
-- and R. Sieder (1982), The European Family. Patriarchy to Partnership from the Middle Ages to the Present. Chicago: The University of Chicago Press; Oxford: Blackwell.
Pounds, N.J.G. (1979), An Historical Geography of Europe, 1500-1840. Cambridge: Cambridge University Press.
Rothenbacher, F. (1997), Historische Haushalts- und Familienstatistik von Deutschland 1815-1990. Frankfurt/New York: Campus Verlag.
-- (1998), 'Social Change in Europe and its Impact on Family Structures'. In J. Eekelaar and T. Nhlapo, eds. The Changing Family: International Perspectives on the Family and Family Law. Oxford, UK: Hart Publishing, 3-31.
Schofield, R., D. Reher, and A. Bideau, eds. (1991), The Decline of Mortality in Europe. Oxford: Clarendon Press.

Publication of Handbook

The data handbook will be published at Macmillan Press in the Series 'Societies of Europe': Rothenbacher, Franz and Peter Flora: The European Population, 1850-1945. London: Macmillan Press (forthcoming).