Poster abstracts

Strand organiser: Rachel Bennett, University of Southampton

Abstracts are presented in alphabetical order of first author's surname.

Gender equality and fertility
Arnstein Aassve, Carlo F. Dondena Centre for Research on Social Dynamics and Department of Policy Analysis and Public Management, Bocconi University; Trude Lappegård, Statistics Norway; Agnese Vitali, Carlo F. Dondena Centre for Research on Social Dynamics, Bocconi University

In this poster we test to what extent gender equality leads to a change, and in particular an increase, in fertility. We focus on Norway, a country that scores highly on gender equality, and has relatively high fertility. We make use of data from 430 municipalities during 2000-2008, including information about municipal fertility levels and a rather detailed description of municipal gender equality. Given the richness of our municipal data, we are able to decompose the gender equality index by six components, each representing a different dimension of gender equality: provision of formal childcare, female municipal council representatives, and four indicators of gender differences in education, age, labour force participation and income. It follows that if gender equality turns out to have a significant impact on fertility, we are able to establish which dimension of the index matters most.

Email: Agnese Vitali:  

Parental migration & child mortality in rural South Africa
Kojo Antobam, MRC/Wits Rural Public Health and Health Transitions Research Unit, University of Witwatersrand; Mark Collinson, MRC/Wits Rural Public Health and Health Transitions Research Unit, University of Witwatersrand; Victoria Hosegood, Africa Centre for Health and Population Studies, University of KwaZulu Natal; Rachel Bennett, University of Southampton

Levels of adult temporary out-migration are very high in rural South African communities as men and women migrate to pursue employment and education opportunities. The rural home is often considered to be the preferred place for children to grow up, and children are commonly ‘left behind’ in the care of extended family when parents migrate. While the economic benefits of migration for sending households are well-acknowledged, much less is known about the impact of parental absence on the health and welfare of ‘left behind’ children. This paper examines the relationship between parental migration and child mortality using detailed longitudinal data from South Africa’s two rural demographic surveillance systems. Discrete time event history modelling is employed to investigate the relationship between mortality in the first five years of life and maternal and paternal migration status, after controlling for individual, household and community level factors. By comparing results for the two study sites, it is possible to examine similarities and differences in the relationships between parental migration and childt mortality in contrasting social and geographical contexts.

Email: Kojo Antobam: 

An investigation into the migratory patterns and outcomes of young people to institutes of higher education in the UK between 2007 and 2011: analysis of the HESA Student Record Data
Neil Bailey, University of Southampton

There is no formal legal obligation to register a change of address in the UK. Therefore the measurement of internal migration around the UK is extremely difficult to quantify and inherent with uncertainty and underreporting. There is no dedicated administrative source that records and measures internal migration flows across the UK. Instead, proxy data, such as the registration and re-registration of patients in the National Health Service (NHS) is used. While the National Health Service Central Register (NHSCR) and the General Practitioners Patient Register Data System (PDRS) are important administrative data sources, they were not created with the purpose of measuring internal migration and as a result some statistical errors and miss/under/over reporting of certain sectors of the population occurs. Thus, migration is recognised as the most difficult component of population to estimate or project. Demographers have long recognised that persisting regularities appear in empirical age-specific migration schedules. When migration is viewed as an event, it is highly selective with regards to age, with young adults generally being the most geographically mobile in any population. It has been suggested that these peaks in the probability to migrate at these ages can largely be explained by moves of young people to and from university. Using the HESA Student Record Data, this research
aims to provide a more detailed understanding of the migratory patterns and outcome of young people to institutes of higher education in the UK.

Email: Neil Bailey:

The Double Burden of Child Malnutrition in Low and Middle Income Countries
Katie Bates, London School of Economics

This research explores a newly emerging nutritional crisis affecting under-fives in low and middle income countries – the Double Burden of Child Malnutrition (DBCM), which can occur at both the population level (countries with a high proportion of children experiencing two different forms of malnutrition: stunting (low height-for-age) and overweight (high weight-for-height)) and individual level (children that are simultaneously stunted and overweight; referred to as ‘stunting-overweight’
individuals). To the author’s knowledge, no quantitative analysis conducted to address the determinants of this emerging crisis. To attend this gap, anthropometric data for under-fives came from all available DHS and UNICEF MICS. Data on socioeconomic context, population health and nutrition availability were collated from the World Bank DataCatalog (WBDC), the UNDP Human Development Reports (HDR), the WHO Global Health Observatory (GHO) and the Food and Agriculture Organization of the United Nations (FAO). For the population level DBCM logistic regression models were fitted – the binary dependent variable indicating the presence of a DBCM (>20% under-fives stunted and >10% overweight) or otherwise. For the individual level DBCM
logistic random intercept models were utilised – the dependent variable indicating whether an individual under-five was ‘stunted-overweight’ or otherwise. A random intercept for country was included. The results show a need to take into account the heterogeneous nature of the nutritional problems among under-fives in LMICs, particularly in relation to the understudied DBCM. Key macro factors associated with both the population-level and the individual-level DBCM are: nutritional availability; socioeconomic context; and the stages of the nutrition and epidemiologic transition.

Email: Katie Bates:

Geographies of ethnic segregation in England and Wales: dimensions, scale and regional differences
Gemma Catney, University of Liverpool

Surprisingly little is known about where members of different ethnic groups in Britain reside, relative levels of concentration and dispersal, and how these have changed over time. This poster draws on work from a wider project ‘Geographies of Ethnic and Social Segregation in England and Wales, 1991-2011’ (The Leverhulme Trust, 2012-14), which aims to fill this gap by producing the first systematic account of local level segregation in England and Wales. This poster will focus on the interrelations of global measures of segregation by ethnic group at a variety of spatial scales (including England as a whole, sub-regions and cities), for 1991 and 2001. Of interest is how measures representing different dimensions of segregation can change at varying spatial scales; this effect is not the same for each ethnic group. For example, members of a given group may be concentrated in specific parts of a city, or spread evenly across a city; analysing segregation at different geographies allows us to explore these scale effects. Some research has outlined how segregation measures are correlated and has suggested that measures of only two of the five dimensions listed by Massey and Denton (1988) may capture most of the variation in population distributions (Reardon and O’Sullivan 2004). This presentation will engage
with these debates to assess how best to characterise ethnic geographies of England and Wales. The remainder of the poster covers planned methodological developments in measuring segregation locally (and at a low-level geography) in England and Wales from 1991 to 2011.

Email: Dr. Gemma Catney:

Determinants of fertility change in Malawi: evidence from 2000, 2004 and 2010 Malawi Demographic and Health Surveys
Jesman Chintsanya, Nyovani Madise, Claire Bailey, University of Southampton

The case of Malawi’s fertility pattern calls for rigorous research as Malawi’s total fertility rate (TFR) has persistently remained high at 6.0 children per woman and just recently declined to 5.7 children per woman in 2010. Modern contraceptive use rose from 7.4 per cent in 1992 to 42.2 per cent in 2010 but without a corresponding decline in fertility. Reducing high fertility by making effective family planning available among high-risk groups, such as women under age 18 or over age 34 and those with more than three children, is recognized as an important strategy to help reduce high maternal deaths at 508 for every 100000 live births. Given this scenario, we set out to answer what are the factors contributing to sustained high levels of fertility and examine whether women are using contraceptives for pacing or limiting number of children. By decomposing the demographic and socio-economic determinants of fertility, we examine the relative contribution to fertility change using the 2000, 2004 and 2010 Malawi Demographic and Health Surveys. As fertility behaviour can vary tremendously and is affected by age, religion, gender, educational attainment, urban or rural residence, as well as regional and cultural contexts and other socio-economic factors, logistic regression analysis will be performed controlling statistically for these confounding factors. To capture spatial variation in fertility behaviour among population sub-groups, the study shall include in the model random effects to account for contextual effects.

Email: Jesman Chintsanya:

French Caribbean migrations: from transatlantic space to transnational space. or : "French Caribbeans go transnational"
Stephanie CONDON, INED (Institut National d'Etudes Démographiques)

As citizens of the French republic, Martinicains and Guadeloupeans who move to mainland France are considered internal migrants. Therefore the dynamics of the migration between the two poles cannot be analyzed from a transnational viewpoint even though patterns of circulation, return, and maintenance of links are quite characteristic of transnational practices. Meanwhile, transport and telephone costs are constantly decreasing and internet communication plays an increasing role in forging and maintaining links between mainland France and the French Caribbean. Migration decisions are taken in a rapidly changing technological context in which social fields can be consolidated via new forms of communication. Thus many skilled young people try out both metropolitan and island labour markets before settling for a longer period. In parallel, increasing numbers of ‘French Caribbeans’, either born in the Caribbean or in mainland France (as descendants of migrants), have sought other destinations, among which the UK capital. Often skilled migrants or students, this movement satisfies various aims: extending human capital through improving their command of the English language and through work experience; expectations of being respected as a ‘normal’ (not ethnicized/racialized) French citizen; a break with what many see as a narrow (French) perspective on career opportunities. However, investigating the scale and the dynamics of this migration is not without its challenges, not least that of identifying the population in available statistics. A combination of sources and methods are thus used to investigate how French Caribbeans are now looking beyond transatlantic space and ‘becoming transnational’.

Email: Dr. Stephanie Condon:

Lives of women living with HIV/AIDS
Tinny Dawar, Sarita Anand, Lady Irwin College, University of Delhi

The poster explores the lives of HIV positive women and the impact of HIV infection and widowhood on their socio-economic status. It also documents the coping mechanisms adopted by these women to deal with the stigma and discrimination they face. The poster is based on data collected during face-to face interviews with women availing counselling services, educational and nutritional support for their children and support to start an enterprise, either from the government hospitals or voluntary organisations in Delhi, India. It was observed that women had little control over keeping their status confidential since many were tested either at the time of their husband’s prolonged sickness or after their husband’s death. Hence family members became aware of their HIV status. Few women reported that their spouses did not share their HIV status with them. In majority of cases, in-laws began taunting, scolding and blaming the women to be the cause of bringing the infection home. A significant number of widows were deserted by their in-laws after the death of their husbands. Hence parental support was valued the most since it reduced the stress level of women and helped them cope with the situation. The counselling of family members if the family is aware of the HIV status of the woman is very important. Another finding of the study is that the responsibility of providing care for their children helped them cope with the situation. It is suggested that the counsellors of the government hospitals must encourage HIV positive women to seek support from NGOs working in the area. Women availed various services such as nutritional and educational support for their children from NGOs. At the same time, being associated with a network of similar women in NGOs gave them a lot of relief.

Email: Tinny Dawar:

Life Opportunities Survey - a large scale longitudinal survey of disability in Great Britain
Fiona Dawe and Helen Tam, Office for National Statistics

The Life Opportunities Survey is a large-scale longitudinal survey of disability in Great Britain. It is a major social survey exploring disability in terms of the barriers to participation that people experience. The wave one results were published in December 2011, providing prevalence rates of impairment among the general household population, as well as exploring experiences of barriers across life experiences. Wave two data collection completed in March 2012 and analysis of the full wave two longitudinal dataset is due to be published in 2013 (with the dataset being deposited at the UK Data Archive shortly after). A short interim report of initial wave two findings, based on data collected in the first year of wave two was published in April 2012. This initial analysis focused on respondents who reported an impairment at one or both waves of the survey. Three groups of respondents were identified: • those who had reported an impairment at both waves of the survey; • those that reported an impairment at wave one but not wave two; and • those who reported an impairment at wave two but not at wave one. This initial analysis was purposely restricted because of the small sample sizes generated from the first half of the wave two sample. The availability of the full wave two dataset provides possibilities for further investigation of these groups, in comparison with respondents with no reported impairments, and their experience of barriers to participation including exploring changes between waves one and two.

Email: Fiona Dawe:

Son Preference and Prenatal sex selection: Recent trends in the UK Asian Diaspora
Sylvie Dubuc, University of Oxford 

Sex-selection against females is well documented in Asia, notably in India and China. Indian census data from 2010 suggests this practice has further increased and spread. Recent research has shown that prenatal sex-selection against females (PNSSaF) is geographically more widespread than previously thought, including accumulated evidence of PNSSaF among Asian Diaspora in western Countries. In the UK, a male biased sex ratio at birth among India-born women, over 1990-2005, reflects the trend observed in India. However, a closer analysis by main Indian region of origin suggests lower prenatal sex-selection against females (PNSSaF) among immigrant Indian women in the UK. Updated trend in the UK are presented; further monitoring will be necessary to confirm (or not) these new results. The causes and policy implications of PNSSaF in the UK/Western Countries are discussed.

Email: Dr. Sylvie Dubuc:

Public perceptions of development, participation, relationship with wellbeing: the case of Makueni district in Kenya
Hildah Essendi. University of Southampton

Although some of the key dimensions of development include standards of living and income, it is also important to recognise the importance of non-monetary factors, particularly in understanding the dynamics of socio-economic development. Whereas researchers recognise that human development entails much more than the rise or fall of national incomes, various indicators regarding wellbeing, health and development in developing countries remain below average. This is in spite of the implementation of many development initiatives in developing countries, particularly in rural areas. In addition, very few studies to understand and implement rural development have sought the views of community members regarding the important aspects and outcomes of
their development and their participation in development. Using the mixed-methods approach, this study investigates community perspectives on the key dimensions of development, and the impact of their perception and involvement in development on their wellbeing. The aspects/outcomes of wellbeing under focus in this study include fertility, use of maternal health services, child morbidity and mortality, vaccination and nutritional status. The study is implemented in Makueni district of Eastern Province Kenya. Data is collected at the community and household levels using focus group
discussions, key informant interviews, in-depth interviews, mapping and questionnaires. The study investigates the perception of community members regarding development, their participation in development and the relationship with their wellbeing. Qualitative data is analysed using QSR NVivo 9 software while the quantitative data is analysed using SPSS and Stata. Data analysis is ongoing.

Email: Hildah Essendi:

Alternative population bases and 2011 Census Outputs for England and Wales
Kanak Ghosh, Office for National Statistics

New questions asked in the 2011 Census for England and Wales allow for the possibility of outputs to be produced on population bases other than the standard definition that is used to count the population (the usually resident population base). This poster describes the proposed alternative population bases, how they can be constructed and provides some background information on the selection of the bases. In addition, the proposed outputs that may be produced for each will be discussed.

Email: Kanak Ghosh:  

Beyond 2011: Administrative Data Quality Framework
Kimberley Hall, Office for National Statistics

The UK Statistics Authority formally established the Beyond 2011 Programme on 1st April 2011 to assess options for meeting future user needs for population and small area sociodemographic statistics in England and Wales. The Administrative Data Quality Framework is designed to provide a framework for assessing the suitability of administrative sources being considered for use by the Beyond 2011 Programme. The framework is based on work undertaken by Statistics Netherlands and looks at three interrelated components:-
•Source - considers the suitability of the data source as a whole;
•Metadata - quality of
information required to understand and use the data;
 •Data - key quality measures or indicators. This framework will be considered as a key component of our overall quality assessment plans and an integral part of wider work to assess alternative options for the production of population and small area socio-demographic statistics in England and Wales.

Email: Kimberely Hall:

The division of household labour in Eastern European families: concurrent theories tested
Paul-Teodor Hărăguş, Babes-Bolyai University, Cluj-Napoca, Romania.

Our intention is to determine which of the most used theoretical constructs (“gender ideology”, “relative resources”, and “time availability approach”) manages to better foresee the division of housework inside the family. We will concentrate our analysis on three eastern European countries (Bulgaria, Romania and Hungary) and we will use the Generations and Gender Project GGP databases. We will analyze what factors can better predict the score of intra-family housework allocation, for men and women, using regression analysis (OLS). The strategy of analysis chosen is to construct different explanatory models for every theoretical perspective. The results shows that
the effect of independent variables are different on men and women. Gender ideology receives a small support (and it has a greater effect on men); relative resources theory is a better predictor for men while time availability approach explains better in the case of women.

Email: Dr. Paul-Teodor Hărăguş:

Female Empowerment and HIV-related Outcomes in Lesotho: Incorporating Levels beyond the Household and the Couple
Katherine Harris, University of Southampton

Using multilevel logistic models this study will explore the relationships between female empowerment and several HIV-related outcomes (HIV-related knowledge and condom use) over multiple levels in the specific context of Lesotho. Studies examining HIV-related outcomes have focussed on individual- and household-level characteristics, ignoring higher level factors which are closely interwoven (Ukwani et al., 2003). The existing evidence for community effects are largely based on qualitative or community-specific studies. This study is interested in the sociocultural construction of gender, which the female empowerment literature considers to be an important determinant of sexual behaviour. However, whilst the literature on female empowerment highlights that women may be unable to make behavioural changes in order to avoid HIV, this area also neglects the influence of factors at the higher level, choosing to emphasise the importance of couple dynamics. Whilst focussing on the level of the couple is both logical and relevant because sexual behaviour is negotiated between two people, couples are also located in a wider social environment (Ingham and Van Zessen, 1997). Both areas of the literature therefore neglect the effect of factors which operate on higher levels, which may lead to overestimation of the influence of individual-level and couple-level factors. The paper thus aims to build upon the limited quantitative evidence base for the importance of community level variables, with the overall aim of exploring the relationships between female empowerment and several HIV-related outcomes (HIV related knowledge and condom use) over multiple levels.

Email: Katherine Harris:

The first birth in Pakistan
Andrew Hinde, Jamal Abdul Nasir, Sabu Padmadas, University of Southampton

This paper reports some results from a study of recent fertility trends in Pakistan. The question we focus on relates to the determinants of the timing of the first birth. A substantial minority of women (about 6 or 7 per cent in the 2006-2007 DHS) conceive their first child before marriage. We investigate the characteristics of these women and compare them to those of the whole population. We then analyse the duration between marriage and first conception for women who conceive their first child after marriage. The results reveal a complex hazard function with two peaks: the first at duration 1 month and a second around duration 15 months. We suggest that this arises because the observed hazard is a mixture of two different functions applying to different groups of women.
(1) A group adhering to the traditional pattern in which conception takes place as soon as possible after marriage. These women have a hazard of first conception which declines monotonically with duration because of a selection effect operating within the group by which the biologically most fecund conceive early.
(2) A group which delays conception after marriage. This group is, in turn divided into two groups: (a) women marrying at very young ages who delay sexual activity for some time after the marriage; (b) women adopting a more ‘western’ form of reproduction in which some delay after marriage is normal.
We hypothesise that, over time, Groups 1 and 2a will form a smaller proportion of Pakistani women and Group 2b will form a larger proportion. An initial comparison of the samples in the 1990-1991 and 2006-2007 DHSs suggests that Group 1 is declining in importance over time in favour of Group 2, suggesting that Pakistani women are adopting more ‘western’ forms of childbearing.

Email: Dr. Andrew Hinde:

Household and community level determinants of school attendance in Sierra Leone: a multilevel analysis
Mamusu Kamanda, University of Southampton

The research applied multilevel logistic regression to the 2008 Sierra Leone Demographic and Health Survey to investigate the effects of household and community level factors on the probability of school attendance. Education is commonly recognised as a human right and a basis for development. The United Nations policy of Education for All is recognition of the right to education with its objective of expanding access to quality education to all children by the year 2015. Sierra Leone, a poor post-conflict state with high mortality and illiteracy rates, is a country in which the promise of Education for All will not materialise by the target date. It therefore forms a suitable context for the study. The investigation found that, net of individual level characteristics,
household and community environment have significant effects on school attendance. At the household level, wealth, education of the household head, and female headship were significant positive correlates of school attendance. Community level female education was also positively associated with the outcome while living in a poor community or in a rural residence had strong adverse effects on attendance. A significant interaction was found between the sex and age of the child, as well as between the sex of the child and female community education. Relative to boys,
girls are significantly less likely to be in school as they age. However, their probability of attending school is elevated by living in a community with a high proportion of educated women.

Email: Mamusu Kamamda:

Using Threshold Messages to Promote Physical Activity; Implications for Public Perceptions of the Dose-Response Relationship with Health
Emily CL Knox & Oliver J Webb - Loughborough University

Physical activity (PA) campaigns often feature participation targets (e.g. 150 minutes/week). These ‘thresholds’ may disrupt dose-response relationships between behaviour and its perceived health
effects. Benefits could be perceived as negligible if behaviour falls below threshold levels but substantial whenever thresholds are met. This study examined if such ‘discontinuity’ effects occur for PA messages. Structured interviews were conducted with a convenience sample (N=1100). Individuals either received a threshold message (150 minutes/week); a generic message encouraging ‘regular participation’; or no message (control). Participants used a scale to rate
perceived health effects of seven PA regimes, ranging from 10-290 minutes/week. Recipients of both message types shared consistent positive perceptions of regimens >150 minutes/week. For regimens <150 minutes/week, however, only the generic message group showed perceived benefits that were above control levels. The generic message group showed an appropriate curvilinear association between PA duration and perceived benefits. In comparison, threshold recipients showed suppressed perceptions for the shorter regimens (10-130 minutes/week), before a discontinuous surge in positive perceptions around the threshold [P < 0.05]. Threshold messages
were associated with lower perceived benefits for modest PA regimens. Consequently, current campaigns may be suboptimal for promoting the goal of small increases in PA at population level.

Email: Emily Knox:

Symbolic data analysis approach to clustering population pyramids
Simona Korenjak-Černe, University of Ljubljana, Faculty of Economics; Vladimir Batagelj, University of Ljubljana, Faculty of Mathematics and Physics; Nataša Kejžar, University of Ljubljana, Faculty of Medicine

In the poster we present an approach based on the symbolic data analysis to clustering of population pyramids. A population pyramid is represented with two frequency distributions over age groups, one for each gender. Such a representation with discrete distributions enables us to preserve information about real distribution and the size of the population of each gender. It is a special kind of symbolic data. For clustering units with variables described with discrete distributions we adapted well-known k-means and Ward’s hierarchical clustering methods such that the obtained optimal cluster’s representative is again population pyramid of all population in the
regions, included in the cluster. Both methods are compatible (they are solving the same clustering optimization problem) and therefore can be used in combination. Properties and advantages of such an adaptation will be presented. The results of applications of the proposed approach on world countries and on US counties will be presented.

Email: Dr. Simona Korenjak-Černe:

ONS Longitudinal Study Beta Testing
Shayla Leib and Jim Newman, Office for National Statistics

The ONS Longitudinal Study (LS) contains linked census and vital event data for 1 per cent of the population of England and Wales. Information from the 1971, 1981, 1991 and 2001 Censuses has been linked across censuses and with vital events data such as births, deaths and cancer registrations. Data from event registrations from 1971 to 2009 are now available on the LS database. The longitudinal nature of the study permits life-course and inter-generational analyses. Individual-level information enables the use of multivariate statistical techniques. ONS are currently linking 2011 Census data into the Longitudinal Study (LS) which will then include five censuses. We aim to launch a new research database incorporating the 2011 Census data in
autumn 2013. There will be two stages of testing to assess the quality of the LS research database. The first stage, alpha testing, is performed by ONS to evaluate the quality of the data and fix problems. For beta testing, researchers will be invited to submit proposals for appropriate projects that challenge and make good use of the updated LS using a pre-release version of the final database and supporting metadata. ONS will be asking for expressions of interest in beta test projects which will benefit both researchers and ONS. This stage is currently planned to take place
between spring and autumn 2013, with a launch event taking place at the end of this period. This poster will give an overview of the information contained in the LS and will highlight the potential for new analyses as a result of the inclusion of the 2011 Census variables. It will highlight the studies importance in furthering understanding of different social and demographic trends; and the sample sizes available for studying particular populations of interest. It is hoped that this overview will help researchers decide whether the LS is appropriate for their research and encourage proposals of beta test projects.

Email: Shayla Leib:

Measuring the urban penalty: hospital mortality in Coimbra, Portugal, 1885-1910
Mafalda Moura Pereira, University of Cambridge

My poster explores how measurement of the “urban penalty” in nineteenth-century Coimbra is potentially skewed by the deaths in the city of non-residents. Over the period Coimbra’s University Hospital offered free medical care, attracting admissions, including of sufferers of chronic disease and unmarried pregnant women, from Coimbra’s rural hinterland. These hospital deaths must stripped out before the legitimate investigation of relationships between city environmental factors (including household- and individual-level features like access to piped water) and (disease-specific) mortality. The study behind my poster establishes the identities and places-of-residence of some 14,700 people buried in Coimbra’s cemeteries. Since about a third of these died in hospital (meaning the deceased might not be a resident), the study cross-referenced burial records with parish deaths and baptisms and with 60,500 hospital admissions (4,700 eventuating in death). Findings show a roughly 3:2 ratio of male to female admissions, a higher incidence of female hospital deaths (8.5% to 7.4%), and a relatively higher proportion of women (43.5% to 31.4%) checking in from urban parishes.

Email: Mafalda Moura Pereira:

The role of demographic histories in explaining the gender gap in job satisfaction
Elena Mariani, London School of Economics

Gender differentials in job satisfaction have received great attention by previous research. The determinants of the gender gap in job satisfaction have been sought among working conditions and different values attached to them by men and women. The role of demographics in these studies has been limited and restricted to controlling for variables such as marital status and number of children. There has been no attempt to try to explain gender differentials in job satisfaction considering the contribution of demographic histories. The present study evaluates the importance of job satisfaction in an analysis of family formation processes. In particular, I use longitudinal data from Germany and the UK to investigate the dynamic effect of first marriage on job satisfaction for men and women separately by explicitly considering the effect of pre-marital cohabitation. The results suggest that marriage has a different impact on the working lives of men and women and such impact depends on the policy environment considered.

Email: Elena Mariani:

A quiverfull of sons : U.S. Protestant sectarian pronatalism since 1985.
John McKeown, JRI

Some U.S. Protestants advocate high fertility. Patricia Goodson's survey data indicates that a significant minority within Evangelicalism believe the Bible prescribes large family size. Academic (and media) attention has focused only on the extreme variaties which exhort early marriage and unlimited fertility. However, “large but limited” pronatalism commending 3-6 offspring is widespread: for example Southern Baptists, the largest U.S. denomination, in 2009 urged their sixteen million members to greater fecundity. This paper is based on analysis of seventeen U.S. Protestant natalist sources. It identifies ten arguments. The extrinsic reasons given for high fertility
are to aggrandize the sect, strengthen America, and stimulate the economy. Intrinsic reasons include divine command and sovereignty, penitential discipline, and natural law. The paper finds that unlimited and limited natalists use similar arguments but differ on the question of planning. It also considers natalist responses to criticisms by environmentalists, on the grounds that the U.S. national footprint has overshot its biocapacity, and that the USA has more births than deaths, for example 4.25 million births and 2.47 million deaths in 2008. Protestant natalism is significant demographically because in the USA not only socioeconomic but also cultural factors influence differences in fertility, and religiosity is an independent variable (Lehrer; McQuillan; Hayford and Morgan). Experiments by Richard Hornok with natalist preaching suggest it produces a "significant attitudinal shift" toward a perceived religious obligation to increase reproduction. This paper concludes that given its location in the USA, with its high per capita footprint, natalism is unsustainable.

Email: Dr. John McKeown:

The impact of demographic changes on infectious diseases transmission and control in middle/low income countries (The DECIDE project)
Alessia Melegaro, Bocconi University; Piero Poletti, Bocconi University; Piero Manfredi, Pisa University; Stefano Merler, FBK, Trento

Population structure and change and social contact patterns are major determinants of the observed epidemiology of infectious diseases, including the consequences on health. Demographic structure and the components of demographic dynamics are changing over time and substantially differ within countries and most critically between countries. However, some of the overall consequences of demographic changes remain unclear, though urbanisation and fertility decline will certainly have a profound impact on social structures, family composition and, as a consequence, on disease spread and on the identification of effective public health measures. DECIDE will explore the following questions: 1. What are the major short- and medium-term impacts of demographic changes on the patterns of infectious disease? 2. How are these demographic changes affecting contact patterns that are of fundamental importance to the spread of infectious diseases? Are there new and different modes of transmission within and between populations? 3. What are the implications of demographic changes? What is the interplay between demographic changes and public health policies in shaping future trajectories of infectious diseases? In order to answer these questions, DECIDE will: 1) analyse large databases of harmonised demographic and health survey data (DHS, HDSS); 2) develop new estimates of social contact patterns and other socio-demographic variables collecting data from representative samples from both urban and rural settings of selected countries; 3) develop a theoretical framework to predict the likely chains through which demographic change influences the burden of infections; 4) develop and parameterise population models for the transmission of infectious diseases to evaluate the impact of public health measures under changing demographic conditions.

Email: Alessia Melegaro:

Improvements Delivered in Migration Reporting
Peter Morgan, Office for National Statistics

Migration statistics are of considerable interest to a wide audience including government departments, local government, academia, journalists and interest groups. Historically migration reporting has been one of the most widely criticised aspects of National Statistics. Criticisms have included: 1. Three separate Government Departments (Home Office, Department for Work and Pensions and ONS) all reported on migration at different times. 2. It was impossible for any but the most expert user to make sense of the different stories that the statistical (International Passenger
Survey) and various administrative sources seemed to be indicating. The Migration Statistics Unit at ONS have improved migration reporting during 2011 to make it much better focussed towards its audience. This has included restructuring the quarterly “Migration Statistics Quarterly Report”, improving the way that data are presented and its accessibility, as well as developing an interactive visual tool showing migration trends over time. The poster will display these improvements to migration reporting for conference delegates to view.

Email: Peter Morgan:

Quantitative Shrinkage: Numerical Data and Manchester’s History of Decline
Fernando Ortiz Moya, The University of Tokyo

The world population keeps growing, accompanied by unprecedented urbanization, with more people living in cities rather than rural areas. This urban growth is leading to the appearance of new urban expressions including megacities, global cities or ‘global cities regions’. However, the opposite trend, that of urban shrinkage, is happening simultaneously. Cities in the pioneer countries of the urbanization process have started to lose inhabitants and economic activity. This unprecedented situation is leading to equally important urban and social problems which have remained unexplored to date. This poster investigates how the analysis of numerical data and its
graphical representation help to notice urban shrinkage. It focuses on Manchester, where shrinkage processes are affecting the whole metropolitan area. Based on an extensive analysis of Greater Manchester’s census data since 1800, the phenomenon of shrinking cities is studied from a numerical lens, linking it with the socioeconomic processes which have shifted Manchester from the industrial capital of the Nineteenth Century to the canonical example of de-industrialized city. This analysis takes into account the rest of the districts of Greater Manchester, not only the core city, providing information district by district and of Greater Manchester as a whole. In doing so, a
wider perspective is obtained, which illustrate the reach of the shrinkage process. This analysis will also help the major actors in the processes of governing and constructing cities to gain a better historical perspective, in order to design strategies to solve the problems derived from it.

Email: Fernando Ortiz Moya:

Distortions in age reporting and their effect on fertility transition in Pakistan
Jamal Abdul Nasir, Andrew Hinde and Sabu S. Padmadas Division of Social Statistics & Demography Faculty of Social and Human Sciences University of Southampton

During the past five decades, Pakistan experienced a slow-paced change in the level of fertility. The recent data show that fertility rates in Pakistan have remained stagnant for the last half decade. The time between 1988 and 2002 is identified as the best duration (fastest as compared to other periods) in the fertility transition of Pakistan. During this time period, the decline in fertility was about 1.4 births per woman. This paper investigates effects of distortions in age reporting for determining the status of fertility change in Pakistan since 1990s using the information collected in three cross sectional surveys namely: 1990-91 PDHS, 2000-01 Pakistan Reproductive Health and Family Planning Survey and 2006-07 PDHS. This paper also presents a new method to adjust for inaccurate age reporting for fertility estimation. The results reveal that accurate age reporting though based on different methods among married women is very poor. The pattern of age heaping as a summary measure and in digit specific manners is almost similar. The most observed pattern of digit preference was: 0, 5, 8 and 2. Digits 1 and 9 were identified to be the most avoided digits. The fertility rates estimation depends on the accurate age distribution. The proposed method namely the method of digit shifts is used to estimate the fertility rates adjusted for age misreporting. In 1990, adjusted total marital fertility is 0.4 births per woman higher than the unadjusted total marital fertility. The fertility estimates adjusted for age misreporting were higher than was the case with the unadjusted rates which leads to very interesting implications. We hypothesise that, over time, fertility transition in Pakistan is not stagnant instead its pace is very slow particularly when distortions in age reporting is incorporated.

Email: Jamal Abdul Nasir:

The SES gradient in infant mortality in Latin America: a success story?
Andrés Palacio, Lund University and University Externado de Colombia

Data from 4 Latin-American countries are used to test the impact of socio-economic status (SES) on infant mortality over time. Bolivia, Dominican Republic and Perú belong to a cluster of late beginners of the modern demographic regime, which are catching up with Colombia, initially the closest to the more modern Latin American countries such as Chile or Costa Rica. Using a pooled sample of Demographic and Health Survey data between 1985 and 2010, the study confirms that income and education drive the relationship between SES and infant mortality. Furthermore, the
relationship is similar enough to think that the gradient is declining across these four countries, contravening recent studies claiming the opposite. On the other hand, interactions between income and education show that the effect of income does change with education: an educational gradient in the richest quartile is confirmed in the pooled and separate samples by country. An implication is that education matters, regardless of income level, and is the key to the success of health care reforms. In short, the convergence to developed country standards relies more than ever in
educational policies.

Email: Andrés Palacio:

Community and household socio-economic influences on dietary intake and obesity in South African adolescents
Rebecca Pradeilles, Centre for Global Health and Human Development, School of Sport, Exercise, and Health Sciences, Loughborough University; Emily Rousham, Centre for Global Health and Human Development, School of Sport, Exercise, and Health Sciences, Loughborough University;. Shane Norris, Birth to Twenty Research Programme, Department of Paediatrics, University of the Witwatersrand; Paula Griffiths, Centre for Global Health and Human Development, School of Sport, Exercise, and Health Sciences, Loughborough University

Childhood obesity is recognized as a public health concern at the global level and is increasing rapidly in Low and Middle-Income Countries (LMICs) especially in urban areas. Childhood obesity is associated with adult mortality and morbidity. The causes of obesity are understood to be multifactorial. The conjunction of poor dietary patterns and insufficient physical activity leads to obesity. Modifications in food consumption and in physical activity patterns are linked to environmental and socio-economic (SE) factors at the individual/household or community level. The importance of the community environment in the energy balance has not yet been investigated in LMICs. Food environment in the community is an important element in the causal chain of
obesity. Few studies have evaluated the associations between community food environment and adolescents’ dietary intakes and obesity in LMICs. South Africa has undergone significant SE changes since the deregulation of Apartheid legislation and is experiencing health and nutrition transitions which are reflected in the rising prevalence of overweight and obesity. In view of these rapid transitions understanding the social context of health is important. This poster will present a review of the results from the literature on the community socio-economic factors that may be associated with adolescent obesity. Particular emphasis will be given to the community influences
on dietary intakes and how this associates with overweight or obesity. Proposed analyses to be undertaken with the 1990 Johannesburg/Soweto born Bt20 cohort will be outlined to further understanding of relationships between community SES, diet, and risk for obesity in LMICs.

Email: Rebecca Pradeilles:

Seasonality of neonatal, infant and adult burials at Christ Church, Spitalfields, 1750-1839
Emily Rousham, Loughborough University; Louise Humphrey, Natural History Museum

Background and aims: The mortality rates of the population of London in the eighteenth century were among the highest known throughout Europe (Landers 1990). Previous studies have reported seasonal fluctuations in mortality in London reflecting the changing prevalence and incidence of infectious diseases. This study examines the monthly distribution of burials in the parish of Spitalfields in order to infer possible patterns of infectious disease mortality. The study also aims to identify whether patterns of burial seasonality differed between infants, young children and adults.
Methods: The monthly distribution of burials was examined from burial records from the parish of Christ Church Spitalfields, London across three 10 year cohorts: 1750-59, 1790-99 and 1830-39. Data were transcribed from parish registers which provided the age, name and date of burial for individuals buried in the church grounds. Results: The number of neonatal and infant burials peaked in March and in late summer (August/September). In contrast, burials of children age 1 to 4.9 years were highest during the winter months, peaking in October/November (in 1750-59) or January (1790-99 and 1830-39). Adult burials consistently peaked in January and February in all
three cohorts. Discussion: The burial records from Spitalfields indicate that infants had a different pattern of seasonal mortality to older children and adults with a marked increase in infant burials in late summer. These seasonal burial patterns may reflect different causes of mortality relating to changing susceptibility to infectious diseases across the life course.

Email: Dr. Emily Rousham:

Kin Influences on Fertility in Thailand: Effects and Mechanisms
Kristin Snopkowski, Rebecca Sear, London School of Hygiene & Tropical Medicine

It has been suggested that human mothers are cooperative breeders, as they need help from others to successfully raise offspring. Studies working under this framework have found correlations between the presence of kin and both child survival and female fertility rates. This study seeks to understand the proximate mechanisms by which kin influence fertility using data from the 1987 Thailand Demographic and Health Survey (DHS), a nationally representative sample of 6,775 women. Kin influence is measured by the length of time couples live with the husband's or wife's parents after marriage. Event-history analysis, multilevel modeling and structural equation modeling are used to investigate both fertility outcomes and potential pathways through which postnuptial residence may influence fertility outcomes, including employment status, maternal and child outcomes, contraceptive use, breastfeeding duration, and age at marriage. We show that living virilocally (with husband's kin after marriage) increases total fertility by shortening time from marriage to first birth, and increasing the likelihood of progression to each subsequent birth. These effects are mediated through correlations between virilocal residence and earlier age at marriage as well as delayed initiation of contraceptive use. We find no influence of husband's kin on maternal or child outcomes. Living uxorilocally (with wife's kin after marriage) also reduces age at marriage, shortens time from marriage to first birth and (marginally) improves child survivorship, but has no effect on other child and maternal outcomes or progression to subsequent births and results in a similar number of living children as women living neolocally.

Email: Dr. Kristin Snopkowski: 

Urbanization and Food Security: The Role of Human Development in a Post-Malthusian Perspective.
Sylvia Szabo, University of Southampton 

 This research investigates the effect of urbanization on food security considering the human development dimensions and critically reflecting on the Malthusian debate. The problem is conceptualised based on the hypothesis that urbanisation effect on food insecurity risks varies significantly across countries by the level of human development. The results based on the analysis of aggregate data on 173 countries, with food deprivation intensity as response variable, show that annual urban growth has a strong negative impact on food security, while the proportion of urban population has a positive association with food security. Evidence suggests that developmental factors, such as income and mortality, have moderate effect on the associations between urbanization and food security. The paper concludes by putting forward a set of key recommendations for policy development and programme intervention.

Email: Sylvia Szabo:

Fertility intentions and the Great Recession
Maria Rita Testa, Vienna Institute of Demography; Stuart Basten, University of Oxford

The onset of the Great Recession in 2008 has had a profound, yet differently felt impact of the society and economy of Europe. In 2011, Sobotka et al. published a paper discussing the effects of economic recessions on fertility in the developed world. They concluded that ‘in most countries, the recession has brought a decline in the number of births and fertility rates, often marking a sharp halt to the previous decade of rising fertility rates’. In order to examine in greater depth how the Great Recession has impacted upon fertility, we can employ the latest round of the Eurobarometer survey package concerning fertility intentions and ideals in 15 EU nations. We have access to data from three rounds, namely 2001, 2006 and 2011. In this paper, therefore, we will examine how overall fertility intentions in Europe have changed since 2001; the extent to which changes have been uniform across Europe; whether men and women are responding differently to the Great Recession and how ideals might translate into ‘reality’ by examining intended future number of children at Parity 1. We will then attempt to explore what are the most likely predictors of changing fertility ideals/intentions across Europe between 2006 and 2011, with a focus on unemployment at different age groups. Finally, we will compare the European evidence with data from East Asia and the United States in an attempt to identify a broader conceptual framework of the relationship between recession and fertility intentions, itself an important mediator to overall pTFR.

Email: Dr. Stuart Basten:

Flexible working, gender equality and fertility
Gesine Tuitjer, University of Osnabruck/MPlfG Cologne

This poster investigates the relation between labour market flexibilization and fertility levels across the EU-27 plus Norway and Switzerland. During the last decade, the labor market experienced tremendous change and so did the family. Internal and external flexibility are linked to issues such as gender equality in care work, the reconciliation of work and care, employment chances for women and precarious employment. These effects influence the fertility level. A positive and significant correlation can be established between internal flexibility and the reconciliation of work
and care, therefore exerting a positive influence on fertility. In contrast, forms of external flexibility such as temporary contracts are correlated to precarious working conditions, poverty and a feeling of insecurity which translates into low fertility levels. Four different production-reproduction regimes can be distinguished, combining labour market patterns and care-arrangements. Special attention is given to the Continental welfare states with the Netherlands, Belgium and France as three different routes to above average fertility levels.

Email: Gesine Tuitjer:

Basic conditions for the post-communist youth generation: a case of Lithuania
Vytautas Valatka, Vilnius University

During 2011, youth unemployment in Lithuania was the 3rd highest in the EU. Various ways have been used to understand the youth unemployment and emigration phenomena. Most research based on opinion surveys, prefer subjective or market based explanations: "low attractiveness of the country", "inadequate match between the qualifications of young people and labor market demands, lack of practical experience, and the low wages offered", etc. This paper analyses and describes the structural conditions which are usually left implicit. 2001 Population and Housing Census data was used for the research. This approach provides four explanations for the current youth state. The first one is that post-communist housing does not meet current society needs - relatively small flats are abundant and many of the owners are unable to renovate. The second one is closed, single cultural society. After the Second World War no former inhabitants were left in the cities. Additionally implementation of continuous development scheme (based on Christaler ideas) during the Soviet times, meant immigration from the Soviet Union was resisted. Today this kind of urban system is unstable. The third reason is misalignment between demographic structure and the state of the economy – youth and members of the baby boom generation have been effected by the 2008 economic crisis. There is a surplus supply of workers in a shrinking job market. Finally, during the period 1989-2001 40% of jobs were lost. It was challenging to constantly adapt to this kind of shock and many children took lessons from their parents.

Email: Vytautus Valatka:

Gender differences in Alcohol and Tobacco use among University Students studying Sport and Heath Science related subjects
Veronica Varela-Mato, Stacy Clemes, Loughborough University

BACKGROUND: University students become less healthy in their independent living. To analyse the level of tobacco and alcohol consumption across both genders of university students studying Physical Activity and Sport Science (A) and Primary Teaching: Specialising in Physical Activity (B), was the aim of this study.
METHODS: A random sample of students (n=179, A=134, B=45) from Vigo University, in Spain, completed a self-reported questionnaire, where alcohol and tobacco consumption where stated on a weekly and daily basis, respectively. Data was analysed using an independent t-test.
RESULTS: No significant differences were found, when analysing alcohol consumption between groups A and B. However, a significant difference was established for Tobacco use (t(173)=2.91, p=0.030). 21% of group A (M=1.36,SD=0.83) reported tobacco use, and 18% (M=5,1, SD=4.73) of this group stated daily use. In comparison, 35% of students in group B (M=6.4, SD=4.48) reported tobacco use, but only a 7% were daily users. Men from group B showed a greater use of tobacco, consuming 8.60 cigarettes/day (SD=7.13), than men from group A (M=4.83,SD=3.99). Women from the same group showed a higher daily tobacco use of 9.92 cigarettes/day (SD=4.90) than women from group A (M=6.46,SD=4.18). But no significant differences were found between genders.
CONCLUSION: The overall percentages for alcohol (A=81%, B=87%) and tobacco consumption (A=21%, B=35%) showed that university education seems to be insufficient to achieve a healthy standard of living, even when they undertake a health-related degree.

Email: Veronica Varela-Mato:

Do cues of neighbourhood mortality and morbidity affect people’s approval of abortion? Testing hypotheses from life history theory
Sandra Virgo, Department of Population Health, London School of Hygiene and Tropical Medicine

Both demographic transition theory and evolutionary life history theory have proposed that fertility responds to changes in mortality. Psychological mechanisms causing individuals to lower their fertility with decreasing mortality might mediate this relationship: several psychological experiments have found that people faced with cues of high mortality express increased fertility preferences. UK research at population, community and individual level has also shown that socioeconomically deprived people are both less likely to terminate pregnancies and more likely to
disapprove of abortion than wealthier people. This might be partially mediated by health disparities, whereby differential perceptions of likely mortality and morbidity contribute to an awareness of reduced actual/disability-free life expectancy for poorer people, and accordingly decreased motivation to terminate pregnancy. The current experiment tested the hypothesis that mortality and morbidity cues influence attitudes towards abortion and childbearing. Participants were randomly allocated to a) mortality cue b) morbidity cue and c) control conditions. In mortality and morbidity conditions participants read fake newspaper articles on increased neighbourhood death via homicide and accidents, and increased neighbourhood chronic illness and functional
limitation, respectively. The control condition was a newspaper article about increased
neighbourhood non-violent crime. There followed questions on attitudes and
background/socioeconomic characteristics. Dependent variables were 1) approval of abortion and 2) motivation to invest in education/career before starting a family. We tested (1) whether main effects of the experimental conditions were found and (2) if these effects interacted with childhood and/or adult socioeconomic status. Results will be discussed in the light of life history theory.

Email: Sandra Virgo:

Beyond 2011: Who do we count and where?
Jennifer Wall, Pamela Dent, Office for National Statistics

The UK Statistics Authority formally established the Beyond 2011 Programme on 1st April 2011 to assess options for meeting future user needs for population and small area sociodemographic statistics in England and Wales. There are a number of population bases for which estimates may be produced. ONS currently uses the usual resident population definition to produce mid year population estimates. The Beyond 2011 Programme provides an opportunity to take a fresh look at the requirements for producing estimates of the number of people, households and families. This poster will explore the different ways in which we could define these concepts in a modern society.

Email: Jennifer Wall:

Maternal work-family interface and maternal wellbeing: a case study of Accra, Ghana
Philippa Waterhouse, University of Southampton

Female labour force participation in Ghana is relatively high due to the increasing requirement of women to contribute to the household economy. The effect of fertility on female economic activity has been long debated. However, results from the Women’s Health Survey for Accra (WHSA) project indicate that children only have a small influence on female labour force participation in the Accra Metropolitan Area (AMA), with women returning to work soon after giving birth (Hill et al 2010). This implies that many mothers with young infants are experiencing multiple role responsibilities. The links between multiple roles and maternal and infant outcomes are complex with both negative and positive impacts expected according to factors such as conditions of work, the household division of labour and the availability of social support (Vaydanoff 2002). Yet, despite the hypothesised importance of these features little information exists in the sub-Saharan African context (Mehra et al 2003).
The present research aimed to gain a greater understanding of the circumstances of mothers with young infants in the AMA. Using qualitative methods the subjective experiences of conflict and enhancement of maternal multiple roles were uncovered and the central importance of work to family life highlighted. Whilst the majority of women reported tiredness, they also felt they were successfully managing their family and work lives. The adoption of work and childcare arrangements by the age of the infant, it seems, is an effective strategy in protecting offspring against the possible adverse consequences of maternal employment.

Email: Philippa Waterhouse:

Overview of 2001-2011 Census question comparability for England and Wales
Karina Williams, Office for National Statistics

This poster explores the impact of 2001 and 2011 question changes on the data that will be Presented in 2011 Census outputs. An overview of question differences between 2001 and 2011 for England and Wales will be provided, and the extent to which these questions are comparable will be discussed. Details of the impact of question changes on data produced, and subsequent outputs, will be described.

Email: Karina Williams: