Abstracts - fertility & reproductive health strand

Strand organisers: Dr. Stuart Basten, University of Oxford; Dr. Monica Magadi, City University

Abstracts are presented in order of scheduled presentation. Please refer to the programme for timings.

Fertility rates and the background conditions of cross-border marriage couples in Japan: An analysis using census microdata samples
Kazumasa Hanaoka, Ritsumeikan University; Shuko Takeshita, Aichi Gakuin University

There has been an increase in the number of cross-border marriages in many countries. In particular, in Asian countries such as South Korea, Taiwan and Japan, local governments and marriage agencies have promoted cross-border marriage of foreign women with native-born males. However, fertility outcomes and background conditions of cross-border marriage couples are not studied in comparison with other couple types. Therefore, using 2005 Japanese census micro data samples and logistic regression models, this article examines differences in fertility related conditions and outcomes by considering combinations of husband and wife nationalities in three types of couples in Japan: cross-border couples, immigrant couples, and native-born Japanese couples. The results indicated that similarities in demographic and socioeconomic characteristics of cross-border couples were higher than immigrant couples except between Japanese husbands and Korean/Chinese wives. The more immigrant and cross-border couples\' characteristics resemble that of native-born Japanese couples, the more they mirror fertility outcomes. Furthermore, our findings seem to suggest that lower fertility rates are mainly caused by the disadvantageous status of immigrant and cross-border couples comprising foreign husbands and Japanese wives. On the other hand, in the case of cross-border couples comprising Japanese husbands and foreign wives, their low fertility rates cannot be improved by controlling demographic and socioeconomic characteristics only. Thus, there seems to be problems associated with cultural differences, including difficulty in communication with partners.

Email: Dr. Kazumasa Hanaoka: kht27176@fc.ritsumei.ac.jp|

The future of fertility in China
Gu Baochang, Center for Population & Development Studies, Renmin University of China; Stuart Basten, University of Oxford

The family planning restrictions in China (often misunderstood as a universal ‘One Child Policy’) is one of the most comprehensive ‘population policies’ in the world. Recently, however, it has come under increasing scrutiny in terms of its efficaciousness, the demographic consequences (in terms of very low fertility, ageing and skewed sex ratio) as well as on a human rights basis. In this paper we use information from a meta-review of fertility ideals and intentions in China to explore the extent to which small family sizes have become socialised and normalised within China and, hence, suggest whether the family planning restrictions in their current form are, to a degree, redundant. We conclude that even when taking into account potential misreporting, stated childbearing preferences in both urban and rural China show remarkable similarities to elsewhere in East Asia. We suggest that reform of the current family planning regime in China is unlikely to lead to a significant increase in the fertility rate.

Email: Dr. Stuart Basten: stuart.basten@spi.ox.ac.uk|; Professor Gu Baochang: baochanggu@gmail.com|

The effects of father absence on fertility in transitional Malaysia
Paula Sheppard, Rebecca Sear, London School of Hygiene and Tropical Medicine

Introduction: Paternal investment in humans is not well understood. In the less developed world, numerous studies have examined the effect of father absence on child health and survival with mixed results. Much less is known about the effects of father absence on later reproductive outcomes.
Methods: Using data collected in Malaysia during 1988 (the Malaysian Family Life Survey, n=567), this study tested the effects of father absence on daughters\' age at menarche, age at first marriage, age at first birth and desired completed family size. Results: We found that father absence up to age 5 (but not at later ages) was associated with earlier age at menarche, adjusting for mother\'s age at menarche, ethnicity and wealth. Father absence, although only during later childhood (6 to 15), is associated with younger ages at first marriage and first birth, even after controlling for other confounders. Father absence does not affect desired family size. Conclusion: The results found in this transitional population in some ways mirror those in developed societies, where father absence accelerates puberty and reproduction, although the precise mechanisms which bring these relationships about are still unclear.

Email: Paula Sheppard: paula.sheppard@lshtm.ac.uk|

The effects of unemployment on fertility
Signe Hald Andersen, Rockwool Foundation Research Unit Denmark; Berkay Özcan, London School of Economics

We analyze the causal effect of unemployment on fertility. Neoclassical theory of fertility has ambiguous (both positive and negative) predictions regarding the effect of unemployment for women. Additionally, existing empirical research shows contradictory results and makes a weak case for exogeneity of unemployment to fertility behavior. We suggest that (unexpected) firm closures constitute an exogenous source of unemployment and adopt it as an instrument to estimate husbands’ and wives’ fertility response, using a unique administrative panel data from Denmark, which includes all residents in Denmark between 1982 and 2006. It contains monthly information about employment, relationship and a very-detailed fertility history -including stillbirths and miscarriages- of individuals as well as information about the firms that they work in. We estimate our models separately for men and women. Our preliminary results show that unemployment as a result of a firm closure negatively affects both women’s and men’s completed fertility and positively women’s timing of the first birth. Men do not appear to delay timing of the first birth due to unemployment.

Email: Dr. Berkay Özcan: B.Ozcan@lse.ac.uk|

Early life environments and first-time parenthood: The experience of Swedish women, 1990-2009
Serhiy Dekhtyar, Kirk Scott, Lund University

It has been shown that detrimental birth contexts, for instance famines, can exert long-lasting effects over female fertility outcomes. Such events, however, are rare, too extreme with respect to the strength of exposure, and therefore offer limited utility for health policy design in Western societies. Moreover, they instrument only one feature of birth environments – nutritional shortage. The role of non-acute contextual characteristics of birth contexts which encompass more than just the effects of under-nutrition for entry into motherhood is still unknown. Moreover, the few Swedish studies that did link early exposures and first-time fertility did not address the issue of postponement with long-enough follow-up periods. We address these gaps by relating entry into motherhood with non-extreme, exogenously-determined contextual characteristics of female birth environments captured by regional-level infant mortality in Sweden in 1970-1976. Doing so allows us to asses not just the nutritional effects, but rather the totality of early health environments, taking pollution, standards of living, and infectious disease burden also into account. In addition to using high-quality Swedish register data, we are able to strengthen causal interpretations of the results by employing family-based comparisons and controlling for unobserved family-level confounding. Our results indicate that even mildly severe birth environments are associated with a reduced hazard of transitioning into motherhood.

Email: Serhiy Dekhtyar: serhiy.dekhtyar@ekh.lu.se|

Did your mother work? Impact of mother\'s occupational status on individual\'s fertility
Maria Rita Testa, Wittgenstein Centre (IIASA, VID/ÖAW, WU), Vienna Institute of Demography/Austrian Academy of Sciences; Valeria Bordone, Wittgenstein Centre (IIASA,VID/ÖAW, WU), Vienna University of Economics and Business

According to the theory of Planned Behaviour people\'s fertility intensions are formed with the contribution of three sets of factors: attitudes toward childbearing which can be either positive or negative, subjective norms which are related to the opinion of relevant others about having a child, and the perceived behavior control (i.e. the perceived ability to realize the fertility plans). In this theoretical framework the reasons for a mismatch between stated fertility intentions and subsequent reproductive behavior are attributed to the errors individuals might have made in forming their decision. An important source for mistakes could be attributed to either an underevaluation of constraints or an over-evaluation of opportunities, such as lack of a suitable partner, infertility or unexpected events, but also learning about parenthood benefits and costs. Drawing on longitudinal data from the Multipurpose Household Survey on "Family and Social Subjects", carried out by the Italian National Institute of Statistics (Istat) between 2003 and 2007, we study fertility intentions of Italian women aged 20 to 49 by the working status of their mother when they were 14 years old. We expect working experiences in the family of origin to be a key determinant of fertility intentions, especially for younger people at the beginning of their eproductive career. Preliminary results suggest that considering whether a person had a working mother contributes to the knowledge on forming, implementing and changing fertility intentions.

Email: Dr. Valeria Bordone: valeria.bordone@wu.ac.at|

Infant feeding and children\'s psycho-social development
Maria Iacovou Almudena Sevilla, University of Essex

Many bestselling childcare books recommend feeding babies to a schedule, and the relative merits of scheduled as opposed to demand feeding are an extremely hot topic of discussion among new parents. However (and in contrast to the enormous literature on breastfeeding), there exists hardly any research on the effects of the different modes of feeding on children\'s outcomes. In this session, I will present the results of what I believe to be the first ever large-scale study of this issue, showing that while schedule-feeding is associated with higher levels of maternal wellbeing, it is also associated with lower test scores for children later in life, of the order of 17% of a standard deviation in SATs tests, and 4-5 IQ points in IQ tests. In an extension to this work, I also examine the relationship between schedule-feeding and children\'s psycho-social development, showing that schedule-feeding is associated with poorer scores on the Strengths and Difficulties assessment, particularly in the domains of hyperactivity and conduct problems. I assess the possible mechanisms behind these relationships, and discuss the types of questions which would need to be carried in future child development surveys in order to take this vein of research forward. The research is based on a sample of 10,419 children from the Avon Longitudinal Study of Parents and Children, a cohort study of children born in the 1990s in Bristol, UK, and the methods employed include OLS, logit and Propensity Score Matching.

Email: Dr. Maria Iacovou: maria@essex.ac.uk|

Revealing the costs of reproduction in Gambian women using multi-process, multilevel models
Rebecca Sear, London School of Hygiene & Tropical Medicine

A question of both theoretical and applied interest is whether women experience \'costs of reproduction\'. Since reproduction is energetically expensive, there should be evidence that reproduction has consequences, in terms of a woman\'s subsequent health. However, these costs have proved surprisingly hard to demonstrate empirically. The problem is that health and reproduction are interdependent: not only is reproduction expected to affect health, but reproduction depends on health. Here, I control for this problem by using a multi-level multiprocess statistical model which simultaneously measures the influence of health on reproduction and the influence of reproduction on health. I use data from a longitudinal study of Gambia women, during a period of high fertility, and estimate the relationships between: number of children (measuring reproductive effort), and (1) BMI and (2) haemoglobin level (as measures of health). Results show that when using an ordinary regression model to test for effects of reproduction on health, reproductive effort is not correlated with BMI and is positively correlated with haemoglobin. When using the multiprocess approach, haemoglobin is still positively related to reproductive effort, but a negative correlation between reproductive effort and BMI is revealed, suggesting that these Gambian women suffer at least some costs of reproduction.

Email: Dr. Rebecca Sear: rebecca.sear@lshtm.ac.uk|

Kinship, co-operation and fertility: quantitative findings from the KASS study of European kinship
Patrick Heady, Max Planck Institute for Social Anthropology

Field research for the KASS ("kinship and social security") project was conducted in 19 localities spread across eight countries in different parts of Europe. Ethnographic fieldwork was accompanied by detailed quantitative interviews in each locality using a computerised kinship network questionnaire to collect detailed data on fertility, mutual assistance and the spatial and social structure of each informant's network of kinship and marriage ties. The aim was to compare and evaluate theories originating from evolutionary and social-cultural anthropology and sociology. The findings show the expected patterns of Hamiltonian kinship altruism and of reciprocity, but there are also marked geographical differences in the extent of kinship interaction and in gender roles, which appear to be due to both economic and cultural factors. Consistently with "cooperative breeding theory”, mothers receive a good deal of support from close relatives – particularly from their partners. Grandparents also play an important role – but one that is balanced by help to grand parents and by a negative correlation between grandparental involvement and domestic help from husbands. Both the "cooperative breeding" and the "kin-influence" hypotheses suggest that fertility should be highest where relatives live close together, since then their ability both to help and to influence parental couples will be greatest. However, comparing fertility levels between the 19 field sites we find that although fertility is positively correlated with levels of kinship clustering among rural areas the opposite is true for urban areas. I will conclude by briefly reviewing some possible explanations for these findings.

Email: Dr. Patrick Heady: heady@eth.mpg.de|

A spatial panel analysis of Italian regional fertility
Agnese Vitali, Policy Analysis and Public Management, Bocconi University

Italy is a case study in lowest-low fertility. Its internal heterogeneity is substantial, and changing over time. If, historically, the South was characterized by high fertility levels, in recent years the trend has inverted as it is the North which now shows the highest regional fertility. This paper uses the theoretical framework offered by the diffusionist perspective to fertility transition to describe the current temporal and spatial trends in Italian provincial fertility in association with a series of indicators of marital behaviours, female occupation, contribution of foreign fertility and economic development. Referring to the spatial econometric literature we make use of geographically weighted regressions and spatial panel regressions to model explicitly spatial dependence in fertility among Italian provinces in the 2000s. Results show that spatial dependence in provincial fertility persists even after controlling for the usual correlates of fertility. Further, we find that the local association between fertility and its correlates is not homogeneous across provinces. The strength and in some cases also the sign of such associations vary spatially.

Email: Agnese Vitali: agnese.vitali@unibocconi.it|

The mismatch between contraceptive use and fertility in sub-Saharan Africa
Akim J. Mturi and Godswill Osuafor, Population Training and Research Unit North-West University (Mafikeng Campus) South Africa

During the last two decades or so fertility was observed to be declining in various countries in Sub-Saharan Africa. The use of contraception has been named as the main proximate determinant of fertility responsible for the decline in fertility. It has been illustrated that a rise in contraceptive uptake reduces fertility levels. For instance, Mauldin and Segal (1988) used data from 86 countries around the globe to fit a model that shows the relationship between total fertility rate and contraceptive prevalence rate (CPR). Many researchers have used Mauldin\'s and Segal\'s model to check the consistency of the relationship of these two variables. Some African countries have been observed not to follow the relationship shown in this model. This phenomenon is referred to as the mismatch between fertility rate and contraceptive use. In other words, there are some countries which have much higher or much lower fertility rates than the ones expected given the level of contraceptive use observed in that country. This paper intends to investigate the mismatch phenomenon. The specific objectives of the paper are twofold. Firstly, the 1988 Mauldin\'s and Segal\'s model is refined by using the most recent Demographic and Health Survey (DHS) data sets for all African countries. Secondly, six countries have been selected to investigate the causes of the mismatch.

Email: Professor Akim Mturi: akimmturi@gmail.com|

Fertility postponement and rising educational enrolment
Maire Ni Bhrolchain, Eva Beaujouan, University of Southampton

The rise in educational enrolment is often cited as a possible cause of the trend to later childbearing in developed societies but direct evidence of its contribution to the aggregate change in fertility tempo is scarce. We show that rising enrolment, resulting in later ages at the end of education, accounts for a substantial part of the upward shift in the mean age at first birth in the 1980s and 1990s in Britain and in France. The lengthening interval to first birth over that period has two components: a longer average period of enrolment, and a post-enrolment component that is also related to educational level. The relationship between rising educational participation and the move to later fertility timing is almost certainly causal. Our findings therefore suggest that fertility tempo change is rooted in macro-economic and structural forces rather than in the cultural domain.

Email: Professor Maire Ni Bhrolchain: mnb2@soton.ac.uk|

Immigration, fertility and tempo effects
Sylvie Dubuc, University of Oxford

The paper discusses migration-specific tempo-effects on estimating fertility of immigrants. Novel fertility patterns of immigrants in the UK prior and after migration are presented. Results support the LFS-OCM methodology proposed to address the problem of overestimation of fertility of immigrants by Period TFRs (due to tempo effects) and also provides novel data to analyse the interrelation of migration and family building hypothesis. Additionally, the timing of family reunion for immigrants\' children temporarily left-behind overall and across main mmigrant/ethnic groups is assessed. Overall results suggest that the phenomenon is unlikely to distort significantly LFS-OCM fertility estimates (i.e. estimates using survey data and a reverse survival method) of established immigrant groups.

Email: Dr. Sylvie Dubuc: sylvie.dubuc@spi.ox.ac.uk|

Migration and fertility convergence
Ben Wilson, London School of Economics

Previous research has attempted to empirically test migrant fertility convergence, but has failed to define this concept in detail, or develop an agreed methodology. This may explain why there are separate tranches of literature on migrant fertility, and why the evidence for convergence remains fragmented, with a confusing array of competing hypotheses and explanations (Forste & Tienda, 1996). This presentation will propose several definitions of migrant fertility convergence, alongside discussion of how these may be tested. Importantly, convergence can be assessed using the number of children born for completed cohorts (Frejka, 2008; Frejka & Sardon, 2007), and this well as those working on demographic projections (ONS, 2007; Sobotka, 2008). Tests of migrant fertility convergence also involve disaggregating migrant generations (Parrado & Morgan, 2008; Parrado, 2011), which requires information on migrant ancestry (e.g. parental country of birth). For the UK, these criteria are met by recently released Understanding Society data. This presentation will introduce these data, in particular questions on migration and fertility. Results of tests for completed fertility convergence will then be shown. For similar reasons to prior research on migrant fertility (Adsera et al. 2012; Adsera & Ferrer, 2011; Mayer & Riphahn, 2000), this analysis uses regression models for count data. Finally, the results of these models will be discussed, including an appraisal of: whether completed fertility converges, the characteristics (and migrant origins) associated with convergence, and the implications for future research.

Email: Ben Wilson: b.m.wilson@lse.ac.uk|

Friends, family and fertility: Does thinking about close social networks influence childbearing intentions? Does thinking about future childbearing influence the composition of a close social network?
Paul Mathews, University of Essex; Maria Iacovou, University of Essex; Rebecca Sear, London School of Hygiene & Tropical Medicine; Ernestina Coast, London School of Economics

Social scientists and evolutionary biologists have both had a long running interest in the relationship between close social networks (primarily family structures) and childbearing. This research adds a psychological aspect to this literature by looking at the effects of priming respondents to think about these components. Data come from the broadly representative Innovation Panel of \'Understanding Society\' (the UK Household Longitudinal Study -UKHLS). We employ randomised experimental methods, whereby groups are created that are theoretically systematically identical except for one manipulated \'treatment\' characteristic; in this case the preceding priming questions, these groups are then compared to look at the effect of the treatment. In this study we first look at the effect of priming individuals to think about their close social network on their fertility intentions. Secondly, we look at the effect of priming thoughts salient to future fertility on the composition of the individuals\' reported close social network. Our initial results show that priming individuals to think about their close social network leads to slightly higher reported fertility intentions.

Email: Dr. Paul Mathews:

Is slowing postponement really the engine for total fertility rate rises in European countries?
Marion Burkimsher, University of Lausanne

Data for 14 European countries in the Human Fertility Database (HFD) have been analysed to assess the demographic reasons for the recent rises in total fertility rates (TFRs). Previous work has suggested that slowing postponement rates have been the main contributor to the rising TFRs. However, on closer examination, three different stages in the latest \'fertility transition\' have been discovered: • In the first stage, in the years just after the turning point in the TFR, postponement rates (ie. the rate of change of mean age at first/second birth) continue to increase (surprisingly). In this phase, TFR increases stem primarily from increased intensity of first birth rates, especially amongst older first-time mothers • In the second phase, postponement rates slow and the this is the main driver of TFR rises • In the third stage, which affects only a minority of countries studied, the \'real\' fertility rates rise again, including a modest rise in larger families. Not all countries pass through all stages, and Portugal is an exception in not seeing any recent increase in TFR. Of particular interest is the increasing spread in age at first birth (and in most cases second birth also). This has happened across all countries studied, whilst some countries have also seen an increase in peak fertility rates.

Email: Dr. Marion Burkimsher: drmarionb@gmail.com|

Family formation via mutual reciprocity between fertility and marriage: A cross-European comparison
Mark Lyons-Amos, Brienna Perelli-Harris, ESRC Centre for Population Change, University of Southampton

The order of marriage and fertility in the European context has become increasingly disrupted in recent decades. Historically, marriage was a pre-requisite for fertility; now the transition from cohabitation to marriage and fertility comprise two processes which jointly constitute family formation. However, existing studies either measure family formation via a single process or as two associated processes with no direct or \'causal\' effect on each other. This approach is inadequate where marriage and childbearing are jointly planned. We propose an extension to existing literature by evaluating the inter-relationship between marriage and fertility in a cross-European setting. We also analyse differences in the timing of fertility in different settings: for example, in Scandinavian contexts family formation can begin with either marriage or birth, in Italian settings marriage is a prerequisite for fertility. Data are taken from a subset of the Harmonised Histories, a series of comparable European social surveys which incorporate both fertility and partnership histories. Using a sub-sample of women in birth cohorts 1945-54, 1955-64 and 1965-74, we employ a cross lagged probit model. This model overcomes the deficiencies of previous association based analyses by evaluating the reciprocal, causal relationship between union status and fertility. As well as establishing the inter-relationship between marriage and fertility, the model is expanded to evaluate national level variation in the fertility and marriage associations via interactions. Preliminary results indicate interdependence between fertility and marriage: evidence of joint planning. However, more work is required to determine the precise nature of country specific effects.

Email: Dr. Mark Lyons-Amos: mja303@soton.ac.uk|

First births in Europe: socio-economic differentials in the effect of economic and institutional contexts over the life-course
Karel Neels, Jonas Wood, Zita Theunynck, University of Antwerp

Postponement of parenthood is one of the most marked demographic changes in Europe since the early 1970s and constitutes a major cause of sub-replacement fertility that emerged in the 1980s and 1990s. Factors driving fertility postponement include increasing enrolment in education of younger generations, changing opportunity resulting from increasing educational attainment and difficult access to the labour markets due to successive economic recessions after the early 1970s. Recuperation of fertility later in the life-course, however, is associated with social policies that reduce the opportunity costs of fertility and support dual earners in combining labour force participation with family formation. In this paper we use data from the 3rd round of the European Social Survey (2006) to assess the impact of economic context, labour market conditions and policy context on first birth hazards of men and women in 14 EU countries between 1970 and 2005. Using multilevel discrete-time hazard models, we focus on the differential effect of these contextual factors on family formation by age, gender and socio-economic position of individuals. Results show that adverse economic conditions - and particularly high Unemployment - significantly reduce first birth hazards of both men and women below age 30, with effect being more pronounced among the higher educated. After age 30, economic conditions continue to affect first birth hazards of men, but not of women. Finally, we draw indicators from the comparative family policy database to assess the impact of institutional contexts on the recuperation of fertility, and differentiate these effects by age, gender and socio-economic position of individuals.

Email: Jonas Wood:

A multi-level analysis of kin influence on fertility in the developed world using the Gender and Generation Survey: the roles of paternal investments and socioeconomic status
Susan Schaffnit and Rebecca Sear, London School of Hygiene & Tropical Medicine

The Gender and Generations Survey (GGS) data offer a unique opportunity to understand between and within country variation in interactions between socioeconomic status and kin and their effects on women\'s fertility in the developed world. Studies on kin effects on fertility are sparse in developed, post-demographic transition settings characterized by low fertility and mortality, long generation intervals, and women\'s participation in economic activities. I will present the results of an analysis of GGS data, representing over 82,000 women in 15 post-transition countries. Using multilevel discrete-time event history analysis and a subset of data including women under age 45 who have given birth within the five years prior to interview, I found that kin availability and paternal investments show highly variable effects on women\'s timing of births between countries. Within countries, the effects of kin and paternal investments on the probability of birth varied by socioeconomic status. Kin availability measures included indications of presence, emotional closeness, and time investments towards focal women\'s offspring. Paternal investment variables, on the other hand, measured fathers\' time investments in childcare and other child-related household tasks. My findings suggest that the costs and benefits of kin for women\'s reproduction vary between abstract socioeconomic divides and more concrete political boarders. I will discuss my results in light of other recent developed world kin studies and will consider the implications of my results for our understanding of post-demographic transition trends.

Email: Susan Schaffnit: susan.schaffnit@lshtm.ac.uk|