Fertility & reproductive health strand abstracts

Fertility: cross-national and global perspectives: Monday 9 September 1.30pm

The future of fertility: Results from a global survey of experts
Stuart Basten University of Oxford; Tomas Sobotka, Vienna Institute of Demography, Krystof Zeman, Vienna Institute of Demography

In Summer 2011 an Internet survey on the likely future trends in fertility, mortality, and migration and the main factors behind them has been conducted among the members of major population associations. The survey, organised as a collaboration of IIASA and Oxford University, will become a basis for new global probabilistic population forecasts by age, sex, and level of education. By allowing a large number of experts to participate and by providing an argument-based underpinning of numerical estimates about future fertility trends, the survey addresses two common weaknesses of population projection-making: 1) very limited or no theoretical foundation and 2) a participation of a small and often closed group of experts in formulating the parameters of projection scenarios. Our study presents first results of the survey module on low fertility, which focuses on countries that are relatively rich and have at present low fertility. The experts were offered 46 arguments potentially relevant for the future trends in fertility, and clustered into six groups of factors. They have selected a country or a group of countries, to which their assessment pertains, and also provided numerical estimates of the likely range of the period total fertility rates in 2030 and 2050. Altogether, 184 experts have assessed the low-fertility module of the survey. These experts represented 41 low-fertility countries on 6 continents. We cluster these countries into regions and analyse the responses along several key dimensions: numerical estimates of future period TFR, the relevance and impact of individual arguments, and the relative importance of the clusters of arguments. We pay special attention to the arguments where experts gave contrasting views about their likely importance and impact on fertility. These diverging views either indicate existence of region-specific factors, or they signal considerable disagreement about the likely impact of some factors on future fertility trends.


Culture, structure and fertility: Education, post-materialism in a cross-national perspective
Ross Macmillan, University of Bocconi

Demographers have long-standing interests in issues of culture and fertility and such issues are central to theoretical discussions around the Second Demographic Transition (SDT). At the same time, there have been relatively few attempts to model culture directly and empirically examine its impact. At the same time, considerations of culture must be situated within the context of other large-scale social change, particularly changing patterns of educational attainment. Using data from women aged 15-49 from 77 countries in the World Values Surveys, this research takes on the culture and fertility question by examining the micro- and macro-level of effects of ‘post-materialism’ on individual fertility behavior. Post-materialism signals cultural shifts from a security-based to a actualization-orientation that echoes arguments about changing ideas about the ‘value’ of children. Results suggest a complex and unanticipated relationship between post-materialism and fertility. First, post-materialism is most important for fertility as a macro-level, cultural phenomenon rather than a micro-level, individual orientation. Second, post-materialist culture has a strong relationship with individual fertility, but the effect is curvilinear with fertility rates highest in culturally ambiguous contexts. Finally, post-materialism shapes the effect of education on fertility where educational effects are muted at both high and low post-material contexts. Implications for theory and research are discussed.


Lengthening birth intervals and their impact on the fertility transition in rural and urban East Africa
Catriona Towriss, Ian Timaeus, London School of Hygiene and Tropical Medicine

There has been concern about the progress of fertility transition in Sub-Saharan Africa: in many parts of the continent family sizes continue to be large. However, examining fertility rates by rural and urban area reveals that the decline is well underway in some populations. This research will investigate whether there is a relationship between low fertility and birth interval length in urban areas of East Africa. Birth spacing practices which served to delay births were an important element of fertility dynamics in many traditional communities. Current research shows that birth spacing continues be an important feature of fertility, and in a number of countries intervals have lengthened to five years or more. Regression models were used to analyse birth interval length by rural and urban area in four East African countries. Early analysis shows that, overall, intervals are longest and lengthening most rapidly in urban areas. Further analysis of these changes will give an understanding of the potential influence that lengthening birth intervals will have upon the fertility transition. Timaeus and Moultrie have argued that many African women who delay a birth do so for prolonged periods and some eventually become permanent postponers. Thus, if intervals continue to increase it will be impossible for populations to return to high fertility levels.


Reproductive and negative birth outcomes: Tuesday 10 September, 9.00am

Social inequalities of adolescent fertility outcomes: Teenage childbearing and abortion trends of three birth cohorts in Finland
Heini Väisänen, London School of Economics

Objective: To investigate why pregnancies occur more often among lower socio-economic group adolescents and study the determinants of choosing an induced abortion over childbirth.
Methods: Nationally representative register data from Finland on three female birth cohorts (1955-59, 1965-69 and 1975-79) are analysed (N=259,242). Cox regression models for socio-economic differentials in each cohort and pregnancy outcome (childbirth or abortion) and binary logistic regression models estimating socio-economic differentials in the odds of choosing an abortion for those who experienced their first pregnancy before age 20, are presented. The socio-economic differentials are measured based on the occupation the household member who has the highest socio-economic status and it is divided into seven groups: self-employed, upper-level white collar employees, lower-level white collar employees, manual workers, farmers, students, and others. Place of residence, ethnicity, relationship status, pregnancy history, and age at pregnancy are controlled for.
Results: Teens from lower SES background have higher risk of both outcomes even after controlling for other factors affecting fertility behaviour, but the differences are more distinctive in the childbearing model. Pregnant teens from upper-level employee backgrounds have from two to three times the odds of choosing an abortion compared to manual workers’ children. Conclusions: Despite the low levels of teen pregnancies in Finland, lower SES background continues to be associated with a higher risk of conceiving and giving birth. This finding indicates that new policy interventions are needed to reduce these inequalities.


Local sex ratios and reproductive skew: the Finnish case
Lassi Lainiala, Anna Rotkirch, Population Research Institute, Väestöliitto – Finnish Family Federation

Sex ratios are known to influence several aspects of marital and childbearing behaviour. Demographers have lately paid attention to the “marriage squeeze” in many Asian countries. The effect of sex ratios on fertility in contemporary Europe has, however, to our knowledge not been explored. Childlessness is increasing throughout Europe and mostly involuntary. “Not finding a suitable partner” is a frequent explanation, but the extent to which this relates to mating markets is unclear. We study the associations between local sex ratios, propensity to marry and to divorce, and fertility in Finland. Like the UK, Finnish fertility is on average comparatively high (TFR 1.83) but polarised. We use both macro-level data of municipality and sub-regional unit level and micro-level register data. Results show that that several municipalities have a very skewed sex ratio, 1.1-1.4 for Finns in their twenties. High regional sex ratio is associated with lower age at marriage of women, lower number of divorces, and higher fertility. Low sex ratios are related to higher age of marriage, more single women, higher share of cohabitation, higher divorce rate, and lower fertility. Thus the regions where women have many children appear to be those with high male childlessness, and the regions where men have many partners have high female childlessness. We suggest that reproductive skew within European societies can partly be explained by local sex ratios and call for comparative research on the topic.


Cooperation and conflict within families: do kin help or hinder reproductive outcomes?
Rebecca Sear, Cristina Moya, Susie Schaffnit, Kristin Snopkowski, London School of Hygiene & Tropical Medicine

The view that humans are ‘cooperative breeders’ is becoming established in the anthropological literature: mothers appear to parcel out the costs of raising multiple dependent children across several other individuals, most commonly fathers of the child or other men, grandmothers or older siblings of the child. But the reproductive interests of different family members will not always be in harmony. These conflicts of interest may show up as reduced reproductive success in the presence of certain kin members. Here we develop a framework for predicting when kin should help and when they should hinder reproductive outcomes. We then present a comparative study of kin influences on fertility, using data from all world regions, to demonstrate some systematic patterns of kin help and hindrance. Datasets we analyse include nationally representative large-scale surveys from both high and low income countries (including Demographic and Health Surveys, Family Life Surveys and the Generations and Gender Survey), as well as small-scale data sets collected by anthropologists from subsistence populations, including hunter-gatherers, horticulturalists and pastoralists. Broadly, we find that the presence of kin sometimes increases and sometimes decreases fertility, but that co-residence with kin tends to be more likely to be anti-natal than living near kin, and that a woman's parents-in-law tend to be more pro-natal than her own parents.


Associations between small area crime rate and negative birth outcomes in Scotland
Tom Clemens, Chis Dibben, School of Geography and Geosciences, University of St Andrews

Negative birth outcomes, which are important determinants of future child development, have been shown to be associated with both social and environmental characteristics of the mother’s area of residence. Air pollution has been identified as a key environmental factor but neighbourhood social stressors such as the prevalence of crime have remained relatively understudied. An important question is the extent to which area based characteristics exert an influence on birth outcomes independently of the individual socio-economic circumstances of the mother. This study examines the effect of small area crime rates in the mother’s place of residence for a number of birth outcomes (fetal development and risk of prematurity) with adjustment for a range of individual maternal characteristics including smoking. A sample of women was drawn from the nationally representative Scottish Longitudinal Study and births to these women (between 1994 and 2008) were identified through record linkage to maternity hospital admissions data. Maternal exposure to crime was estimated from the crime domain of the Scottish Index of Multiple Deprivation (SIMD) at small area level (datazones). The association between crime rates and birth outcomes was estimated from multilevel linear and generalised linear models. The preliminary findings, generally, indicate a significant relationship between levels of recorded crime at the mother’s place of residence and negative birth outcomes which remain significant after adjustment for a range of important individual confounding effects. These findings add to the growing body of evidence highlighting the independent association between local area characteristics and negative birth outcomes.


Fertility preferences and intentions: Tuesday 10 September 11.00am

Correspondence between intended and realised fertility: the role of education
Maria Rita Testa, Wittgenstein Centre (IIASA, VID/ÖAW, WU), Vienna Institute of Demography/Austrian Academy of Sciences; Maria Iacovou, University of Essex

The intended number of children is a strong predictor of subsequent fertility. However, the lack of long-span longitudinal surveys has hindered in-depth studies on the correspondence between intended and actual childbearing in Europe, especially the role of fertility postponement in the mismatch between the two variables. This is a relevant line of research, given the increasing share of couples who delay starting a family. Using longitudinal data from the British Household Panel Survey (BHPS) and its successor, Understanding Society (UKHLS), which cover the years between 1991 and 2012, we link planned fertility, as reported by women and men in their 20s and 30s in the first wave, with completed family size, as reported by the same women and men at the end of their reproductive careers, i.e., in their 40s and 50s at the last wave. We aim to reconcile the contradictory findings of highly educated women planning larger families than the medium to low-educated ones but having smaller families than their less educated counterparts by the end of the reproductive years. Our basic idea is that an increase in education from a low level is associated with a decrease in birth intentions, whereas an increase from a higher level is associated with an increase in fertility intentions.


Fertility intentions and health status among women in slum ans non-slum areas in eight Indian cities 
Sangita Kulathinal, University of Helsinki, Minna Säävälä, Population Research Institute, The Family Federation of Finland

Harsh and unpredictable environmental conditions make it ‘bio-logical’ for people to develop faster, mature earlier, and reproduce sooner rather than later. Reproductive intentions and behavior are not solely determined by social learning but may respond also to ecological context. Thus far we have mostly aggregate level analyses on the effect of higher mortality and morbidity on the timing of first birth (e.g. Nettle 2011). We have little knowledge on how individual health and risks to longevity, independently of SES, are reflected in fertility preferences and intentions in developing country conditions.  

This paper examines the relationship between reproductive intentions and health condition of women aged 15 to 29 in eight Indian cities. The data consists of the National Family Health survey (NFHS) 2005-6. We will examine whether individual differences in health (measured by weight, height and anemia) and awareness of the risk of death (manifested by the fact of having lost one’s parents), relate to the ideal number of children and preferred waiting time before the next desired birth.

We hypothesize that morbidity and perceived mortality risk relate to fertility intentions not only in the ecological level but also when using individual as unit of analysis. We will examine whether, when controlling for education, sex of the existing child(ren) and the urban risk environment, those women who have poorer health status and whose own parents are dead, would prefer more narrowly spaced births and would have a higher ideal number of children.


Nettle D 2011. Flexibility in reproductive timing in human females: Integrating ultimate and proximate explanations. Philosophical Transanctions of the Royal Society (Biological Sciences) 366: 357–365.


HIV/AIDS and sexual behaviour: Tuesday 10 September 1.30pm

Attitudes towards sexual health and sex behaviour among older adults with HIV in rural southern Malawi
Emily Freeman, Centre for Research on Ageing, University of Southampton

Adults aged beyond the reproductive years are estimated to contribute 14.3% of adult HIV disease burden in sub-Saharan Africa. Research regarding older adults’ sexual health attitudes and behaviours has typically focused on risk of transmission among HIV-seronegative or sero-unknown individuals. However, the sexual health attitudes and behaviours of seropositive individuals have implications for experiences of HIV after age 49, and onward transmission. This paper answers the research question ‘what are HIV-infected older adults’ attitudes towards sexual health and sexual behaviours?’ Data were produced in rural southern Malawi using repeat in-depth interviews (N=136) with men (N=20) and women (N=23) aged 50-90. A third of respondents had HIV. Interview data were supplemented by data from focus groups with older people with HIV (N=3), key informant interviews (N=19) and observations made over 11 months of fieldwork. Attitudes towards sexual health and sexual behaviours were tied to broader understandings of ageing, the body and illness. Both older age and HIV were understood to diminish an individual’s finite store of vitality, recognised by diminishing ability for farm, house and ‘bed’ work (sex). Work was in turn associated with what it meant to be alive: to be an ‘adult’. The adult identity represented the core identity respondents associated with and aspired to. By limiting ability to work age and HIV threatened respondents’ ‘adult’ identities. In response, they employed a range of narratives that realigned behaviours with the ‘adult’ identity. In these narratives, understandings of sex and the body underpinned an emphasis on either continued sexual vitality and activity or the avoidance of sexual risk-taking.


Gender Disparity in HIV Sero-prevalence and Associated Gender Variables: A national-level analysis of the association between gender inequality and the feminisation of HIV/AIDS
Katherine Harris, University of Southampton

The quantitative link between HIV prevalence and gender inequality is not well established at the national level, with papers from Greig & Koopman (2003) and Kashkooli (2006) suggesting that sub-Saharan Africa does not fit the pattern of a negative correlation between levels of HIV prevalence and gender equality. Furthermore, the sex ratio of HIV prevalence varies significantly from country to country, yet thus far has only been the subject of one quantitative analysis (Hertog, 2008). This paper therefore aims to extend the quantitative evidence base regarding the country-level variation in HIV prevalence and the sex ratio of HIV prevalence. Firstly, using country-level aggregate data for sub-Saharan African countries, bivariate analysis and partial correlation coefficients are used to assess the direction and magnitude of the relationships between indicators of gender equality (in social, economic and political domains) and the dependent variables (level of HIV prevalence and HIV prevalence sex ratio), whilst controlling for important institutional and socioeconomic differences between countries including wealth, population size, health expenditure, population growth, size of urban population and maturity of the epidemic. The paper uses data primarily from UNDP Human Development Report (HDR), and national household surveys which include a HIV component. Secondly, an OLS regression is conducted in order to better describe the form of the relationship between HIV prevalence and gender inequality. Preliminary findings suggest that the factors driving high HIV prevalence in a country are unlikely to be the same as those factors related to the discrepancy in HIV prevalence between the genders. 




Fertility: migration, residential contexts and spatial differences. Wednesday 11 September 9.00am

Residential Context, Migration and Fertility in Britain
Hill Kulu, University of Liverpool

This study examines fertility variation by residential context and its causes. While there is a large literature on fertility determinants in industrialised countries, little research has investigated spatial fertility variation. We study fertility variation across regions with different size and within urban regions by distinguishing between central cities and suburbs of the cities. We use longitudinal data from Britain and apply event history analysis. We investigate to what extent the socio-economic characteristics of couples and selective migrations explain fertility variation between residential contexts and to what extent contextual factors play a role. We also study childbearing behaviour of people who move from one residential context to another. Our analysis shows that the fertility levels decline as the size of an urban area increases; within urban regions suburbs have significantly higher fertility levels than the city centres.


Disentangling the quantum and tempo of immigrant fertility
Ben Wilson, London School of Economics

Migrant fertility has become an increasingly prominent explanation for recent fertility change. Immigrant Total Fertility Rates (TFRs) are consistently higher than native TFRs in the majority of European countries (Sobotka, 2008). However, research has cast doubt on the use of tempo-distorted measures of fertility for evaluating the contribution of immigrants to aggregate fertility (Toulemon, 2004, 2006). Furthermore, previous research has almost exclusively analysed incomplete fertility profiles, an approach which provides little insight into the absolute impact of immigrant fertility. Considering the UK, this research uses survey estimation and count regression techniques to disentangle tempo variation from the quantum of immigrant fertility. This approach is
crucial for testing migrant fertility hypotheses because migration is known to influence the timing of childbearing (e.g. Andersson, 2004). The results support several hypotheses, with evidence of disruption followed by elevated fertility, particularly for recent cohorts of women arriving after age 25. However, the results show no evidence of disruption for immigrants from Jamaica or Bangladesh, instead suggesting cultural maintenance for Bangladeshis and adaptation for Jamaicans. After exploring these results in detail, this paper discusses the benefits of this approach for understanding migrant fertility and informing population projections.


Fertility and values: a look at regional differences within the European context
Lara Patrício Tavares, Instituto Superior de Ciências Sociais e Políticas (Universidade Técnica de Lisboa) and DONDENA Centre for Research on Social Dynamics (Università Bocconi); Bruno Arpino, Department of Political and Social Sciences and RECSM, Universitat Pompeu Fabra and DONDENA Centre for Research on Social Dynamics (Università Bocconi)

Using data from the Eurostat Database and European Value Surveys we assess if recent fertility trends in Europe are associated with a change in values. A special emphasis is given to Spain and Italy that, together with the other Southern European countries, are often seen as a homogeneous group sharing the same ‘traditional’ values and demographic behaviours –as opposed to Scandinavian countries which are seen as progressive. We show that Italy and Spain are not that similar in terms of values. We also show that similarities at the country level with respect to TFR hide considerable variation at the regional level. We argue that an analysis at the regional level, as carried out here, is crucial to better understand changes in fertility levels. Our analyses provide evidence that recent fertility trends are associated with value dynamics, namely that the highest increases in TFR happened in regions where both individualism with respect to relationships and individual autonomy grew at the same time that individualism with respect to children diminished. We also provide empirical evidence in support of McDonald’s theory that both gender equity at the institutional level and within the family are necessary for fertility to rise.


A Spatial Analysis of Childbearing in Cohabitation
Agnese Vitali, Carlo F. Dondena Centre for Research on Social Dynamics, Bocconi University; Arnstein Aassve ,Carlo F. Dondena Centre for Research on Social Dynamics, Bocconi University; Trude Lappegård, Statistics Norway

The disconnection between marriage and childbearing represents one of the main recent change in the families, with the Nordic countries acting as the forerunners in this new trend. In this paper we study the diffusion of childbearing in cohabitation as opposed to marriage across Norwegian municipalities. We rely on spatial panel econometrics to investigate how childbearing in cohabitation spreads geographically in relation to three indicators: importance of religion, female educational expansion and economic uncertainty. Our findings indicate that female educational expansion is the most important predictor of childbearing in cohabitation. The innovation of this paper is found in the access to unique data which offer detailed information on all municipalities within a country. Using new advanced modelling techniques this paper provides new and improved knowledge about the diffusion of childbearing in cohabitation.