List of provisional abstracts

Theme 1: Cohabitation and marriage: Modern family dynamics in the Nordic countries
Theme 2: Fertility trends in the Nordic countries
Theme 3. Long term perspectives on family and women's work during the 1900s
Theme 4a: The History of Population Statistics
Theme 4b: Historical Population Estimates
Theme 5: National systems of population statistics in the Nordic countries
Theme 6. Population forecasts
Theme 7. Third world demography
Theme 8. International migration
Theme 9. The immigrant population
Theme 10. The integration of immigrants in the labour market
Theme 11 [cancelled]
Theme 12. Domestic migration and regional population trends
Theme 13: Historical fertility and nuptiality patterns, 1700-1900
Theme 14a: Historical perspectives on infant and child mortality
Theme 14b: Public health in history
Theme 15: Mortality and health studies
Theme 16: Historical demography of indigenous people in the Nordic countries
Theme 17: Historical demography - Migration

Theme 1: Cohabitation and marriage: Modern family dynamics in the Nordic countries

Co-ordinator: Eva Bernhardt, Center for Women's Studies, Stockholm University Discussants: Gunnar Andersson, Max Planck Institute for Demographic Research, Rostock; An-Magritt Jenssen, Dept of Sociology and Political Science, Norwegian University of Science and Technology, Trondheim

Love and Ownership: Marriage, Cohabitation, and buying a Home in Sweden

Nathanael Lauster, Sociology Department, Stockholm University

Cultural analysis suggests that the ability to buy a home may serve as a symbolic requisite for entry into marriage. With reference to class, the ability to buy a home serves as an indicator of one's success in life, and one's preparedness to settle into a more 'formal' family. At the same time, the shared investments involved in looking to buy a home are likely to increase the degree of commitment in a relationship. My research attempts to determine the impact of various factors associated with the ability to buy a home on observed marriage and cohabitation patterns for individuals in contemporary Sweden. This approach refines the emphasis on often-used demographic variables (employment, income, education, etc.), to consider a variable, the ability to buy a home, more culturally meaningful to actors engaged in family formation decisions.

The 1992 Swedish Family Survey contains life history data on entry into cohabiting unions and marriage for 1,666 men and 3,317 women born between 1949 and 1969. The population registry, in turn, contains rough data on income and geographic location for these respondents. By geographically matching this data with regional and recent historical variations in housing prices and programs (ability to secure loans, etc.), I hope to explain a great deal of the variation in both marriage and cohabitation.

I hypothesize that the ability to buy a home will have a significant impact on marriage and cohabitation. Those more able to buy a home are more likely to enter unions, and, in particular, marital unions. Those less able to buy a home are more likely to avoid unions or remain in cohabiting unions rather than entering marriage. Recent historical variations in housing prices and programs are likely to explain much of the decline observed in entry into marriage between the late 1960s and early 1990s. Regional variations in housing prices, availability, and programs are likely to explain many regional differences in marriage and cohabitation, particularly rural and urban divides. Policy implications of findings will be briefly discussed.

Leaving home in a multiple destination perspective: evidence from Danish register data

Isabella Gomes Carneiro and Lisbeth B Knudsen, Danish Center for Demographic Research, University of Southern Denmark, Odense University

This paper analyses the process of young adults leaving the parental home. This has received increased attention from the sociological, economical, geographical and demographic literature, since the recognition of its importance in understanding the life course, the family formation process and intergenerational relationships. The increasing complexity of the living arrangements observed in the second demographic transition is one of the many obstacles to the study of this phenomenon. The present paper attempts to shed light into these intricacies by using Danish register data. The leaving home process is, after all, intimately related to other demographic processes that take place at early adulthood. It intends to observe how factors affecting the decision of young women to leave the parental home interacts differently give three possible destinations taken: living independently, living in a legal union, and living in a consensual union.

Most of the existing research on leaving home in the European context has been concerned with countries where the average age of leaving home is extremely high. Nordic countries, on the other hand, show a much younger age at departure. Despite such intriguing contrasts, there have been very few researches into the leaving home process in the Nordic countries. Furthermore, the leaving home process in Denmark has never before been studies, a feature that makes this study particularly important.

The source of data is the ‘Fertility of Women and Couples Data Set’, which originated basically from the Fertility Data Base (FTDB), which is register based. A group of women, 13-17 and still living with parents in 1981, will be analyzed longitudinally from 1981 to 1994 in order to capture the timing of the transitions mentioned, hence establishing the duration in the parental home until the occurrence of the transition.

Attitudes to cohabitation and marriage among young adults in Sweden

Eva M Bernhardt, Center for Women’s Studies, Stockholm University

Sweden is well-known as the for-runner when it comes to non-marital coresidential relationships. Unmarried cohabitation is a social practice with historical roots in Sweden, but it became really widespread from the end of the 1960s. Nowadays, it is clearly “deviant behavior” to get married without previous cohabitation. Many cohabiting couples marry between the first and the second child. It has been shown that pregnancy and childbirth still tend to trigger a change in marital status. Nevertheless, there is probably growing acceptance of childbearing and childrearing within cohabiting unions. Those who choose cohabitation rather than marriage for the whole life course are, however, a minority even in Sweden. To understand family change and to get a handle on what to expect in the future, it seems crucial to know what are the attitudes to cohabitation and marriage among the generations born in the 1970s who are currently establishing themselves on the labor market and entering the family formation phase.

Attitudes to cohabitation and marriage will be explored in the proposed paper, which will present results from a recently undertaken mail questionnaire survey among young adults in Sweden. The survey, Family and Working Life in the 21st Century, includes 2,300 respondents, 22, 26 and 30 years old, both males and females. The results of the survey show that 85 percent of the men and 87 percent of the women consider it OK to cohabit even if one has children. Thus, acceptance of childbearing within cohabiting unions is very widespread among young adults in Sweden. Most likely, however, it should be interpreted as an expression of tolerance towards how others choose to live their lives, more than an intention regarding their own lives. Among those not yet married (the overwhelming majority in these ages) 21 percent answer “yes” and 34 percent answer “perhaps”, when asked whether they think they will be married in five years. Only 30 percent answer “no” and the remaining 14 percent had no opinion. Clearly, a majority of young adults, not yet married, consider marriage a definite option in their future lives.

This paper will focus on factors influencing attitudes to childbearing within cohabiting unions, running separate (logistic) regressions for men and women, in order to detect possible gender differences, controlling for age, parents’ education, childhood family structure (intact/ non-intact family), and other factors. The effect of current parental status (number, age and coresidence status of possible children), current partnership status (single/LAT-relationship/ cohabitation/ marriage), current employment situation, educational level, gender role attitudes, familism, career orientation, importance of religion, and satisfaction with various aspects of current life situation are among the factors to be included in the analysis. For those not yet married, I will also include expectations about getting married in the next five years.

On the whole, I expect small effects of the explanatory variables. However, men are likely to be less positive to childbearing within cohabiting unions than are women, as are those with higher education, a more traditional gender role attitude, and those who consider religion important in their lives. In future research, I expect to show that a positive attitude to child-bearing within cohabiting unions decreases the likelihood of a future transition to marriage.

Language and mate selection in Finland

Fjalar Finnäs, Institutet för finlandssvensk samhällsforskning, Vasa

In addition to the well-known two-sex problem we discuss other methodological problems connected with studies of the mate selection process. Based on an exceptionally detailed data set we illustrate the necessity to focus on local marriage markets, and the problems connected with this. The empirical study focuses on mate selection in Finland with respect to the language of the spouses, taking several characteristics of both spouses into account. The studies show that among the Swedish-speaking persons the proportion with a Finnish-speaking partner correlates with the language structure of the place of residence. Further it decreases with an increasing level of education and increases with age. Bilingual couples are more common in consensual unions than in formal marriages.

Socio-economic status and divorce in first marriages in Finland

Marika Jalovaara, Population Research Unit, Department of Sociology, University of Helsinki

Various studies report an inverse association between socio-economic status and the risk of marital disruption. Using register-based follow-up data on first marriages in Finland intact at the end of 1990 and divorces in 1991-93 (n=21,309), this study aimed at gaining a better understanding of socio-economic differentials in divorce risk by disentangling the influences of various aspects of the socio-economic status of the spouses. Indicators of socio-economic status include each spouse's education, occupational class, economic activity, and income as well as housing tenure and housing density. When examined individually, divorce risk was inversely associated with socio-economic status for all its various indicators except wife's income.

All of these factors had an independent effect on divorce risk. The effect was, however, weak for the spouses' occupational rankings and housing density, and it was positive for the wife's income. Given the multifaceted nature of these socio-economic differentials, it appears unlikely that one single explanation could account for them all.

Cohabitation: For better or for worse?

Ingvild Hauge and Turid Noack, Statistics Norway

With increasing numbers of cohabiting couples and children growing up with parents who are not married, Norwegian statistics on family formation and family dissolution have encountered new issues. Regular statistics on marriages, separations and divorces do not paint the entire picture anymore.

Based on new figures for Norway the paper will explore family formation and family dissolution, the latter in particular, for cohabiting couples with common children. Preliminary figures indicate various fluctuations in the family relationships of these cohabitants. How many cohabitant relationships end every year, either due to the couple moving apart or due to death? And how many cohabitants with common children choose to marry, and if so, after how long? Survey data has established that cohabitation is considerably less stable than marriage, even when there are children involved in the break-up. But how many children experience this type of family dissolution? And how many see their parents entering a de jure relationship? What trends can we discern? Register data are used to create time series, so that it is possible to track developments over time. The time period will be limited to the last ten years.

The paper will aim at making Nordic comparisons, to the extent that such comparisons are actually possible. Are the statistics available? What story does it tell?

Once a parent – always a parent

Elisabeth Landgren Möller, Statistics Sweden

Most parents have joint custody of their children. It applies to practically all parents married to each other. Parents who are not married to each other can get joint custody by informing the Social Welfare Board either in conjunction with confirming paternity or later. Most non-married parents have joint custody.

Three-quarters of 0-17 year old children in Sweden are living with both their biological parents. One quarter are living either with a lone mother/father or with one biological parent and a step-parent. But 87 percent have parents with joint custody. Young children have cohabitant, non-married parents more often than older children. It applies to 49 percent of one-year old children but only 7 percent among the 17-year old. Sole custody has become unusual. Today 96 percent of one-year old children with cohabitant parents have joint custody. Cohabitant parents often marry after a while. Others separate before or after they have been married. Regardless of marriage, the joint custody of parents is automatically kept after separation. However, either parent can apply for sole custody. That can be allowed by the District Court only if it considers it to be in the best interest of the child.

Having custody is not the same as caring for and living with the child. Most children with separated parents live with one of them. But parents have a shared responsibility to ensure that the child is able to spend time with the parent with whom it is not living.

Dissolved families: A prospective cohort study of family strain before parental separation following children born in Denmark 1973.

Mogens Nygaard Christoffersen, The Danish National Institute of Social Research

Both in Scandinavian and North-American studies it has been estimated that about one third of a birth cohort will experience a divorce during their childhood (Bumpass & Rindfuss, 1979; Christoffersen, 1992, 1993; Jensen, 1991). The study includes all the Danish children born in 1973 who have experienced family separation between the years 1979 and 1991, when the children or adolescents were about 6 to 18 years old (N=24.096). The controls (years at risk) were constructed by the total birth cohort (69.623) who weren’t exposed to family separation during the current period.

Statistics cover several health aspects, education, income, social networks, family violence, parental self-destructive behaviour, parental alcohol or drug abuse, and parental unemployment. By using unemployment statistics from 1980-93, it was possible to follow the extent of the parents’ unemployment (number of days) for individual years. Statistics have also been collected for the same period which year by year cover a number of health aspects (eg. admittance of children and parents to hospital, deaths), family aspects (eg. family breakup, teenage motherhood, and victim of violence, placement of children outside the home), family resources (eg. income, vocational training), self-destructive behaviour (eg. criminal convictions, suicide attempts) and other social circumstances including schooling and vocational training as well as employment of the children when they become adult.

Whether their parents are living together, married or not, the personal identity number was used to link the children and their biological parents to information in 15 registers. Subsequently, the database had been made anonymous by deleting the identity numbers. The study analyses in what way the family situation prior to the family separation differs from the controls. The cohort data were analysed by means of logistic regression to isolate the potential influence from parental constrains.

Theme 2: Fertility trends in the Nordic countries

Co-ordinator: Gunnar Andersson, Max Planck Institute for Demographic Research, Rostock Discussant: Gerda Neyer, Max Planck Institute for Demographic Research, Rostock

Delayed fertility in Denmark: An overview of recent trends

Lisbeth B. Knudsen, Danish Center for Demographic Research, University of Southern Denmark, Odense

[Abstract missing]

Fertility developments in Norway and Sweden since the early 1960s

Gunnar Andersson, Max Planck Institute for Demographic Research, Rostock

The purpose of this paper is to give an overview of recent trends in childbearing in the two neighbouring countries of Sweden and Norway. We make use of indexes produced by applying indirect standardization (by event-history techniques) to register data of the two countries in order to describe and contrast the fertility developments over the last four decades. The indexes enable us to decompose overall fertility trends into birth-order specific components. In addition, by combining the same kind of data from two countries, we get an accurate picture of various differences in levels of fertility between the two countries. We demonstrate how Swedish fertility has shown strong fluctuations during the whole period under study while Norwegian fertility has evolved more gradually over time. A turnaround from decreasing to increasing levels of childbearing is evident in 1977 for both countries while a sudden shift to shorter birth intervals in the 1980s is specific to Sweden and contributed to its more spectacular increase in fertility during that decade.

The first shared birth in second unions in Sweden and Hungary: a gender perspective

Livia Sz. Oláh, Dept. of Sociology, Stockholm University

With the growing prevalence of sequential partnerships among individuals of reproductive ages childbearing occurs in increasingly different family settings in most industrialized countries. A number of studies have shown that fertility decisions in unions where at least one of the partners had a previous relationship are influenced, at least partially, by other factors than it is the case in first partnerships. Especially, children of one or both partners from earlier relationships can affect the new couple’s childbearing decisions. Also, the effect of such children may vary depending on whether or not they become part of the new household. As children usually stay with their mothers after family dissolution, the gender dimension of stepfamily fertility needs to be studied explicitly. In our study we address this issue by focusing on the first shared birth in second unions in Sweden and Hungary. We use the method of hazard regression to analyze data extracted from the Swedish and the Hungarian Fertility and Family Surveys of 1992/93. The results suggest that the union-commitment effect is present in both countries as the risk of having a first shared birth is not significantly different for childless couples as for couples with at most two pre-union children, and it does not matter whose children they are, his or her. We also find some support for the parenthood-status hypothesis in both Sweden and Hungary in terms of higher conception risk in second unions where the woman is a parent but the man is childless as compared to partnerships where both partners have pre-union children. In addition, the findings confirm our expectations that the impact of pre-union children living with the couple is stronger on the first shared conception than that of children living elsewhere.

Trend and Seasonality in Swedish Fertility 1850 to 2000

Per-Gunnar Cassel, Demography Unit, Stockholm University

Monthly data on the number of births are found in Sweden’s official statistics. The data collected here will be for the years 1850 to 2000. The series analysed in this study are the number of births and the crude birth rates. These series exhibit a regular monthly seasonal variation characterized by two peaks: one in March and theother in September. Over the years, however, this pattern has been slightly changing. One aim of this paper is to show, in tables and diagrams, the structure of this change. Another aim is to depict the trend of the selected time series as monthly time series.

The Bureau of the Census in the USA has developed methods, which were later refined at Statistics Canada, whereby a time series showing seasonal variation can be decomposed into a trend, a seasonal variation and an irregular component. The time series department at Statistics Canada has developed a program under the name X11ARIMA that has been used in this study. X11ARIMA meets well our demand for doing statistical analysis. The program is furthermore well known and extensively used by governmental statistical offices all over the world.

The procuration of monthly trend estimates rather than yearly estimates offers some advantages. Not only can we discern ‘within-year’ movements in the trend that would otherwise be obscured, but we can also date the turning points in the trend more precisely.This taken together will facilitate a causal explanation of how the trend moves.

An everlasting question that arises with all short-term statistics is: “Does the values observed for the last month(s) or quarter(s) in the series warrant the conclusion that the trend is turning?” The answer to that question can be given by the use of a regular hypothesis test, which by the way is carried out by X11ARIMA. However, one cannot entirely rely upon such a test. A causal understanding of how the trend moves should consolidate the answer and knowledge of the past will substantiate that understanding. Therefore, the results of this study have a value that goes beyond its purely historical interest.

Trends in childbearing of foreign-born women in Sweden

Gunnar Andersson, Max Planck Institute for Demographic Research, Rostock

In this paper, we present an investigation about recent patterns and trends in childbearing of foreign-born women in Sweden. It is performed in a similar way as a previous study by Andersson (1999) who analyzed child-bearing patterns of Swedish-born women in that country by applying event-history techniques to register data. The present extension to the foreign-born population gives insights about existing differentials in fertility between women born in different countries and we display period trends in fertility by birth order for some aggregated groups of foreign-born women. In addition, we get a picture of the impact of the migration process on the childbearing behavior of migrants; we examine whether the migration results in a disruption in childbearing and whether we find any signs of an adaptation to the fertility levels of the Swedish-born population. Our study is based on longitudinal information on childbearing derived from the Swedish population registers. Our data cover around 510.000 women born abroad (in 1925-1985 coming from more than 200 different countries) who immigrated to Sweden up to 1999 at an age no higher than 35. The Swedish-born population is included as a reference group and we present relative risks of first, second, and third births for women stemming from the various countries.

Theme 3. Long term perspectives on family and women's work during the 1900s

Co-ordinator: Maria Stanfors, Lund University Discussants: Katarina Katz, Dept of Economics, Stockholm university, Lars Svensson, Dept of Economic history, Lund university

The Role of Education on Postponement of Maternity in Britain, Germany, the Netherlands and Sweden

Siv Gustafsson, Dept of Economics, University of Amsterdam

This paper analyses the effects of education on postponement of maternity, by distinguishing the level of education and time being in school controlling for women’s birth cohorts. In this paper we separate between level of education completed and time of leaving full-time education as determinants in duration analyses to explain timing of first birth. There are three hypothetical outcomes of the timing of maternity according to educational levels. In the first outcome, all women, irrespective of education level acquired, take a similar time period after completing school until giving first birth. The second outcome is that high educated women wait longer after completing school until having their first child and the third outcome is that high educated women give birth quicker after finishing school

The empirical analysis has been carried out using household panel data from four countries, namely Britian (BHPS), Germany (GSOEP), The Netherlands (OSA) and Sweden (HUS). We use the cross sections of a recent wave of the respective household panels and connect the information on birth history and education history which had been collected earlier in the panel data period, to the date of the cross section. We follow all women from age 15 until they either have a first birth or are censored at the latest cross-section not yet having given birth to a child. We construct variables on durations of schooling, time at completing full-time education, duration of time between ending full-time education and time at first birth.

Impact of educational attainment on the timing of births and number of children. The case of Denmark

Lisbeth B. Knudsen, The Danish Center for Demographic Research, University of Southern Denmark, Odense

Previous analyses in Denmark has revealed the existence of a strong relationship between a woman’s educational attainment and the timing of the first birth and the number of children she has throughout her reproductive age span. A similar relationship has been found regarding the occupational position, as women in higher, demanding positions delay their childbearing and have less children, on average, before a given age. For men, the relationship is somewhat different, as men in higher positions do not limit their number of children to the same degree as women in similar positions, which indicates that the woman’s possibilities to combine work and family is the most important when it comes to the preferred family size.

This analysis focus on the impact of educational attainment on childbearing, based on register data on the total female population in fertile age and their current partners in Denmark from 1980-1994. The women’s educational attainment will be analysed in relation to their timing of first birth, the spacing and the total number of children. The number of children of their current partners, if any, will be included as a variable in the analysis, as well as their educational and occupational position. Moreover, findings from previous research regarding the period before 1980 will be discussed.

Obstacle or Prerequisite? The Role of Economic and Social Conditions for the Level of Swedish Birth Rates 1965-2000.

Åsa Löfström, Dept of Economics, Umeå university

The methodology used in this study is time series analysis. We will analyse different aspects of economic life, on macro- and micro level, which may be of crucial interest in explaining the ups and downs in Swedish birth rates during the last 25-30 years. The dependent variables in the models we will estimate are number of births and total fertility rates (TFR) respectively. The independent variables are sorted into three different groups (at least): A) Economic indicators e.g. economic growth, labour market situation (unemployment and employment) etc for different years. B) The frequency of social reforms and the expansion of existing reforms directed towards families or children during the period (child allowances, day care facilities for working parents, parental leave etc.). C) Changes in personal ntury

Hettie Pott-Buter, Dept of Economics, University of Amsterdam

Historical international comparison of long-term trends in female labour, family and fertility is largely unexplored territory. This paper contributes to this territory by comparing these trends in the Netherlands with those in the Denmark, Germany, Norway and Sweden. During the twentieth century in all countries comparable important changes have taken place in labour force participation and in the division of caring and breadwinning responsibilities within the family. In general government measures and statutes reflect the changes and socio-economic developments. This paper reveals the different policies that were introduced in the mentioned countries to promote, to support or to prevent the division of labour between women and men. Some government measures are discussed in more detail, such as protective and restrictive labour laws for married women, income tax measures and social security policies. The drastic change in the 1980s and 1990s in Dutch policies coincides with a rapid increase in labour force participation of (married) women. From extreme low levels of female labour force participation around 1900, Dutch figures are now comparable with the Nordic countries.

Between work and welfare: Labour force transitions among single parents in a changing labour market.

Marit Rønssen, Division for Social and Demographic Research, Statistics Norway

In Norway, as in most other Western European countries, there has been a steady rise in female employment during the last decades. A striking feature since the beginning of the 1980s has been the rapid increase in the labour market participation of mothers. Today, about four out of five married or cohabiting mothers with children under the age of 16 are employed. This well exceeds the overall employment rate of women and also exceeds the overall employment rate of men.

In a recent analysis (Kjeldstad and Rønsen 2000) we find that the growth in maternal employment has not affected single mothers in the same manner. While in 1980 about 60 per cent of single as well as married and cohabiting mothers were employed, at the end of the 1990s the rate in the former group had lagged greatly behind, being still no higher than about 65 per cent. When comparing single and married or cohabiting fathers we further find that being a lone parent also reduces employment activity among men.

The lower employment rate among single parents is common for all Nordic countries, but is not the general picture for Europe as a whole. In e.g. Belgium, France, Italy and Austria single mothers are more likely to be employed than married mothers are. Many see the particular Nordic pattern as an expression of a well-established social security system and quite generous economic support to single parents. This is especially true of Norway where single mothers and fathers may receive a special social security transitional allowance ("overgangsstønad") if they are not able to support themselves.

In the 1990s the so-called work line has become more prominent in Norwegian social policy. Incentives to encourage employment activity have been directed also at single parents, in the beginning of the decade by making part-time work more economically attractive, and towards the end by reducing the maximum period for transitional allowance. At the same time, the 1990s were a period of great fluctuations in the business cycle, the national unemployment rate reaching a top level of 6 per cent in 1993, and falling to 3,2 per cent at the end of the decade.

So far our cross-sectional analyses indicate that single parents experience greater problems on the labour market during spells of high unemployment than married and cohabiting parents do. Hence adverse macro-economic conditions may have swamped the policy efforts to encourage labour force participation. In the proposed paper we shall investigate this topic further, analysing individual transitions in the labour market during the 1990s. As a contrast, single parents will be compared to married or cohabiting parents, and gender differences will be highlighted by comparing single mothers and single fathers.

The analyses will be based on data extracted from a large data base that has recently been developed at Statistics Norway (FD-Trygd). The data base contains information from several administrative registers (population, employment, unemployment, social security, income etc.) and the data are prepared especially for life-cycle analyses. Presently the base covers the period 1992-1997/98 which will determine our analysis period.

Gender structures, breadwinner models, and demographic change in a generational perspective: Approaching the second phase at the turn of the millennium?

Kari Skrede, Division for Social and Demographic Research, Statistics Norway

The paper surveys the development of the process towards gender equality in Norway, with particular focus on the structural changes that have taken place in the younger generations where the opportunity structures have facilitated more equal gender roles, both with respect to shared breadwinner responsibilities and with respect to family tasks and roles. A central issue in the paper is whether the process towards increased gender equality has lead to converging lifestyles or to increased social dispersion.

Women’s work and the family from the 1970’s in Finland and Norway. What did family policies do?

Tapio Rissanen, Work Research Centre/Research Institute for Social Sciences, University of Tampere, Christin Knudsen, Norwegian Social Research, Oslo

Finland and Norway are two countries with similar family policies and comparable total labor force participation of women. The development of both women’s employment and policy measures have, however, historically been embedded in very different economic and cultural settings as well as political climates, and despite current similarities of total numbers, the underlying picture is manifold.

Women’s high labor force participation has a long history in Finland. Women were an important source of labor in the agriculture and forestry all the way up to the 1950’s. The unit for both production and reproduction were the households or the farms. Women took also part in the growing manufacturing and service sector of the economy, female workers made up as much as 41 percent of this workforce during the 1920 – 1950’s.

In Norway the industrial revolution took a much more dominant place in the late nineteenth century. Already from the last part of the nineteenth century and up to the mid-twentieth century did the majority of unwed women work for wages. But in this country it was a distinct difference between this group of women and those that were married. Married women did not work unless they had to or really wanted to (2-4 % did).

Common for both countries from the 1960’s was the growth of the public sector including the welfare sector. It is said that the seventies was the decade when married women started to enter the labor force in Norway, however, usually when the children had reached a certain age. In Finland, the seventies was the decade when the mothers of young children entered the labor force. In Norway this took place to a large scale in the eighties. The big difference was that Finnish mothers worked full time while part time was common in Norway. In the mid- to late eighties nearly four out of five Finnish mothers were in the labor force. By 1998 four out of five women with young children were employed in Norway.

Finland and Norway are the only countries to have introduced a Cash-for-Care for parents with children one and two years that do not make use of a public subsidized day care. Finland did so in the late eighties while Norway did so a decade later.

This project will present the development of women’s labor force participation during the 1900’s in Finland and Norway, with emphasis on the pattern starting from the 1970’s. We will look at this development in a broader perspective, where the countries’ policy making and cultural norms and values will be considered and related to the changing employment patterns of women and in particular mothers. We will also relate fertility patterns to the changing family situations and consider the evolving gender issues in the two countries.

Theme 4a: The History of Population Statistics

Co-ordinator: Gunnar Thorvaldsen, University of Tromsø

The Development of some Census Variables

Einar Lie, Statistics Norway

I flere land har statistikkens inndelinger av befolkningen etter sosio-økonomisk gruppering vært studert relativt inngående, særlig gjelder dette Frankrike og Storbritannia. Disse studiene viser at prinsippene og motivene bak kategoriseringene har endret seg sterkt over tid. Bak grupperingene finner man ulike hensyn som ønsket om å få frem et bilde at befolkningens fordeling etter rang og status, etter hvor de hører hjemme ut fra yrkesmessig kompetanse, eller å få kategorisert befolkningen med tanke på å drive sosiale eller demografiske undersøkelser.

I Norge har kriteriene for inndelingen etter det mani dag kaller sosio-økonomisk gruppering, vært et lite påaktet forskningsfelt. Det finnes ingen systematiske historiske studier av dette. Hensikten med dette bidraget vil være å gi et riss at hva slags hensyn som lå bak inndelingene i tellingene i det 19. århundret. Litt fortettet kan man si at inndelingskriteriene gikk fra å skulle si noe om hvem folk var, til hva de drev med. Tellingene fra til midten av århundret gir et fortettet bilde av det Jens Arup Seip kalte "samfunnets stenderske ramme". Med de nominelle tellingene (1865) og dypere endringer i politikk og sosial struktur, forandret dette seg. De norske folketellingene fikk sammenlignet med for eksempel Sverige og Danmark en usedvanlig detaljert livstillingsgruppering. Samtidig var deler av den økonomiske statistikken, særlig opplysninger om produksjonsmengder og -verdier, i flere henseender relativt svakere utbygd. Dette gjorde livsstillingsstatistikken til et kjernepunkt i den statistiske kunnskapen om den økonomiske utviklingen i Norge.

Et hovedpunkt i dette bidraget vil være det som kan omtales som "Kiærs system" (etter A.N. Kiær, leder for Det statistiske kontor/Det statistiske sentralbyreau fra 1867 til 1913) i den sosio-økonomiske inndelingen - hvilke prinsipper som lå bak dette systemet, og hva slags samfunnsbilde det var med å skape. Perioden med numeriske tellinger behandles mer kursorisk."

The early Production of Statistics in Sweden

Peter Skiöld, University of Umeå

Two aspects of population statistics in Sweden have been stressed as important in a European context; time and quality. The establishment of the Table Commission (Tabellverket) in 1749 gave Sweden an early start. There are reasons to claim that Sweden was unique in the field of population statistics, at least during the eighteenth and early nineteenth centuries.

Why was it then Sweden that played a leading role? How was it possible for the country to fulfil the ambitious plan for national population statistics? And why were these initiatives taken here in the first place? Which were the motives behind the actions? To answer these question we need to view Sweden in a wide and long-term perspective. Several factors interplay and are influential during the pre-history of Tabellverket, but also for the maintaining of the establishment. These factors are not separated from each other by sharp borders, instead the factors involved are strongly related.

The present paper discusses the role of taxation records and administrative books, the national Lutheran church and the clergy, population development, political changes, the official administrative system, the scientific context, the importance of a strong state and an accommodating people, and strong personalities in networks.

The Population Issue and the Censuses in early 20th Century Norway

Kjartan Soltvedt, Statistics Norway

"Demographic transition" is generally regarded as one of the key concepts in the discipline of demography, if not the key concept. However, as has been argued repeatedly in later years, it is by no means an unproblematic concept. This paper will focus on one of the problematic notions embedded in the concept, the notion that the fertility transition can be seen as a uniform process, by and large a corollary of economic development and modernisation. Furthermore, the paper will examine the relationships between population statistics and population policy by examining the inter-war population debate in Norway.

In a seminal analysis of the fertility transition in England and Wales and its conceptualisation in early British population statistics, Szreter (1996) has argued that the popularity of the concept of demographic transition stems largely from the model of social structure it was premised on. Szreter identifies the origins of what he calls the "professional model of society" to the first scientific analysis of the fall in fertility, the report on the fertility of marriage from the 1911 census of England and Wales. According to Szreter, the British enumerators chose to interpret the fertility decline as an effect of the deliberate birth regulation. Furthermore, it was emphasised that birth control was a social practice instigated by the upper echelons of society and then gradually adopted by other sections of the population. By questioning this interpretation, Szreter also offers an alternative description of the fertility transition. By deconstructing the original classification of occupations, he concludes that there existed multiple sets of fertility regimes. Thus, the original construct of the fertility transition as a process where diffusion of contraceptive practises gradually was embraced by different sections of the population appears to be a product of how the statisticians divided the population into different categories. As an alternative interpretation, Szreter argues that changes in fertility must be seen as consequences of shifts in the perceived cost of childbearing.

The first Norwegian analysis of the decline in fertility was conducted as a part of the 1920 census. By and large, the newly appointed head of the Norwegian Statistical Institute, Gunnar Jahn, relied heavily on the British approach in designing the fertility analysis used in the 1920 census. The critique the demographic research in Britain has been subject to seems therefore also relevant when discussing the early Norwegian analyses of the fertility transition. However, it should be stressed that whereas in Britain there is continuity with regard to demographic research from 1911 onwards, in Norway a continued tradition from the work of Jahn and into the post-war era is less pronounced. The impact of the intellectual constructs of the early fertility analyses on more recent demographic research has therefore been of a lesser extent than what came to be the case in Britain.

But whereas the impact on later Norwegian demographic research by these early attempts to grasp the complexities of the fertility decline has perhaps been negligible, the same cannot be said with regard to social policy. In the 1930ies fears of Norway's imminent depopulation became pronounced, and had a heavy impact on the emerging debates on social reforms and the design of the future social democratic welfare state. Although few openly prenatalist policy reforms resulted from this debate, the impact of the depopulationist debate of the 30ies should not be understated. Much of what was only discussed in inter-war years came to be implemented in the post-war era. Although the "baby-boom" of the immediate post-war years effectively quieted fears of depopulation, the social reforms, which were introduced in this period, had its origins in the discussions of the inter-war era, and still more important, were premised on the then prevailing perceptions of society and social classification.

Thus, the prevailing perceptions of social differentiation in the inter-war debates on social policy emerge as relevant topics for discussion. To what extent were the perceptions of social differentiation, which permeated the debates of the 1930ies, embedded in the new in the welfare institutions emerging in the aftermath of World War II?

Theme 4b: Historical Population Estimates

Co-ordinator: Gunnar Thorvaldsen, University of Tromsø

The Population of Sweden 1571-1751

Lennart A. Palm, Dept of History, University of Gothenburg

The famous Swedish official statistics that started in 1749 have only been published on provincial (län) or national level up to 1805, probably due to big lacunae in its local material. In a forthcoming book (The Population of Sweden. Population size in its parishes and communes 1571-1997. With special attention to the period 1571-1751) I have tried to overcome the lack of earlier information on parish level by publishing official parish figures for 1751 and 1780, estimating the size of parishes where the statistics have been lost. For the years 1571, 1620, 1699, 1718, 1735 I have made calculations, also on parish level, using different tax lists that enumerate the households. A lot of effort has been spent on making the household figures complete. The same labor has been directed towards finding reliable regional mean household sizes (MHS) to multiply the household figures by. To enhance studies on very long periods for all the Swedish territory of today former Danish and Norwegian parts of the country have been included (all parts included referred to as Sweden below).

The population totals for the period 1571-1699 show an astonishing average growth of c. 0,6 percent per year. This made it necessary to investigate if this is really compatible with an assumed comparatively high mortality in that period. Tax material from c. 1620 was critically scrutinized and showed, albeit its incompleteness in some respects, evident traces of what J Hajnal called an East European system: age at marriage about 21 years, only 30 % unmarried women (of whom one fifth were widows) above 15 years of age, crude birth rates of 44 or more per 1000, and a lifetime marital fertility of seven to nine children per women.

As much of the source material divides the population according to age only in two groups - above and under 15 years - a method was developed for calculating maximum age of marriage using Hajnals formula (Hajnal 1953 p 129) for computation of singulate mean age at marriage (SMAM) on the assumption that the age pyramide of 1750 for Sweden could be used and that the married women were the oldest and filled the top of the pyramide, the unmarried the bottom and that all women had married at 50. The last two of these assumptions make SMAM as high as possible. At that point the assumption of the age pyramide was questioned and it was argued that using the 1750 age pyramide was only prudent - probably the 1620 pyramide had been somewhat flatter (which would allow for a still lower maximum). Maximum age at marriage for 125 000 Swedish women c. 1620 calculated in the prudent way was 24,4 years, compared with statistical age at first marriage 1861/1880 or 27,1. Simulations on later and more complete data on age and marriage status suggested that maximum age tended to be five years or so over statistical age.

Fertility c1620 was computed using a formula suggested av Henry & Pilatti Balhana (1975) but on the conditions that we do not know the proportion of married women of 30 years of age, only how many women in total over 15 years of age that were married. Fertility was then computed on different assumptions about the proportion women age 30 and the proportion women age 30 that were married. The most plausible proportions gave 7,1-8,8 children per woman. The general picture of the demographic system in Sweden c. 1620 was corroborated by evidence also from one parish in northern Sweden with unique sources from 1559 and from Denmark and Norway in the middle of the 17th century, and some contemporary direct evidence from 1621 and 1639 on peasant girls beeing married as young as 12 to 15. The resulting population totals are tens of percents lower than established handbook computations. However the latter figures were not founded on contemporary information on MHS and in fact based on the conviction that annual average growth rates in the 16th and 17th centuries did not surpass 0,1-0,2 %. In my article I show that much higher rates are well allowed for by an ”East European” demographic system in those days prevailing in Sweden. That system seems to have been succeeded by a more Western system at some time around 1700.

Population and Vital Statistics by Age, Sex, and Marital Status over 200 years: Revised and Extended Estimates for Norway 1801-2001

Kåre Bævre, Dept of Economics, University of Oslo

Norwegian population statistics of the 19th century is generally held to be of a quality unique to the period. As in the other Scandinavian countries this opens for unique opportunities for quantitative studies of pre-industrial and early industrial demographic history. The work documented in this paper aims at facilitating such work by establishing a comprehensive database of national population statistics for the two hundred year period 1801-2001. There are three main improvements upon earlier work and the original published statistics. First, annual estimates of the national population by age and sex is extended back to 1801 (previous estimates ending in 1846). Second, estimates are made also with respect to the distribution on marital status by age and sex for the same time period. Third, the old estimates of population by age and sex in the period 1846-1930 are improved upon by employing modern statistical techniques and new sources of information. In particular information from primary sources such as parish registers and emigration protocols, as well as previous studies of this material, is used were possible to remedy the lack of information in the published statistics. As the documentation and evaluation of the original statistics is spread on a large number of publications, a comprehensive treatment also facilitates accessibility. This is important, since the reliability and amount of information available differs substantially over the 200-year period.

Danish Population Estimates 1665-1840

Hans Chr. Johansen, Danish Center for Demographic Research, University of Southern Denmark, Odense

In an early phase of the interest in historical demography most of the efforts were concentrated to family reconstitution studies using parish register information at the individual/micro level and because of the laborious work in collecting the data most studies were limited to one or a few parishes. In later studies attempts have been made to use early nation wide or regional - often imperfect - statistics or to create such statistics from other types of sources in order to avoid the time-consuming collection of individual data and in order to achieve results which hopefully will be more representative for larger areas. The paper will be an attempt at evaluating the quality of the Danish early population statistics for a macro analysis and at constructing population time series and some demographic measures for the period from 1665 to 1840. The start in 1665 has been chosen because the Danish population suffered great losses from epidemics and war during the 1640s and 1650s. The losses hit the various parts of the kingdom very differently and macro studies based on incomplete data are therefore much more risky to undertake during this period than in later years. The analysis stops in 1840 because after that year there exists detailed census statistics with short intervals and also yearly detailed statistics of births, marriages and deaths.

Theme 5: National systems of population statistics in the Nordic countries

Co-ordinator: Kåre Vassenden, Statistics Norway, Kongsvinger

Introduction: The Nordic model

Kåre Vassenden, Statistics Norway

[Abstract missing]

Statistical system of population changes in Finland

Matti Saari, Statistics Finland

Statistics Finland has received the structural data on population in electronic format since 1976 and the basic data on population changes since 1975 from the Population Register Centre. The latest system renovation in the production of these data took place in connection with the production of the data for 1999. Production of the basic data for structural data on population has been started on 1 February the next year. The basic structural data on population forms the final data for the modifications to be made with the PL/1 program. The data for 2000 was complete in mid-March 2001. The structural data on population are used to convert files with the super-star program into PC-Axis files from which tables are produced for publications and information service. Tables are also made of the structural data with the TASSU program. The family file, which also contains data on cohabiting couples, is also produced annually from the structural data on population.

The basic data on population changes are transferred at Statistics Finland to the population change database (VÄMTIK) and new information, such as the population change code, is deduced. VÄMTIK is used with IDEAL program. Aggregated data from the structural data on population are also taken to the VÄMTIK system at the beginning of the year on top of old data, which are then updated. Each mid-month statistics on population and population changes are produced for the previous month. Only when the annual statistics are compiled the information on the correction forms is taken into account. Next the correctness of different items of population change data concerning the same person are verified by taking into consideration the dates of reporting and event. Then a sequential file on a certain type of population change is produced from the VÄMTIK system, at the same time deleting any inconsistencies between the information on the events. The sequential file is checked and any errors detected are corrected and then tables are printed with the TASSU program. The data are also formed into PC-Axis files from which the majority of the tables in the publications can be printed. Because of the revisions, the files for 1999 were not completed until September 2000.

A new Total Population Register at Statistics Sweden: Better quality and more possibilities

Anna Wilén and Ingvar Johannesson, Statistics Sweden

The Total Population Register (TPR) of Statistics Sweden contains data that are collected from the Swedish Population Registration System. Every day Statistics Sweden receives information on changes of the registered population. The TPR is the basis for all official Population Statistics and most samples of surveys are chosen from the TPR. Furthermore, the TPR supply information to other types of statistics and it is largely used in commissioned works. The most used items of the TPR are; personal identity number, name, address, place of residence, sex, age, marital status, citizenship, country of birth and relations between adults and between children and adults. A new TPR-system came into use at the end of 1998. Then a number of new items were included in the system, such as dates of all events and relations between persons.

In connection to the new system, some activities have been carried through in order to improve the quality of data. A new routine for re-reporting of false data to the Population Registration Authorities is used and some historical information from the old TPR-system has been inserted into the new system. There has also been a changeover from a weekly to a daily collection of data. Some new registers have been created; a Multi-Generation Register, a register that contains all changes of personal identity numbers and a register containing the total stock of personal identity numbers that have existed over the years. Some on-going quality improving activities include the integration of historical data and improvements of the Multi-Generation Register.

The present Norwegian population statistics system: An overview

Kåre Vassenden, Statistics Norway

[Abstract missing]

Theme 6. Population forecasts

Co-ordinator: Nico Keilman, Dept of Economics, Unived world demography

Co-ordinators: Johanne Sundby, University of Oslo and Veijo Notkola, Dept of Sociology, University of Helsinki

AIDS in Namibian society

Veijo Notkola, University of Helsinki

Sub-Saharan Africa is the only region of the world that has developed a generalized, self-sustaining, heterosexually transmitted AIDS epidemic. According to UNAIDS estnt in Namibia and then South Africa but the rapid spread of HIV. Southern Africa now contains most of the countries hardest hit by AIDS.

Documentation of the process of demographic transition in sub-Saharan Africa is partial in its coverage, of limited historical depth, and lacking in temporal detail. Little is known about the extent of fertility increase and mortality decline during the colonial period, about pattern in a population for which mortality data exist already for the period up to 1990 and which is known to have experienced an extremely rapid rise in HIV infection (and therefore mortality) during the 1990s. It is also possible to analyse interactions between fertility and mortality and to compare the mortality data for Ovamboland with the mortality of other Southern African populations. In addition, little information exists about cultural factors connected to AIDS epidemic in Namibia and how campaigning for the prevention of the spreading of HIV/AIDS has been carried out in Namibia.

The main aims of the project are: A) to measure the rise in mortality resulting from the AIDS epidemic in Ovamboland by comparing data for the 1990s with those for 1965-1989; B) to compare the pattern of mortality change in Ovamboland with that revealed by the other recent mortality data that exist for Southern Africa; C) to measure the changes in fertility associated with the HIV epidemic in Namibia; D) based on these results, to develop appropriate mortality and projection models for contemporary Africa; E) to analyse cultural behaviour and the spread of AIDS in Ovamboland; F) to analyse the role of the media in reporting of the HIV/AIDS pandemic and in campaigning for the prevention of the spread of HIV.

Male contraception prevalence and factors associated with contraceptives among men aged 15-59 years in Ngara -district, Tanzania,: 2000

Fabian N. Ndenzako, University of Oslo

This study examined the contraception prevalence and factors associated with contraception use in the district. The factors included education, knowledge and attitude of men about male contraception, availability, religion, and fertility preferences. Data were obtained from the representative sample of 275 men aged 15-59 years from 18 villages in the district general population. Majority of men had completed standard seven 66% (182), and 82.5% were married. Men started their first sexual intercourse at late teen's years and get married at early 20's. Men contraception prevalence was found to be 19%, which was slightly lower compared to the National male contraceptive prevalence (22%). Only 10% had used contraception in their last sexual intercourse, majority mentioning to have used abstinence (9.5%), followed by condoms (7.3%) then injections (Depo-Provera) 5.8% to their wives and non had used vasectomy as a contraceptive method. 9.5% of men had regular sexual partners and only 7.6% had used condoms in last sexual intercourse with their regular partners. Knowledge on contraceptives was high, 96% (264) of men knew condom as a contraceptive method, followed by abstinence 70% (194), then coitus intruptus 51% (140) and vasectomy was lastly to be known 48%(131).

Despite the high knowledge about condoms as contraceptive method, it was less used compared to abstinence due to false believe and negative attitudes men had about it. Almost twice as men (85%) knew female sterilisation, compared to vasectomy (48%). Majority of men wanted average of 4 children (20.7) only 1.8% wanted as many as 20 children. More men (45.1%) preferred boys as compared to (9.5%) who wanted girls, majority mentioning that boys were important for inheritance of family properties, (27.3%) as well as helping them when they become old (11.3%). Among those not using contracepives, 21.2% reported that they want more children, 16.4% reported difficulties in use of male methods and 5% reporting poor availability of the methods and 2.6% didn't know the methods. 60% men, whose wives were not using any contraception, were willing to allow and support their wives to use contraceptives if they were given adequate information. However, when men were asked whom do they believe is suppose to start contraception use discussion in a family, (67.5%) reported that husband should start discussion (24.4 %) believing that, they are head of families and 18.2% reporting to be the ones who have to take care of children. 12.4% reported wife has to start discussion about contraception use, equal number of men reporting that wife is the one who become pregnancy (3.6%) and felling pain during delivery.

85% knew places where contraceptives can be obtained, hospital, clinics and dispensaries being frequently mentioned. Most men had seen condoms and majority (73%) believe that condom could prevent sexually transmitted diseases, suprising 20% didn't know that and 1.1% didn't believe it. 45% of men agreed that, condoms can prevent HIV transmission, and 41% they didn't agree and 14% didn’t know that at all. 22.2% didn't know how many times can a condom used per sexual intercourse and 1.1% believing that a condom can be used more than once for sexual intercourse.

Contraceptive use among lactating women in Botswana

Edwin M. Itshekeng, University of Oslo

Breastfeeding is still very prevalent in developing countries. This practice operates as a natural but limited means for preventing even higher fertility in these countries. Conversely, use of modern contraceptives by women in developing countries has also increased in recent years. The immediate policy relevance of an understanding of breastfeeding behavior and contraceptive use lies in the fact that both factors are the most influential intermediate fertility determinants. This paper examines breastfeeding women's use of modern contraceptives in Botswana. It will provide information on the prevalence of breastfeeding and contraceptive use among childbearing women. An association of these practices with selected demographic and socio-economic variables of women who were currently breastfeeding will also be shown. The data comes from the fertility and mortality survey carried out in 1993 by the department of Demography (University of Botswana). The survey was intended as a national survey thus necessitating designing a nationally representative sample of Botswana. The procedure used in selecting the sample was multistage cluster sampling, with probability proportional to size in order to obtain a self-weighting sample at the national level. The ultimate sampling units were households but other units (clusters) were used at the primary and secondary stages. The individual questionnaire, in particular contained a variety of questions on a woman's pregnancy and birth history, her educational and employment background, contraceptive knowledge and use and her breastfeeding history. The results showed no striking urban-rural differences (as expected) in contraceptive usage among lactating women who still breastfed their last child. It was also revealed that breastfeeding women with high parity (more than 5 children), who were usually older (35-39 years) were less likely to use contraceptives for birth spacing but instead adopted prolonged breastfeeding. There was also no significant association of education with contraceptive usage among currently breastfeeding women, although this association tended to be very high ( among non-breastfeeding women.

Maternal health in Mozambique: Improving maternity services

Johanne Sundby, Momade Bay Usta, Emmanuel Rwamushaija, Univ. of Oslo/ Jose Macamo Hospital, Maputo

UNFPA and NORAD has invested in a reproductive health program in Maputo City, Mozambique. One component of the project was to upgrade one maternity to become a fully finctioning referral maternity hospital. Until this project started, 1999, there was only one obstetric hospital that could provide 24 hour obstetric emergency services in the city of Maputo, with its 1 million inhbitants. This hospital is severely overcrowded and admits far too many uncomplicated deliveries that bypass the peripheral referral lines. The intention of the project was to provide better maternal services for the peripheral parts of Maputo.

The project invested in obstetric equipment like a sonograph, refurbished the theatre, opened an outpatient surgery for minor gynecological surgeries, rnovated the wards, trained the staff, employed some more trainded doctors, and imporved the overall quality of care by starting regular staff rounds, producing guidelines for treatment and structuring services. An operations research activity follow the project, and they have worked a lot on the statistics.

After the launch of the 24 hour service (April, 2000) there has been a steady increase in number of admissions to the maternity, now to about three times the number. Each month, more than 1000 deliveries take place. There has also been a change in the referral lines. Now many of the peripheral maternities in and outside the catchment area refer to this upgraded hospital rather than the central maternity. The number of caesarian deliveries have increased from one or two elective cs. per week, to almost 12 % of all women being delivered by cs. The absolute number of cs. in the central hospital has decreased while the percentage has gone up due to a shift away from normal deliveries there. The number of complicated cases treated at the upgraded hospital has steadily increased, and the burden of the current staff is high. Staff structure and cost has to be considered carefully in the future.

Theme 8. International migration

Co-ordinator: Lars Østby, Statistics Norway

Introduction: Scandinavian migration in a comparative perspective

Lars Østby, Statistics Norway

The organiser of the session intends to start the session with a presentation of Scandinavian migration in a comparative perspective, as a backdrop for the formal papers to be presented. There will be made some comparisons based on as recent figures as possible (Council of Europe, Eurostat, John Salts EMIN), concerning migration flows as well as stocks. Comparisons will be made between the Scandinavian countries, and the European perspective will also be shortly introduced.

Rethinking migration theory in the light of restrictive immigration policies

Jørgen Carling, Centre for Development and the Environment, University of Oslo

In this paper, I propose a framework for the study of contemporary migration which can help explain how migration flows are shaped by the interplay of migration aspirations and migration policies. Preliminary findings from research on emigration from Cape Verde are presented as an empirical application of the framework.

Migration theory and empirical investigation of migration have usually been based on a dichotomous distinction between migrants and non-migrants: two groups separated by the demographic event of migration. The migrants are those who have concluded in favour of migration in a decision-making process, and subsequently migrated. However, restrictive immigration policies call for a rethinking of this model. In many traditional sending countries of migration to Europe, large numbers of people wish to emigrate, but are unable to do so as a result of immigration policies in receiving countries. These people can be called potential migrants, apart from the non-migrants, who have decided that they wish to stay.

The framework suggested in this paper distinguishes between non-migrants, potential migrants and migrants. First, the transition from non-migrant to potential migrant is a question of aspiration to migrate. Second, the transition from potential migrant to migrant is a matter of individual ability to realize this aspiration. In this paper, both aspiration and ability to migrate are discussed theoretically and applied to the empirical context of Cape Verdean migration to Europe.

International migration and return migration of Swedish engineers before 1940

Per-Olof Grönberg, Department of Historical Studies, Umeå University

The purpose of this paper is to examine international migration and return migration of Swedish engineers, 1890-1940. Though a small group within the large emigration context, engineers played an important role in the industrialisation of Sweden and therefore they are an interesting group to study within the scope of emigration history and historical demography. From the source material, consisting of directories from six technical schools in Sweden, patterns of international migration and return migration and the are revealed. One hypothesis is that North America was the most important destination for the engineers. In this paper, the transatlantic migration streams of Swedish engineers will be compared to international migration and return migration of engineers within Europe, the Nordic countries and the rest of the World. The connections between the engineers’ geographic origin, their education and different international destinations will be discussed with regard to out-migration as well as return migration. The importance of return migration from different countries on different branches and regions in Sweden will be discussed as well as the importance of international experiences on the careers of the engineers after returning to Sweden. The patterns of the engineers’ international migration and return migration will be compared to general patterns in these fields. The occupational careers of the engineers after return to Sweden will be discussed in the light of a comparison with the occupational careers for a group of Swedish engineers who never went abroad.

Labour market integration in the Baltic Sea Region- history, current trends and potential: A study of international labour migration between the regions around the Baltic Sea including all four Nordic countries.

Lars Olof Persson and Jörg Neubauer, Nordregio, Stockholm

A study of international labour migration between the regions around the Baltic Sea including all four Nordic countries.

The Nordic countries established a common labour market in the mid 1950s. This resulted in comparatively intensive labour migration between Finland and Sweden during the 60s, whereafter migration has decreased. From time to time particularly bottleneck problems at one Nordic nation’s labour market have been solved by temporary immigration from another Nordic country. Since the fall of the Iron Curtain, immigration of labour force from former socialist countries has generally increased. This holds also true for the Nordic labour market where a share of immigrants stem from countries around the Baltic Sea. After Finland’s and Sweden’s entrance to the single European labour market, there are few signs of again changing migration pattern. With the foreseeable enlargement of the European Union within the next ten years, there is an intensive debate on the size and character of potential migration, particularly from East to West. Accordingly, the emerging integration of the international labour market around the Baltic Sea is an issue of high policy relevance.

The final purpose of the project is to discuss the emergence and the potential characteristics of a common Baltic Sea Region labour market. In this paper we will quantify and analyse the composition of labour migration in the Baltic Sea Region as well as estimate the potential size and orientation of future international migration in the area. The findings will be supported and illustrated by the use of official statistics.

Theme 9. The immigrant population

Co-ordinators: Anita Lange, Statistics Denmark and Svein Blom, Statistics Norway

Patterns of Mixed Marriage in Denmark: 1977/99

Vera Botelho, Dept. of Statistics and Demography, Odense University Isabella Carneiro, Danish Center for Demographic Research, Odense University Lisbeth B. Knudsen, Danish Center for Demographic Research, Odense University

This paper examines the extent of mixed marriage in Denmark and its implications for the Danish socio demographic structure. The research focuses on two main issues. The first is how much and with whom, do Danish intermarry. Rates of out-marriage according to sex, age and nationality of the partners, are some of the measures considered in this analysis. The second issue is regarding the relationship between mixed marriage and the immigration process. We attempt to observe how mixed marriage patterns in Denmark are influenced by the ‘exposure’ to immigrants. Empirical grounds for the analyses are provided by data retrieved from Denmark Statistics. Preliminary findings suggest that rates of intermarriage among Danish have been increasing during the last two decades.

Vilka nationella invandrargrupper är störst var någonstans i Norden? Regionalgeografiska tendenser.

Eric De Geer, Uppsala University

Har människornas rumsliga preferenser samt statsmakternas agerande resulterat i slumpmässig fördelningar av de nationella invandrargrupperna? Är den största gruppen störst overallt? Avvikelser från detta "normala" studeras här nu på olika rumsliga nivåer.

Vid den grafiska presentationen av empirin liksom vid bearbetningen och analysen används främst kartografisk teknik och metodik. Datamaterialet har erhållits dels från officiell statistik och dels från datakörningar på respektive länders statistiska centralbyråer. Ett varmt tack till dem for hjälp härvidlag.

Ett problem är valet av variabel. Bör man välja det gamla kriteriet medborgarskap eller det sedan en tid tilgängliga födelseland? Ett test med de 48 viktigaste nationella grupperna i Sverige visar att utfallet varierar från den ena ytterligheten till den andra. En tredje variabel, språket. kan bara testas for Finlands del på grund av beklaglig brist på data från de øvriga nordiska Iänderna. De empiriska resultaten redovisas i en rumslig fallande skala från landsnivå. via länsnivå till kommunnivå (regionalt partiellt) och slutligen exempel på utfall intrakommunalt från de fyra huvudstadsregionerna

Orsakerna till de redovisade, øverraskande stora differenserna på samtliga regionala nivåer kan i detta sammanhang endast diskuteras översiktligt De påvisade förhållandena ger en bas till såväl vidare diskussion som till utväljandet av regionalt begränsade områden för intensivstudier utifrån andra dicipliners synvinklar.

Economic difficulties and divorce risks for four immigrant groups in Sweden

Diana Corman, Demography Unit, Stockholm University

The main objective of this paper is to investigate the interdependency between economic conditions and union stability. Previous studies suggest that immigrants have higher risks of unemployment and of union dissolution than those of native Swedes, but we have little knowledge on the interconnections between the two events. Unemployment and union dissolution point to the hinders in the process of integration but also to the different adaptive strategies used by different groups of immigrants. We have access to retrospective life histories of two thousands people born in Poland, Turkey, Chile and Iran and who lived in Sweden at interview (1996-1997). Iran and Turkey have in common the same dominant religion, Islam, and the similar culture. We can couple Chile and Poland on similar religious grounds, as they both belong to the Roman Catholic church. The four chosen immigrant groups have different degrees of cultural and geographic "distance" to Sweden and therefore are expected to have different integration potentials and opportunities. Other explanations for divorce risks are the immigrants' differential access to socio-economic resources. Polish women may choose a different adaptive strategy to the family's economic hardships than Turkish women, and might choose to divorce their current partner instead of staying put with a partner they are dissatisfied with.

Integration ur ett familjeperspektiv

Gun Alm Stenflo, Statistics Sweden, Stockholm

[Abstract missing]

Innvandrernes medbragte uddannelse

Henrik Mørkeberg, Statsitics Denmark

[Abstract missing]

Theme 10. The integration of immigrants in the labour market

Co-ordinators: Kirk Scott, Dept of Economic History, Lund University; Pieter Bevelander, IMER, Malmö University

Do Country-Specific Skills Lead to Improved Labor Market Positions? An analysis of unemployment and labor market returns to education among immigrants in Sweden

Ann-Zofie Duvander, National Social Insurance Board, Department of Research Analysis and Statistics, Stockholm

The gap in labor market rewards between immigrants and the native-born is sometimes explained with reference to immigrants' lack of country-specific skills. This study investigates whether speaking and understanding Swedish well, having an education obtained in Sweden and living with a Swedish partner improve immigrants' position in the labor market. The findings show that these characteristics do not substantially reduce the risk of unemployment, and the risk remains clearly above the level of native-born Swedes. However, employed immigrants with a Swedish education and very good language skills are as likely as Swedes to be educationally over-qualified for their job. In sum, country-specific skills are helpful in the process of reward attainment, but do not go all the way in accounting for the labor market disadvantage of immigrants. The residual may be due to discrimination.

[Title missing]

Martin Klinthäll, Dept of Economic History, Lund University

[Abstract missing]

[Title missing]

Pieter Bevelander, Malmö University

[Abstract missing]

[Title missing]

Kirk Scott, Lund University

[Abstract missing]

Theme 11 [cancelled]

Theme 12. Domestic migration and regional population trends

Co-ordinator: Mats Johansson, Swedish Institute for Growth Policy Studies (ITPS), Östersund

Befolkningens bevægelser på husstandsniveau

Henning Christiansen, Statistics Denmark

Hidtil har bevægelsesstatistikken (fødte, døde, flytninger og vandringer) været kendetegnet ved at være en hændelses(person)baseret statistik. Som noget nyt er det nu muligt opgøre antallet af hele husstande/dele af husstande, der flytter/vandrer. Enlige, enlige med børn, par, par med børn mv., der flytter kan belyses .

En nyhed, der ved samtidig intruduceres i bevægelsesstatistikken er oplysningen om hvorvidt: 1) Den fraflyttede bolig efterlades tom eller beboet samt 2) Den tilflyttede bolig stod tom eller beboet ved tilflytningen. Det er således blevet muligt (juni 2000) at belyse flyttemønstrene indenfor landets grænser og se flytningen i sammenhæng med den efterladte og tilflyttede bolig. Eksempler på hvad der kan belyses: 1) Antallet af flytninger hvor hele familien samlet flytter 2) Antallet af flytninger hvor unge flytter hjemmefra til egen bolig 3) Antallet af flytninger hvor der flyttes fra egen til kærestens bopæl 4) Antallet af flytninger foreteaget af unge der flytter hjemmefra til bofællesskaber. Yderligere kan husstande berørt af fødsel og dødsfald også belyses. En Nyt artikel herom blev for 1. gang offenliggjort d. 21. juni 2000 i Nyt fra Danmarks Statistik nr. 254.

Domestic migration trends in Sweden

Tor Bengtsson, Statistics Sweden

Domestic migration trends over municipality border is discussed from the following perspectives: levels of in- and out migration for a municipality, age distribution of the inhabitants for a municipality, typical regions according to patterns of domestic migration, economical consequences of domestic migration.

New regional migration patterns in Sweden

Ola Nygren, Royal Institute of Technology, Stockholm

Inter-regional migration and business cycles has showed a stable pattern in during the post-war era. During economic upturns, aggregated net migration has been relatively small and the regional structure in terms of jobs and population has remained rather stable. In downturns, metropolitan areas and other big cities have been net migration winners, mainly due to job losses in small regions with obsolete industrial structures.

In the later part of the 1990:s, this pattern was broken. Despite a strong economic upturn, resulting e.g. in a net growth of some 300 thousand jobs, migration patterns has mainly remained the same as the first half of the decade, which to a large extent was characterised by economic recession and huge job losses. The skewed allocation of new jobs certainly accounts for part of the explanation for the new migration trend, but not all.

Another observation from the recent three-year economic upturn, is the lack of the traditional corresponding increase in gross inter-regional migration flows. A slow long-term decrease in inter-regional migration has been observed for many years, but the recent absence of an up-swing in migration flows is so notable that it may raise questions to whether it could hamper the functioning of the national labour market and consequently economic growth.

A third observation from the 1990s is the weakening of the inter-dependence between regional labour market changes and inter-regional net migration patterns. Persisting high unemployment rates as well as an increasing mis-match between demand and supply of labour are supposed to be the most important explanatory factors in this respect. The paper deals with inter-regional migration between 109 local labour markets in Sweden, focusing recent trends.

The typical migrant

Helène Marklund, Statistics Sweden

Domestic migration in Sweden last year is described from the following perspectives: a comparison of the total number of domestic migrations in Sweden with Norway, Denmark and Finland; migrations over various domestic borders; migration-intensity for different age groups; the migration behaviour of men and women; and, the migration behaviour of tenants and homeowners. Taken together, these perspectives provide a comprehensive description of the typical domestic migrant in Sweden.

Modernization and Migration Flows of Two Cities, An Extended Gravity Model: Case Studies of Brisbane and Stockholm

Carl-Johan Rohlin, Applied Population Research Unit, School of Geography, Planning and Architecture, The University of Queensland

Wilbur Zelinsky phrased a Hypothesis of Mobility Transition in 1971, daring to relate all aspects of mobility to the Demographic Transition and stages of modernization. The hypothesis suggested that, not only are fertility and mortality related to modernization, but also the other component of the population equation: migration. This paper presents research that seeks a relationship between the levels of modernization of city regions and the migration flows to and from city regions. If a relationship between modernization and migration can be determined, what aspects of modernization are well correlated with the migration flows?

Brisbane and Stockholm represent two small global cities or city regions. Correlations between migration and key factors relating to modernization are examined. Both international and internal migration of each city region is analysed. The migration flows to and from more than one hundred regions for each city were included in the analysis. The size of in-flows, out-flows and net-migration were compared with the modernization level of origins and destinations. A range of proxy-variables was used to measure the defined regions modernization levels. Proportion working in the service sector followed by urbanization rate were the selected proxy variables for modernization suggested with both a method of fois a relationship between modernization and migration, and provides an interesting analytical model for the two city regions. It is possible that further research can provide a model for predicting future migration flows of global cities.

The crisis of small and medium-sized towns: Dual Sweden revisited?

Mats Johansson, Swedish Institute for Growth Policy Studies (ITPS), Östersund

During the second half of the 90s population development has once again been in focus for the regional policy debate in Sweden. Now it is not only the out-migration from the traditional depopulation areas that has been in focus – instead it the population decrease of the small and middle-sized towns and municipalities that have experienced a ‘population crisis’. The problems of these oved in to the local or regional centres with an accentuated lop-sided age structure as one consequence. This paper analyses the processes behind this partly new migratory movements and its consequences for the small and medium-sized towns and municipalities.

Regional policy: Creating functioning or caretaking regions?

Lars Olof Persson, Nordregio, Stockholm

In spite of 35 years of active regional policy in Sweden, the regional population trends have remained more or less unchanged. Long term net out migration has reshaped the demographic structure of numerous regions. No doubt, negative populations spirals have eroded the basis for functioning labour markets. It is easily calculated that more or less the entire remaining labour force soon will be demanded for taking care of the ageing population. which generally spells inequalities in living conditions between regions - have become too severe to be treated within the existing policy frameworks, the general response is to invite a Parliamentary commission to monitor the situation, redefine the objectives and to agree on new policy measures. This happened the first time in the mid 60s, as migration flows from Norrland towards the southern urbanizing and industrializing regions reached unprecedented high numbers. The reaction at that time was the ambitious set of objectives: equal access to jobs and services in all parts of the country.

Fifteen years of implemention of this first regional policy, however, could neither divert nor reverse the concentration processes. Hence, the same problems reappeared in the late 80s. The new objectives agreed upon in the new commission set up in 1987 reveal the parliamentarians deep trust in the power of central policy. The objective for regional policy was now said to be to "create functioning regions economic, social, cultural and ecological terms". The at that time common view of the state as a responsible, knowing and helping hand is obvious. The Commission expected sector co-ordination and cooperation between actors to occur, initiated and led by the counties, the prolonged arm of the state at the regional level. However, since no specific suggestions were made on how this decentralization of regional policy should be implemented, it became a quite ineffectual expectation.

Challenged by slow and uneven economic growth next Parliamentary commission was set up in January ´99 (SOU 2000:87). In many ways, the problems perceived looked much the same as in the late 80s: strong growth in a few metro regions, particularly in Stockholm and a continuous depopulation of small local labour markets with lagging industry. Notwithstanding the Swedish economy being healthier than for a long time before, the commission’s point of departure is that increased economic growth in all regions will be the main objective also for regional policy. No doubt, the commission launches and argues for the idea of regional policy to be the superior economic and industrial policy. It strongly rejects economic theory, which is ascribed to treat the national economy to be localized to just one single point in space. The logical consequence of this reasoning – to suggest a superior sector co-ordinating role for regional policy both at the central and regional level, is however not put forward in the report. Beside the objective "increased growth in all regions" – indeed a big task for a small policy - the main objective for the proposed new policy is now "well functioning local labour market areas". Again, an ambitious and attractive, but equally vaguely defined objective, readily consented by all political quarters.

In the analytical sub-contracted reports, the Commission clearly displays the current wide variation of local labour markets in terms of economic and social performance. According to the commissioned experts’ forecast, 55 out of the 60 smallest regions are to lose still more of their population already as a necessary consequence of the demographic structure. (SOU 2000:97 Report 19).

But for the first time, in its suggestions for new policy measures, this Parliamentary Commission seems indifferent to the forecasted future Sweden walking in two different directions. In fact, the commission has agreed on that maintenance of regional population numbers is neither a task nor an option for regional policy. The responsibility for the state in ”long term subsidy-depending regions” is said to be nothing more or less than to secure ”likeworthy” basic service provision.

To conclude - the perceived dominating regional problem in Sweden since more than 50 years - the great geographical divide of the labour market through concentration of population and market based economic activity to a few larger city regions and the eventual marginalization of a large number of small local labour markets, has been the prime motive for two generations of regional policy in Sweden: the national policy initiated in the mid 60s and later supplemented by EU structural policy. The shortcomings of these two generations of regional policy in changing the concentration trend have become obvious to parliamentarians as well as the voters. Currently, as a surrendering response, the regional dualization of Sweden seems to have become fully accepted more or less by consensus by this Parliamentary commission. If not in the rhetoric of growth in all regions and nicely functioning local labour markets, but still so in policy practice suggested.

Theme 13: Historical fertility and nuptiality patterns, 1700-1900

Co-ordinator: Martin Dribe, Lund University

Fertility and Economic Stress in Eurasian Perspective 1750-1900

Tommy Bengtsson, Lund University

What is lacking in aggregate studies of the fertility responses to short-term changes in harvest (prices or real wages; for an overview see Galloway 1988, Bengtsson and Reher 1998) is the impact of different household and family characteristics, such as social status, household composition, and previous fertility history. Aggregated studies not only lack detail but also analysis of the causal mechanisms: whether the fertility response is due to malnutrition and bad health of parents, to splitting up of couples due to temporary migration, or to planning births. The aim of this paper is to study fertility from a household perspective, taking different household characteristics into account, and to discuss the causal mechanisms behind short-term economic fluctuations. The analysis is made for five populations in China, Japan, Belgium, Italy and Sweden during the pre-industrial period. We use longitudinal data on individuals from the EurAsia Project on Population and Family History and employ a combination of time series and event history analysis. In the paper we show that the fertility among lower socio-economic strata, day-laborers and artisans, in the European sites and in Japan are influenced by short-term economic stress. In the Chinese populations it is the farmers themselves as well as the artisans that are vulnerable to changes in food prices, while soldiers and functionaries, with fixed payment in kind, are not. We also find evidence that the response was planned in several populations, including the Italian one, while this was not found for the Swedish population. Thus the same socio-economic groups suffered from short-term economic stress in Europe and Asia apart from China, which had an entirely different distribution system but the way the vulnerable groups responded were different.

Industrialization and fertility patterns in Laplandish mining communities at the turn of the century 1900

Stefan Warg, Umeå University

Uppsatsen behandlar fruktsamhetsutvecklingen i Kiruna och Malmberget 1900-1920. Utgångspunkterna för arbetet anknyter till E.A.Wrigleys respektive M. Haines modell för demografisk utveckling i gruvindustriella miljöer.Dessutom kommer resultaten att kopplas till diskussionen rörande stopping vs. spacing. Metoden bygger dels på Coales fertilitetsindex, och dels på analyser av två giftermålskohorter, representerande giftermål ingånga 1897-1903 respektive 1918-22. I det senare fallet har cox-regressionsanalyser av sociala- och kulturella variationer i födelseintervall genomförts. Resultaten pekar mot att även om indextalen för äktenskaplig fertilitet var höga under den tidigare delen av undersökningsperioden, kan viss förekomst av barnbegränsande åtgärder i form av förlängda födelseintervall konstateras för gruvarbetarbefolkningen. På 1920-talet har indextalen sjunkit samtidigt som födelseintervallen minskat, vilket indikerar förekomsten av ett "stopping behaviour" inom populationen vid denna tidpunkt.

When the Swedish Family Became Urban”

Hans Nilsson, Linköping University Lars-Göran Tedebrand, University of Umeå

[Abstract missing]

Marriages and Women without Husbands in Danish Towns in the 19th century

Anette Jensen, Danish Center for Demographic Research, Odense University

In the 19th century most of the Danish population got married at least once in their life, and marriage was the starting point for creating their own families. Researchers of historical demography have for many years focussed on marriage, marriage rates and age at first marriage in relation to the economic development and these factors influence on the fertility rates. Even though marriage was the standard, approximately 10% of the adult women population were living on their own whether as widows or as unmarried. If we want to describe women’s life in the past, it is consequently necessary to discuss the situation of these women who where living alone.

The purpose of the paper is also to portray these women who were living on their own. I will treat widows and unmarried separately, even though they have much in common, but they had as an example a different status in relation to the law, and it seems as the fact that widows once had been married gave them more opportunities in their widowhood than those women who never married had. Based on studies in Danish towns in the 19th century I will characterize these women who were living without husbands. What kind of household did these women join? By which means did they support themselves? How can we explain why they were living on their own? The paper will also deal with some aspects of the problems of finding sources that tell us something about women without husbands.

Regional Variations in the Age at Marriage in Sweden 1750-2000

Christer Lundh, Lund University

The aim of the paper is to study the trends and regional variations in the mean age at first marriage in Sweden during the period 1750-2000, and to analyse possible connections to economic, demographic and institutional change. Did trends in age at marriage depend on changes in the standard of living, variations in birth cohorts or in the sex ratio, or was it due to changes in marriage preferences as a result of the overall modernisation of society? Can regional variations in the mean age at first marriage be explained by differences in the economic or demographic structure or were they due to cultural differences? For the twentieth century, the calculations of the singulate mean age at marriage are based on SCB population statistics. For the period prior to that, the study relies on Gustaf Sundbärg’s methods of reconstructing the Swedish population by sex, age and civil status in order to be able to calculate the singulate mean age at marriage.

Theme 14a: Historical perspectives on infant and child mortality

Co-ordinator: Sören Edvinsson, Umeå University

Social differences in infant mortality in Denmark 1785-1840

Grethe Baangaard, Danish Center for Demographic Research, Odense University

Before 1835 Danish Public Statistics on mortality were divided into ten-year age groups, and because of that, we only know little about the distribution of infant and child mortality for the youngest group. Comprehensive collections of data from parish registers are necessary in order to shed a light on this problem. The early mortality decline in Denmark began in 1775, but the decline was steadier after an epidemic of smallpox in the years of 1786-87. In this study a cohort of newborn babies from the period 1785-1840 from a number of Danish rural parishes are followed through the first years of their lives, and different variables are tested against the mortality-risks of these children. Social differences for mothers and babies are the main focus, and there seems to have been a much more varied picture than expected.

Working class families and infant deaths. A study on infant mortality in Linköping, Sweden, mid 19th century.

Magdalena Bengtsson, Dept. of Health and Society, Linköping university

Earlier studies have shown that infant mortality was relatively low in Linköping in the middle of the 19th century, around 150 deaths per thousand. There were few outbreaks in infectious diseases, such as measles and whooping cough, and smallpox had practically disappeared. The risk of getting infected by contaminated water still existed, as the system of water mains and sewage was not installed until the 1860’s. At the end of the 19th century population increased due to immigration and this affected housing in a negative way, especially for the lower classes. In 1840-49 two thirds of all children in the city came from working class families and these children experienced a relatively high mortality, 186 deaths per thousand.

The purpose of this study is to investigate infant mortality in working class families in Linköping in the middle of the 19th century. The questions in focus are concentrating on how economic conditions and household structure act together and how they influence infant mortality. Biological factors, such as mother’s age at first birth and parity will also be taken into account.

Decline of infant mortality 1814-1878, Asker and Bærum, Norway

Eli Fure, Riksarkivet, Oslo

We know from national aggregate statistics that infant mortality declined in Norway during the 19th century, but finer details about the decline have to be studied at the individual level. For the parish of Asker and Bærum close to Oslo, a population database has been established which is well suited for analyses of infant mortality. Almost 14000 legitimate children are included in the analysis.

The paper shows that the decline in mortality was strongest in the youngest age intervals. After the age of six months there was no decline in mortality in the period. The probability of dying increased significantly if earlier siblings had passed away shortly after birth. The neonatal mortality did not depend upon season of birth, older children had significantly higher mortality during winter. The importance of doctors and midwives did not increase during the period.

This pattern indicates that changes in endogenous causes, i.e. causes related to the mother, her pregnancy and the birth were more important to the decline in mortality than exogenous causes, such as changes in epidemics or changes induced by local or central authorities such as better medical care.

The Cause of Death Pattern among Infants, Stockholm 1878-1925

Eva Bernhardt, Stockholm University, and Bo Burström, Karolinska Institute

Stockholm in the late 19th century was a densely populated urban area, undergoing rapid expansion of its population, both through natural increase and through substantial in-migration. It was also an extremely dirty city and demonstrated a gloomy mortality picture, showing that the living conditions in Stockholm were extremely dangerous to the health of its inhabitants, compared to other cities of comparable size. However, in the last decades of the 19th and the first decades of the 20th century there was a remarkable change, especially for infants and children.

In this paper we propose to investigate the decline in infant mortality in Stockholm in this period, focussing on the changes that occurred in the cause of death pattern among infants. This task is facilitated by the existence of an excellent data source, the Stockholm Historical Database (SHD). It contains the computerised population registers for the parishes of Maria and Katarina (that is the area of Södermalm) 1878-1925, which together had about 50 000 inhabitants at the beginning of the period. It is possible to analyse infant mortality by cause of death (the major killers being diarrhoeal and respiratory diseases), since the information in the population register has been matched with the death certificates in the Health Board archives.

Preliminary analyses of data from 1885-1900 showed that 3,990 of 49,779 children died during their first year of life, 14% of whom died in their first month of life. Diarrhoeal diseases caused 35% of the deaths; pneumonia and bronchitis 26%; other infectious and childhood diseases (including measles, pertussis, diphtheria and tuberculosis) caused 13% of the deaths. The remaining quarter of deaths was due to other and ill-defined causes. No major changes of mortality rates among infants occurred during the period. Diarrhoeal diseases declined somewhat towards the last years of the 19th century, while other causes showed only annual fluctuations.

As we have good information on age at death for the infants, we will divide the first year of life into segments, which will enable us to study trends in for example neonatal and post-neonatal mortality. Our previous studies have suggested that the decline in mortality began in older age groups and in higher social classes. In later extensions of our analysis we will study changes in the cause-specific pattern of death in sub-groups of infants born in and out of wedlock and by social class of the head of household.

The sins of their fathers: Caring for children with hereditary syphilis in Sweden 1840-1930

Anna Lundberg, Dept of Historical studies, Umeå University

The paper looks into the rise of paediatric care for children with hereditary syphilis, focusing especially on the rise of so-called "Welander-homes". Children that suffered from this disease were pitied already in the 1840s, but their lives were considered blessed from unnecessary pain when cut short by death. Many were treated in isolated cellars in public orphanages. In the early twentieth century, increased medical knowledge, a passionate professor of venerology and private funds made it possible to start asylums for these infants where they could be restored to health during the course of several years. By the 1930s, infant mortality had decreased slowly but significantly and the disease was far from as common as it had been 90 years earlier.

Inequalities within the Family. Infant Mortality in Nineteenth Century Sundsvall

John Rogers, Department of History, Uppsala University Sören Edvinsson, Demographic Data Base, Umeå University, and Anders Brändström, Demographic Data Base, Umeå University

The role of the mother for the survival of children has been recognized in several studies in the developing world. One important aspect of this is the mother’s education and literacy. When infant and child mortality has been studied in historical contexts, the important role of the mother has been emphasized together with factors such as socio-economic and environmental conditions. Mother’s literacy and its significance for her position within the family has however only occasionally been explored in relation to the survival of children. This paper tries to rectify the situation

All children (ca 55.000) born in the Sundsvall region during the 19th century are followed from their birth, death or out-migration. The reading marks in the catechetical registers are used as a measure for the literacy of the parents and its impact on survival among children in 19th century Sundsvall is analysed. Not only literacy level per se is analysed, but also differences in literacy level and differences in social background between the spouses are considered. During the transformation from an agrarian society to an industrial, the importance of mother’s literacy increased. The connection was stronger in the urban environment, in contrast to what has been found in earlier studies.

Theme 14b: Public health in history

Co-ordinators: Jan Sundin & Marie C. Nelson, Department of History and Department of Health and Society, The Tema Institute, Linköping University; Sören Edvinsson, Demographic Data Base, Umeå University

The origins of public health measures in Norway in the 17th century

Ole Georg Moseng, Dept of community medicine, Univ of Oslo

The appointment of the first doctor with public funding in Norway in 1603 stands out as a symbol of the rudimentary health care system introduced in many European countries in the 17th century. He represents, in many ways, a process in which the concern for physicians, surgeons, midwifes and apothecaries mirrored the state’s growing willingness and ability to assume responsibility for the health of its subjects.

Two other aspects of this process also seem important: Firstly, the authorities shouldered the burden of founding and developing institutions for the disabled, the poor, the mad and the beggars – in short, those who were unable to take care of themselves. Secondly, the most striking achievement concerning health was probably that the most dangerous of all epidemics, the plague, was conquered by way of administrative measures based on quarantines and isolation.

The increasing awareness of health conditions appears closely related to fundamental manifestations of modernisation. The anti-epidemic campaigns may, for instance, be linked to the concept of mercantilism. For a state, which, to an increasing extent, built its strength upon a large and prosperous population – that was able to pay the taxes demanded and to raise the soldiers needed – it would be a logical consequence to make sure that the objects of taxation did not pass away during epidemics. The poor relief institutions, which hardly were distinguishable from hospitals, workhouses or houses of correction, were also considerably influenced by mercantilism, but also capitalism and, to a certain extent, protestantism. The authorities faced a double task of fighting poverty as well as the poor. And the state revealed its double face of Janus: Controlling and repressive on one side, caring and paternalistic on the other.

At the same time, medicine was no less based on the dogmas of Galen than in the middle ages. It took nearly a couple of centuries before the medical scientific revolution found its way into the doctor’s office.

Scarlet Fever and the Fate of Children in Late Nineteenth Century Sweden

Marie Clark Nelson, Department of History and Department of Health and Society, Linköping University

During the late nineteenth century the decline in childhood mortality was interrupted for several decades. The causes of this setback were complex indeed, with elements related to urbanization, industrialization, social changes, and even the spread of epidemics of childhood diseases. This paper explores the role played by the epidemics of scarlet fever in the increase in deaths at this period of time. An overview of the effects of the epidemic on the aggregate level (county) is supplemented by an in-depth study of a specific epidemic.

Birth rate and birth giving during the depression of 1930’s: Class, Poverty and Unemployment in the industrial City of Tampere

Jarmo Peltola, Tammerfors

The official unemployment rate rose in Finland less than elsewhere in the Western Europe during the 1930’s depression. Many other usual manifestations of a depression failed to appear as well in the then still predominantly agricultural country. Self-sufficient economy had important role for the livelihood of many families, especially on the countryside. The depression had a severe impact on landless agricultural and forest workers and their families who lost their jobs for some years and were often forced to rely on poor relief. A system of relief work was created in order to help the unemployed, but such jobs were not available in all parts of the country.

Tampere stood out as different in comparison with the rest of the country. The town was the sole fully industrialized, somewhat larger and vigorously growing industrial town in the country. 73 % of active labour force belonged to working class. The main part of the industry produced goods for the home markets. The core labour force, especially in the textile factories, consisted of young women, though 10 percent of the working women were married. The metal and construction workers were usually men. Wood, paper and shoe industries hired both men and women.

The depression brought the town’s labour market in to a standstill. 20 per cent of the textile workers, 30 per cent of the metal workers, 40 per cent of the wood and paper workers and up to 70 per cent of the construction workers did loose their jobs. In addition, the working week was reduced into 3 to 5 days in all those major companies that managed to avoid bankruptcy. Civil servants also faced wage-cuts. Due to the decreasing consumption demand in the town the shops’ sales dropped, as well as the demand of services, which naturally had an impact on the level of consumption of those working in the service sector. The direct total decrease of jobs was 5 600 (1932), which included c.a. 15 per cent of the economically active labour force in the town.

The level of social security was low. The basic social assistance consisted of means tested poor relief, which assisted people were obligated to pay back. The state and the municipalities established relief work-sites where men could get work for short periods. Those women who supported their families or themselves could work in the workhouse established by the municipality. The municipality organised also food-aid, the food being delivered, as a compensation for work, for those who were officially accepted into the unemployment registers. Only those who were entitled to poor relief could become registered as unemployed.

The birth- and marriage-rates dropped significantly during the depression. The amount of children born fell from 15 per 1000 inhabitants per year - where it was before the depression - to 10 during the deepest slump. Similar drop took place also in Helsinki and other parts of Finland, but proportionally the drop was the deepest in Tampere. In this paper I discuss how the depression affected births and marriages. Placing the case of Tampere in a wider national and European context will shed light also on the question why did birth- and marriage-rates fall. The different positions of the mothers belonging to different social classes is also touched.

Theme 15: Mortality and health studies

Co-ordinator: Helge Brunborg, Statistics Norway

Old-age mortality in Norway

Helge Brunborg, Statistics Norway, Oslo

The mortality of old people in Norway has been declining much less than for younger age groups over the last decades. For the very old, those over age 95, there is even a tendency of increasing mortality. This is contrary to the trends in most other countries, except for the Netherlands. Incorporating this trend in population projections has substantial effects on the number of the oldest-old perithout a foreign background, who migrated in Sweden by crossing a county border during the period 1970-1999. The control group consist of individuals who did not migrate during the same period and who reside in their county of birth. Background variables are: gender, age, income, educational level, socio-economic level, marital status and duration of stay in the new county.

Migrants and regional variations of cardiovascular disease mortality in Norway

Erik H. Nymoen, Statistics Norway, Oslo

Regional differences in mortality of cardiovascular diseases (CVD) in Norway have been examined in several studies. Among Norwegian counties opposite time trends as well as different levels has been observed. Regional differences in CVD mortality has been related to various regional characteristics prevalent during the period where deaths are observed. Migration is a possible source of error in geographically based studies of the associations between disease and the environment. The importance of individual data for more than the period just before death is enhanced by etiological theories emphasising the effects of different experiences of changing welfare conditions. E.g. if poor living conditions in childhood increases>

Co-ordinator: Peter Sköld, Umeå University

Befolkningsudviklingen i Grønland 1800-2000

Jette Jensen, Nuuk University

Paperet vil indeholde en beskrivelse af den grønlandske befolkningsudvikling gennem 200 år med udgangspunkt i opgørelser og folketællinger tilbage til kolonitiden. Med udgangspunkt i befolkningstilvæksten analyseres udviklingen i dødelighed, fertilitet, flytninger og vandringer. Det søges herunder belyst hvilken indflydelse den førte kolonipolitik har haft for udviklingen fornævnte kvotienter og ikke mindst befolkningens størrelse, sammensætning og geografiske koncentration. Der vil i denne forbindelse ske en gennemgang af de væsentligste programmer iværksat af den danske regering. Paperet er et forsøg på at tegne et demografisk portræt af et isoleret, arktisk samfund der undergår en forvandling fra koloni til moderne velfærdssamfund på 200 år.

A geographical analysis of Spanish Flu morbidity and mortality in Norway 1918-1919

Svein-Erik Mamelund, Oslo University

The Spanish Influenza of 1918-1919 is o Norway and two dependent variables, morbidity and mortality after infection of the Spanish Influenza, are analysed using weighted least squares. The independent variables are the rural-urban dimension, exterior or interior situation, and presence of different means of communication, climate, dominant economic sector, income, financial circumstances, religion, ethnicity, and standard of housing in each district.

The ecological data comes from a variety of sources, but most are from Health reports and the census of 1/12 1920. The findings of this paper are not consistent with the view that Spanish Influenza was a uniform and egalitarian killers as many contemporary writers and some recent scholars have suggested. First, the proportion of Sami out of total population was positively associated with Spanish Flu mortality. The explanation may be biological and epidemiological, i.e. a lack of inherited and acquired immunity. Second, the effect of percentage Adventist-Mormon out of total population on influenza-pneumonia mortality was negative. The result may be explained by the high standard of nutrition and regular lifestyle in these religious groups. Third, average height of recruits also seems to matter. Where average height was high, influenza and pneumonia mortality was low. This finding is possibly due to different standards of past nutrition and disease history. Finally, wealth per person was found to be negatively associated with influenza-pneumonia lethality.

Migration pattern and family structure of the Swedish Saami population between 1960-90

Per Sjölander, Södra Lapplands Forskningsenhet, Vilhelmina

[Abstract missing]

Theme 17: Historical demography - Migration

Co-ordinator: Lars-Göran Tedebrand, Umeå University

Family names as indicators of in-migration to Norway prior to 1801.

Gunnar Thorvaldsen, Registreringssentralen for historiske data, University of Tromsø and Sølvi Sogner, Dept of Hsitory, University of Oslo

As part of a three-volume publication on the history of immigration to Norway, Solvi Sogner writes about the period before 1815. The 1801 nominative census is the period´s major cross-sectional source for population studies, but regrettably gives no direct information on birthplace. In order to compensate for this, the family names in the digital version have been grouped according to onomastic criteria, and non-Norwegian names classified by national origin. On this basis, a tentative regional distribution of immigrants during early modern times has been mapped.

The Meaning of Being Male or Female Migrants. Demo-geographical Aspects on the Spatial Mobility and Marital Path of Young Single Men and Women Heading for Sundsvall, Sweden, in the 1870´s

Marie-Christine ( Lotta ) Vikström, Department of Historical studies, Umeå University

It is often hard to unravel the many movements of past migrants on an individual level. Unfortunately, most quantitative sources, such as the censuses, are unable to give complete data on the migrant´s place of birth, departure and new destination. However, thanks to the computerized parish registers stored at the Demographic Data Base, Umeå University, it is possible to cover a large part of the migrant´s geographical background, their time of residency and place of new destination. In order to shed further light on the migratory path of individuals on a sequential level, the study focus on young, single men and women heading for Sundsvall during industrialization. Firstly, the degree of urban background, the Sundsvall residency and place of new destination will be analyzed between the genders. The aim is to distinguish differences, or similarities, in men and women´s wayof moving to and fro Sundsvall. To what extent, for instance, did the urban background determinate their will to stay or move on. Secondly, the marital behaviour of the migrants is analysed with regard to their geographical background. For instance, did a common migratory past have any influence on the migrant´s choice of partners? In the light of these two topics, the presentation will consider about 1.500 young unmarried migrants who arrived at Sundsvall. a saw mill town situated about 400 kilometers north of Stockholm. in the 1870´s.

Migration of Rural Families in Nineteenth Century Southern Sweden. A Longitudinal Analysis of Local Migration Patterns

Martin Dribe, Department of Economic History, Lund University

Most research on historical migration patterns has focused on more long-range migration processes such as rural-urban, inter-regional or inter-continental flows, neglecting the more local migration patterns, which could also have a profound impact on the rural preindustrial economy. This paper analyzes the migration of rural families in nineteenth century southern

Sweden using a longitudinal data set at the micro level. The analysis shows a rather high mobility of families, of which an overwhelming majority was over very short distances. Access to land and social networks, family composition as well as family type played important roles in determining the migration decision of these families. The results clearly show that migration was an important part of the lives of preindustrial families, and that the new migration pattern emerging in the second half of the nineteenth century, when rural-urban as well as international migration increased considerably, was not so much a result of a higher mobility in general, but rather a response to changing incentives making long distance migration more attractive.

NDF hovedside | Hva er demografi? | Demografistudier | Demografi-lenker | Om NDF | Arkiv

Sidene på er validert ved hjelp av W3Cs HTML-validator, og følger W3Cs spesifikasjon av HTML 4.0 Transitional. Sist oppdatert 1. januar 2002.

Valid HTML 4.0! Valid CSS!