Go to content Go to menu

10th Demographic Conference of "Young Demographers" – Presentations and online discussion forum

List of abstracts – oral presentation


Naser ABUSRIHAN,1 Khawla ZOABI,2 Jon ANSON 3

1,3 Ben-Gurion University of the Negev

2 The Arab Academic College for Education in Haifa, Israel

Fertility decline in a rapidly changing society: The Bedouin in the Negev (South of Israel)

Bedouin fertility, which used to be amongst the highest in the world, has dropped dramatically over the past two decades, from over 10 children per woman to less than six, and continues to decline gradually. We interpret this rapid decline, and the current relative stability, in the light of changes in Bedouin society and, in particular, the role of women in that society. In the past 70 years, Bedouin society in the Negev (southern Israel) has been reborn and recreated along familial and tribal lines, which have offered a first line of defence against poverty on the one hand, and the State on the other. However, over the past decades, three process have begun to undermine this family-centred, patriarchal structure.

1. The rapid growth of population, from little over 10,000 in 1951 to almost 120,000 fifty years later, and 250,000 today.

2. The education of the younger generations and the displacement of the traditional elders by younger generations

3. The growth of female education and employment outside the home.

Using data from a recent survey of Bedouin households, as well as ethnographic data from women discussing their view of childbearing, we shall present an interpretation of the decline in Bedouin fertility as a strategic change, which is still unfolding. Educated, younger women appear as the vanguards of this change, but it is permeating all levels of Bedouin society.

Keywords: Bedouin; Israel; Fertility; Patriarchy; Family





Cristina AVRAM 1

1 Charles University

Demographic transition on a regional level in the Czech Republic

Demographic transition has an important role in the socio-economic progress. Its influence on development of the society continues until today through the changes of population size and structure. However, the demographic transition did not take place uniformly in the whole country, but with regional particularities.  Therefore, the aim of the paper is to identify the differences in the demographic transition in the Czech Republic on the regional level, and to examine their correlation to ageing.

Census (1869 - 2011) and vital statistics (1872 - 2011) data was used for the research. The historical data was recalculated to the 2011 territorial administrative divisions, for comparability reasons. The calculation of indicators was preceded by the reconstruction of maps (1869 - 1910) due to the unavailability of historical maps. The indicators used for the analysis of demographic transition were: crude death rates, crude birth rates, infant mortality rates, Coale indices, and crude marriage rates.

The results of the paper will contribute to the examination of the relation between the onset and pace of demographic transition on the tempo and quantum of ageing on the regional level in the Czech Republic.

Keywords: Historical demography; Demographic transition; Regional level; Ageing



1 Stockholm University

The Effect of Resident and Non-resident Children on Men’s and Women’s Chances of Re-partnering after the Separation of the First Childbearing Union

The majority of people re-partner after the dissolution of their first childbearing union. In this study we examine how the presence of children and the residential arrangement of these children influence the chances and timing of re-partnering (and remarriage). We address three research questions: (1) Is there a gender gap in re-partnering? (2) Can the gender gap be attributed to the presence of children from prior unions? (3) How does residential arrangement of children affect chances of repartnering of men and women? We use data on the 11% sample of the total population from the Finnish longitudinal registers.  We employ discrete-time event-history analysis to examine gender differences in re-partnering and re-marriage after the dissolution of the first childbearing union. The analysis relies on childbearing events and cohabitation and marriage histories and covers the years between 1987 and 2012. The results show that the gender gap in chances of repartnering is diminishing over time. Our models reveal that childless separated and divorced men and women do not differ in their re-partnering chances. Women with resident children are the least likely to re-partner, and non-resident children lower their re-partnering chances as well.  The re-partnering chances of men don’t seem to be affected by non-resident children, and although the effect of resident children is negative, it is much smaller compared to women with resident children. Our study shows that the presence and residential arrangement of children are the key factors in both chances and timing of re-partnering and remarriage.

Keywords: Re-partnering; Remarriage; Resident children; Gender gap


Nava DIHI,1 Jon ANSON 2

1,2 Ben-Gurion University of the Negev

The social structuring of mate selection: Assortative Marriage Patterns in the Israeli Jewish Population

How blind is love, and how free are we to choose our partners? We show that partner selection is socially constrained, and the freedom to choose is limited by at least two major factors or capitals: on the one hand, material resources and education, locating the partners on a scale of personal achievement and economic independence. On the other, the partners' ascriptive belonging to particular ethnic, or origin, groups, differentiated by the groups' social prestige, as well as by their culture, history and even physical characteristics. However, the relative importance of achievement and ascriptive factors, as well as the overlap between them, varies from society to society, depending on the society's structure and the factors shaping it. Israeli social structure has been shaped by the waves of new immigrants who arrived over the years. The timing of their arrival, their patterns of physical settlement and their occupational inclusion or exclusion have together created a mosaic of social groups whose principal common feature has been the country of origin from which they arrived. The analysis of marriage patterns helps illuminate the social meanings of the groups and their borders. To the extent that ethnic group membership has meaning for individuals and influences their life choices, the ascriptive factor will gain in importance relative to the achievement factor in their choice of marriage partner.

In this research we examine Jewish Israeli marriage patterns by looking at the marriage choices of 5,041 women aged 15 to 49 who were single at the census in 1983, and who were married at the time of the 1995 census, 12 years later. The database for this study was a file linking respondents from the 1983 and the 1995 censuses. In both cases, 5 percent of household were randomly

chosen, so that our sample includes about 4 percent of women in Israel in 1983. We present three basic analyses:

1. Who was still single in 1983, using personal and household data from the 1983 census (binomial model)

2. Who married between 1983 and a1995, using personal and household data from the 1983 census (binomial model)

3. What were the personal characteristics of the womens’ partners in 1995, using data from the 1995 census (loglinear model)

We show:

1. that material and cultural capital both operate to delay marriage and to increase the probability of remaining single; and

2. while there is a clear association between ethnic group membership and education, endogamy and homogamy both operate as separate forces which constrain (but do not determine) the choice of marriage partner, and thus both serve to reproduce the current pattern of relationships, as well as identifying patterns of proximity and distance between the different groups.

Keywords: Israel; Nuptiality; Homogamy; Endogamy



1 Tallinn University

After a Child Died: The Changing Fertility Response During the Course of the Demographic Transition

The fertility transition has long been seen as a response to falling childhood mortality. Although contested using aggregate-level data, the relationship has consistently been proven using individual-level data and event-history methods. These studies, however, have been mostly done using Cox proportional hazards model. This is problematic since the assumption that everyone will eventually experience the event of interest is not held with fertility. Also, the Cox model is unable to disentangle risk (stopping) and duration (spacing) effects. Here a cure mixture model is used that does not have these faults. Thus, the main research questions are: i) did the death of a previous child raise the propensity of a subsequent birth; ii) did it shorten the time to next birth; iii) did these effects grow larger during the demographic transition as conscious parity-specific family limitation became dominant. As a dataset the population registry of Estonia is used and the reproductive histories of women born 1850-1899. The results prove that the death of a child raised the propensity for having another birth and this effect grew for later cohorts. The effect on the timing of next birth was, as expected, shortening. But this effect decreased in later cohorts. This means that the increase in deliberate fertility behaviour during the transition led to the increase of parity-specific aims and that dead children were replaced. On the other hand, this increase in agency meant that couples also controlled the birth intervals more and planned the next birth for a suitable time.

Keywords: Infant mortality; Child mortality; Fertility; Cure model


Sijie HU 1

1 London School of Economics and Political Science

Descendants over half a millennium: marital fertility in five Chinese lineages, 1400-1900

The paper studies marital fertility of five Chinese lineages in Zhejiang and Jiangsu Region from 1400 to 1900. The paper mainly deals with three research questions: how high exactly was the Chinese marital fertility from 1400 to 1900? Was it stable over time? Was it a 'natural fertility' regime or did the Chinese control fertility within marriage? By exploiting new genealogical data of detailed records on individual births, deaths, and marriages, and studying more than 40,000 individuals in the five lineages, a unique marital fertility pattern is reconstructed and reexamined. The measures used in this paper to estimate the marital fertility levels consist of both standard fertility measures relative to females, age-specific marital fertility rates and total marital fertility rates, and also a non-standard fertility measure relative to males, namely net fertility of married males. The paper also runs a new test to examine the existence of the early stopping behaviour in marriages, a type of parity-dependent control. On the one hand, contrary to conventional wisdom on Chinese fertility, the results show that the marital fertility rates in the period were much lower compared to those of Northwest Europe in similar periods. On the other hand, in line with the classic ideas, the paper finds no signs of parity-dependent controls within marriages, which proves that Imperial China was still largely a "natural fertility" regime.

Keywords: Chinese historical marital fertility; Chinese genealogies; Parity controls



Kateřina MALÁKOVÁ 1

1 Charles University

Malignant neoplasms mortality: analysis of potential factors of level and structure of mortality due to malignant neoplasms in the EU focusing on Czechia

Malignant neoplasms diseases, alias cancer, are becoming a major problem, not only in developed countries. That diseases are the second most common cause of death (following circulatory diseases) and relative share of them among the causes of death are increasing over time. Nowadays, cancer plays an important role in current and future population development and population health.

This study is concerned with describe trends and the current situation of the level and structure of mortality from malignant neoplasms and also on research of possible factors (such as lifestyle, level of prevention, availability and quality of health care) that may be related to the incidence of neoplasms and the mortality from neoplasms. The studied area is the European Union (EU) with a focusing on Czechia in the latest years.

The results of the analysis show significant differences between countries of the EU due to many studied factors. Simultaneously, it has been found that European countries, which are geographically or culturally close, evince similar patterns of behaviour and mortality due to malignant neoplasms than mutually distant and dissimilar countries. This is also substantiated on Czechia, where closeness to neighbouring countries is ascertained.

Keywords: Malignant neoplasms; Mortality; Risk factors; European Union; Czechia




Collet MUZA, 1 Kudzaishe MANGOMBE 2

1 Charles University

2 University of Zimbabwe

Population development in Africa with special regards to ageing

The purpose of this study is to examine regional population ageing differentials and similarities in Africa. Population ageing is a global trend of the 21st century and has got huge socioeconomic implications. Demographically, Africa is world second largest and poorest continent, yet Africa’s population ageing is not fully understood. Africa is often characterized as a homogenous region which is not useful in terms of policy formulation. Selected indicators and determinants of population ageing were taken from World Population Prospects (UN 2017) for periods 1995-2000, 2010-2015 and 2025-2030. We applied k-means method of clustering to standardised ageing indicators to group countries according to their heterogeneity and homogeneity. Tukey’s (HSD) post-hoc test was used to validate the cluster analysis differences. Three clusters were identified for each respective analysis period with high, medium and low population ageing. The results reveal that majority of countries 58%, 50% and 54% were at low, medium and medium levels of population ageing in 1995-2000; 2010-2015 and 2025-2030 respectively. In 2010-2015, the increase of countries in cluster 2 (medium) and simultaneous decrease in cluster 1 (low) represents population development associated with demographic transition process. Throughout all analysis periods Northern and Southern African countries (13%) formed a relatively constant homogeneous spatial unit with highest population ageing. Each cluster form a spatial unit in which appropriate policy measures can be formulated and implemented. We recommend that each spatial regions with similar ageing demographic profiles to share experiences and knowledge in line with the Agenda 2030 for Sustainable Development Goals.

Keywords: Demographic revolution; Ageing; Regional differences; Clustering



1 Institute of Atmospheric Physics CAS, Prague, Czech Republic

2 Department of Demography and Geodemography, Faculty of Science, Charles University, Prague, Czech Republic

Do we have to worry about cosmic-origin background radiation?

Nowadays the cosmogenic radionuclides in our environment are likely play a greater role than in the last decades. The concentration of cosmic radiation and cosmogenic radionuclides reaches its highest values during periods of low solar activity when the Earth’s magnetosphere is more penetrable. We evaluate the lifetime attributable risks induced by an increasing concentration of cosmic radiation and cosmogenic radionuclides during the long solar minimum in 2020–2040 for the specific conditions in the Czech Republic.

We focused on quantifications of the relationship between radiation exposure and the resulting potential risk of carcinogenesis or cancer mortality in the Czech Republic. Three different scenarios based on dose radiation level were explored. This analysis was performed separately for men and women exposed to radiation at age 10 and age 30. The total annual dose of natural background radiation was then used to model different scenarios of the potential increase in the lifetime cancer risk. The possible effects of increased radiation impacting the lifetime attributable risks were computed for three different levels of radiation (2.71; 2.85; 3.3 in mSv).

We present estimates of the lifetime attributable risks of incidence and of mortality for three solid tumor cancer sites (colon, lung and stomach), based on a combination of the Excess Relative Risk (ERR) and the Excess Absolute Risk (EAR) models.

Keywords: Solid cancer mortality and incidence; Lifetime attributable risk; Age at exposure; Solar activity; Natural background radiation dose



1 University of Cambridge

Infant mortality decline in London, 1891-1911: an exploratory spatial analysis of patterns and determinants.

Understanding the mortality transition in England and Wales from the mid-nineteenth century onwards remains a fundamental challenge for historical demographers. The decline in infant mortality is particularly peculiar as it lagged behind overall mortality decline by half a century (c.1900). Explaining this later, but rapid, infant mortality decline is of special demographic importance, as infant deaths accounted for a substantial proportion of all mortality. Thus, the improvement in this demographic measure allowed the national life expectancy at birth to rise above 50 years for the first time.

New spatial techniques and digitised Integrated Census Microdata, have allowed detailed spatial analysis of infant mortality patterns and determinants to be undertaken. Key hypothesised determinants were analysed against log-transformed infant mortality rates for London’s Registration Sub-Districts. Specifically, CAR Leroux spatial regression models were fitted for each census dataset (1891, 1901, 1911), to control for spatial dependencies inherent in geographically aggregated datasets.

Mean infant mortality in London declined from 157 per 1,000 live births in 1891, to 116 in 1911. Three key determinants of this demographic measure were identified: 1) Total Fertility Rate; 2) Percentage of Married Women Working and 3) Institutions per 100,000 population. All three of these variables were significantly positively associated with infant mortality in each census dataset model.

Keywords: Infant mortality; Mortality transition; Census data; London; Inequalities


Helena M. REPCZYŃSKA, 1 Wirginia WOJTCZAK 2

1,2 Adam Mickiewicz University, Poznan

Did industrialization change family model? Evidence from the city of Lodz (Poland) in the late 19th and early 20th centuries

It has been emphasized in the literature that industrialization induced women to enter the labour market leading to changes in their reproductive behaviour: postponing matrimony, setting a family later in life and procrastinating reproduction. This work presents the results of the pilot study into the marital fertility in the 19th and early 20th centuries in the city of Lodz. Reproductive histories of women are taken from the municipal registers of Lodz residents. In the early 19th century, Lodz became the most important centre of textile industry in Poland, known for its dynamic urban and industrial development as the "Polish Manchester". Industry accelerated the migration processes, especially from rural areas. Villagers "brought" with them rural traditions. They were accustomed to a different work pattern and to having a large number of children.

In this work the following questions are answered: 1) Did the rapid social and economic changes at the end of the 19th century and early 20th centuries result in blurring the differences in fertility patterns between Lodz inhabitants coming from various social backgrounds? 2) Did migrants from non-industrial areas to Lodz contribute to the “city family model”?

In order to address the above questions the following data were evaluated using the analysis of variance: average ages of mothers at the moment of giving birth, between-birth intervals and the number of children according to social and occupational status, as well as parents’ place of birth. The impact of parents’ place of origin on family size was assessed with logistic regression.

Keywords: Birth intervals; Number of children; Mother’s age at birth; SES; Industrial/ non-industrial family model


Tim RISWICK,1 ChiaChi Xingchen LIN,2 Yinghui HSIEH 3

1 Radboud University

2 TamKang University

3 Tzu Chi University

Between Rivalry and Support: The effect of sibling size and composition on child mortality in Taiwan, 1906-1946.

Household structure and organization have become central issues in social science research mainly because both can have huge implications for social inequality. It determines who is available, who is culturally or socially held responsible, how resources within the household are allocated, and how members of the household interact with each other. Considering the importance of the availability of kin for providing care and other resources to young children, previous studies have mainly focused on the importance of parental support. Yet, other household members may just be as important.

This paper will address sibling effects by investigating infant and child mortality in two different regions in Taiwan, 1906-1946. Although in both regions extended family structures are considered the ideal type of household, this paper choses these regions because they also differ in cultural practices and economic conditions. In other words, this paper aims to take regional differences into account while investigating sibling effects in the sense of rivalry and support. By using reliable longitudinal datasets for Zhu Bei and Dong Gang, this study can take changing household composition through time into account. By doing so, it can overcome shortcomings of earlier studies which investigate historical Asian communities and add to the nuclear and stem household dominated literature by providing insight into sibling effects in a region where extended household structures are the ideal. We expect that especially having older brothers will have a negative influence on infant and child survival, while the historical context may explain the size of the effect.

Keywords: Sibling rivalry and support; Resource dilution; Infant and child mortality; Taiwan




In need for policies and tools - new modeling approach for demographic challenges

The aim of this paper is to provide a technical guideline and to establish a framework for modeling savings decisions with better fit to the data. Due to longevity and declining fertility, there is a great concern if the publicly run pension systems are able to provide a sufficient coverage for the old-age expenses. Moreover, the empirical findings show that the private savings seem to compensate the lower old-age pension benefits only partially. The current OLG models seem to fail to reflect empirical evidence on savings decisions overestimating the compensation by private savings. Even with the development of the imperfectly rational agent models, the literature still tends to focus on using simple heuristics as households’ decision rules in order to fit the empirical evidence better. We want to address this issue while retaining the assumption of households solving an underlying intertemporal optimization problem, to account for non-perfect adjustment of agents to the settings' changes.

We construct an overlapping generations model with intra-cohort heterogeneity, uninsurable income shocks and uncertain life span. To this structure we add a convex transformation of decumulative distributions of all uncertain events, i.e. idiosyncratic shocks and uncertain life span. This allows us to account for many different behavioral aspects of decision making, such as time inconsistency and failure to asses adequately the probabilities of various shocks, while adding very little additional structure. Moreover we discuss other possible modeling choices that could bring results obtained within OLG model closer to the empirical evidence regarding saving decisions.

Keywords: Longevity; Irrationality; Pension systems; Macro modelling; Computable General Equilibrium Models


Torsten SAUER,1 Roland RAU 2

1,2 University of Rostock

Predicting Death Using Random Forests

Machine learning methods have become very popular in various scientific disciplines. Using Breimann's random forests and data from the National Health Interview Survey (NHIS) and its mortality follow-up, we wanted to know 1) Could these methods be used to predict the occurrence of death? 2) Which variables are important for these predictions?

We checked the accuracy of the forests by estimating the area under the ROC curve (AUC) for test data and showed that they perform relatively well, with an AUC from 0.83 to 0.87. Nevertheless, because of the high imbalance of death and survival cases in the test data the precision is very low.

To indicate the predictive power of every variable we estimated the mean decrease in accuracy (MDA). Not surprisingly "age" is by far the most predictive, followed by "mobility limitations" and "self-rated health". Typical sociodemographic mortality determinants like "sex", "education", and "income" seem to be very weak in their predictive ability in each of the six selected intervals after the interview.

Keywords: Mortality; Machine Learning; Random Forest; National Health Interview Survey


Maria Dolores SESMA CARLOS 1

1 Radboud University

Experiences of Internal migration, social structure and old-age mortality: a longitudinal study in the Netherlands (1850-1940)

Studies have found a longevity advantage, but also a socialization and adaptation effect, among rural migrants following a long exposure in urban settings. The influence of diverse migrant trajectories and individual inscription in the social structures for a long-follow up period in old-age mortality is still uncovered, especially during the industrialization period. In the context of The Netherlands during the epidemiological transition, we first analysed old-age mortality differentials associated to sociodemographic characteristics. Second, we studied whether the exposure to the urban environment, captured through diverse migrant trajectories before age 50, had an influence on later-life mortality, controlling for the previously mentioned sociodemographic factors. Micro level longitudinal data was derived from The Historical Sample of The Netherlands (1850-1940) for Dutch individuals ≥ 50 years old  from birth cohorts 1850-90 over a follow-up period of 50 years (n=8,773). Event-history analysis was applied to model old-age mortality over the life span of permanent residents on one side, and on the other side of migrants with diverse exposure time to the urban life, considering the successive birth cohorts and the gender differentials.

Keywords: Mortality; Internal migration; Social structure; Event-history analysis; Old age

Jitka SLABÁ 1

1 Charles University

The effect of unemployment experience to the first birth fertility, Czechia 2008–2016

In the last few decades, a childbearing postponement to the later ages of women has challenged developed countries. As a result, the mean age of mothers at birth has increased. Besides, fertility has decreased below the replacement level in European countries despite the prevailing ideal of a two-child family model. On the macroeconomic level, the economic growth affects a fertility level positively. Therefore, a slowdown in economic development connected with an increase in unemployment rate causes a fertility decrease. On an individual level, the relationship between economic situation and reproductive behaviour is ambiguous. The aim of the submitted paper is to analyse the effect of unemployment experience on the first child fertility behaviour of Czech women in the period of 2008 to 2016 using individual data from the survey “Ženy 2016” (Women 2016; N=1257). Previous analysis of the survey shows that more than 30% of women postpone their plan of first birth, one quarter of them due to material conditions. Current work shows that women who experienced an unemployment more often postpone their first birth. Furthermore, they more often delay their first birth due to material conditions. It seems that individual experience of unemployment affects the first child fertility behaviour.

Keywords: Unemployment; Fertilty; Czechia



1 Masaryk University Brno

Milestones in the lives of young subjects in Moravia in the eighteenth century

In the twenty-first century the stages of childhood and youth are identified quite easily – they are defined by the law. As John R. Gillis points out, the education provides milestones, which divide this period into phases. In the past, however, this division was far from clear. This paper will focus on milestones in lives of young subjects in a Moravian town and city in the second half of the eighteenth century.

First of all, the moment when children left their parents’ home was a turning point. The author will try to find out how old they were and if this age is similar to the rest of Europe. Moreover, not all the children went in the same direction – most of them became servants in households of other subjects, some boys entered apprenticeship and young men were enlisted. Many young people even returned to their parents’ house. Secondly, the end of youth was usually marked by marriage. It is, therefore, necessary to ask not only about the age when young people got married, but whom they married and under what circumstances. Finally, a number of men and women did not enter marriage. In these cases it is much more difficult to determine when or if ever they became adults, i.e. independent members of the community.

The research is based on lists of subjects and orphans from the estate of Slavkov u Brna. The source enables us study to what extent were the phases of youth influenced by the gender, social status, place where the young people were born and age when they were orphaned.

Keywords: Youth; Subjects; Servants; Early modern period


Martina ŠTÍPKOVÁ 1

1 University of West Bohemia

Marital status and social networks of the elderly

Quality of life (not only) at older age is closely related to of individuals’ social networks. The characteristics of social networks include objectively measured indicators, such as number of social contacts or amount of support received from the others, as well as subjectively perceived loneliness. Both objectively being alone and subjectively feeling alone, have a negative impact on physical and mental health and survival. Loneliness is most common among the elderly. The proposed paper focuses on the age group 50+ and analyses characteristics of their social networks in relation to their marital status.

Longitudinal survey SHARE is used. Waves 2011 and 2015 focused on close social network (max. 7 persons with whom respondents talk about personal matters). The objective measures include the number of network members and relation to respondent (relatives, non-relatives). Subjectively perceived loneliness is measured as satisfaction with social network.

Preliminary results of multilevel analysis show that divorced and widowed seniors have smaller social network. Family members prevail in social networks of all marital status groups but divorced (and, to a smaller extent, also widowed) respondents have a higher proportion of non-relative persons. Satisfaction with social network is the highest among married respondents, lower among widowed and the lowest among divorced individuals. Most of these differences can be explained by the objective characteristics of their social networks, especially by the number of relatives in the social network. This suggests that having more friends does not lead to the same protection against loneliness as having family in one’s social network.

Keywords: Divorce; Marriage, Old age; Wellbeing; Widowhood



List of abstracts – posters



1 Department of Demography and Geodemography

Cancer screening coverage among Czech women

It has been proven by multiple studies that cancer screening reduces cancer mortality. In the Czech Republic population-based screening programmes were introduced in 2002 for breast cancer and in 2008 for cervical cancer.  This work focuses on breast and cervical cancer screening coverage among Czech women. Data from Health insurance have been used for calculating age-specific coverage ratios as well as crude and standardized coverage ratios in 2009–2017 (as earlier data were not provided). It has been found that especially the breast cancer screening coverage rates are influenced by age structure and there is big difference between crude and standardized rates. The cervical screening coverage has grown during the reference period. The gynaecologist check-up attendance rates were also calculated and have been compared with cervical screening rates, as cervical screening is a part of regular gynaecologist check-ups and gynaecologists should also send women to the breast cancer screenings every other year.

Keywords: Screening; Cancer; Epidemiology


Andrey FROLOV 1

1 Charles University

Mortality modeling and forecasting: past, present and future

Mortality forecasts are usually based on statistical modeling of historical data. Early models for mortality rates were deterministic, for instance the Gompertz curve (1825), Perks (1932) logistic model (a generalization of the Gompertz curve), Heligman and Pollard (1980) curve. Progress in computational algorithms helped handling complex models, and the number of parameters was no longer an issue.  This trend called for more advanced techniques to manage the mortality problem. In 1992, Lee and Carter presented a stochastic model to fit and predict mortality rates for United States.  There were several reviews and extensions of the basic Lee-Сarter model, particularly Lee (2000), Pitacco (2004), Wong-Fupuy and Haberman (2004), Cairns (2006), Renshaw and Haberman (2003), Renshaw and Haberman (2006) as well as others. Although these researches increased forecasting accuracy of the original Lee-Carter model, it is still a challenging task to forecast unsteady mortality developments.  Bayesian approach to the Lee-Carter model was developed by Pedroza (2006).  As for the most recent technique in mortality forecasting we should name the neural networks and machine learning in general.  Only a few research articles applied neural networks to forecast mortality, namely Atsalakis (2007), who proposed a neural network with fuzzy logic inference, Abdulkarim and Garko (2015) and the most recent paper by Hainaut (2017). The presentation includes further comparative analysis of the methods in terms of their significance. The main goal of the presentation is to conduct a comparative analysis of existing mortality forecasting methods and to identify areas for their further improvement.

Keywords: Mortality forecasting; Stochastic models; Comparative analysis; Review of methods



Oldřich HAŠEK 1

1 Charles University

Demographic processes seasonality from a perspective of web searching engines statistic: an effective helper, or an impasse?

Nowadays, the internet connection is a part of daily life for a major population of developed countries. The internet is for a lot of people a way how to find an answer to different types of questions (e.g. advice, entertainment possibilities or shopping ideas). Internet searching engines gain enormous amounts of information each day. This information is often not fully utilized. Is it possible to use it meaningfully in demography? Data from web searching engines are time detailed. That allows us the time detailed observations of selected demographic processes. For example, data can be used for studying seasonality. The paper assesses the usage of this data in the case of the Czechia and their ability to replace the data obtained from the official statistics.

The paper introduces the relevance of using the obtained data, considering different demographic processes. One of the successful examples of this method of analysis is the use of internet data for the international comparison of the seasonality of marriages.

Keywords: Divorce; Marriage, Old age; Wellbeing; Widowhood


Ladislav KÁŽMÉR 1

1 Institute of Sociology, Czech Academy of Sciences

Smoking, gender and macro-social gender inequalities. Multilevel evidence from 21 European countries.

The aim of the paper is to examine the relationship between the overall level of gender inequalities within society and inequalities in adult smoking rates between males and females.

Data on current smoking prevalence among both males and females, available from a large international European Social Survey (ESS, 2014), were applied. The survey comprised data on more than 40.000 respondents coming from 21 countries. At the same time, recent data on Gender inequality index (GII, 2014) were merged to each country participating in the ESS survey.

Analysis was conducted in two steps. First, age-adjusted within-country differences in adult smoking prevalence rates were related to the national GII. Next, multilevel logistic regression models were conducted to test the cross-level interaction of the effect of gender on adult smoking status with the country’s GII. In multilevel models, the statistical control on a range of individual socio-economic variables was applied as well.

Results of the study point to the significant effect of macro-social gender inequalities on the prevalence of smoking among genders. Last but not least, both theoretical and empirical discussion on the phenomenon is provided.

Keywords: Smoking; Prevalence; Gender inequalities; Cross-national research; Multilevel modelling


Oliwia KOMADA 1


Child allowance as a pronatalist and income stabilization tool

Empirical studies show that economic uncertainty has a large, negative effect on completed fertility. We show that the overlapping generations model with endogenous fertility and uninsurable earnings risk provides a reliable fit to this pattern observed in the data. Moreover, we show that the individual fertility decision may not coincide with the social optimum in such a framework. Suboptimal choice occurs due to the external effects of children and precautionary savings. Consumers override link between private fertility choice and pension benefits level. Earning risk yields to precautionary savings and thus reduces fertility. Internalization of child-related externalities and income stabilization should be the aims of policymakers. We show that child allowance is a proper tool for this purposes. We build a computational model to qualify the effect of child allowance on fertility rate and welfare. We show that child allowance is welfare improving and increase fertility rate. Policies that stabilize income may have a positive effect on fertility.

Keywords: Fertility Choice; Uninsurable Idiosyncratic Income Risk; Life Cycle; Child Allowance


Barbora KUPROVÁ,1 Petr TUREČEK 2

1,2 Department of Demography and Geodemography, Faculty of Science, Charles University

Presence of Grandmothers in the Families at turn of the 18th and 19th Centuries and their Possible Influence on Infant Survival (the Example of Škvorec Manor)

The evolutionary anthropologists formulated the hypothesis (30 years ago) which is called "Grandmother hypothesis". This theory attempts to explain the menopause which is limited the reproductive life of a womenˈs life. Anthropologists show that menopause allows women to support (to look after) of their own grandchildren (Hawkes, 2003). Studies which tested this theory by using the current subpopulations’ data had similiar results: the presence of grandmother corresponds to the lower age of her daughters (or daughters-in-law) at the birth of their first child and their shorter birth-birth intervals and to lower infant (or child) mortality of grandchildren (e.g. Beise and Voland, 2002). What is the situation at the turn of the 18th and 19th centuries? How often did maternal or paternal grandmother live close to the birthplace of her grandchild? And what was the effect on the survival of this child? The aim of this paper is discovered the possible effect of maternal and paternal grandmothers on infant (child) survival at the turn of the 18th and 19th centuries in Škvorec manor.

Keywords: Grandmother hypothesis; Infant survival; Presence of grandmother; 18th and 19th centuries


Mukesh PARMAR 1

1 Jawaharlal Nehru University, New Delhi, India

Sex differential in Life expectancy and Lifespan disparity in India: 1991-2015

Life expectancy (LE) and Lifespan disparity (LD) are two side of the same coin. On one hand LE measures average number of years lived at a particular age group, LD measures the average remaining life expectancy by averting premature deaths. LD declines when deaths are averted at early ages, which compresses the distribution of deaths, and expands when saving lives occur at late ages, which  leads  to  increase  in  the  average  remaining life expectancy. Thus, with decline in LD, LE increases as deaths are postponed. LD is the relatively less researched area in India.

Data is taken from the abridged life tables provided by Sample Registration System (SRS) of India for three decades viz. 1990-95, 2001-05 and 2011-15. LD is calculated from the standard formula given by Zhang and Vaupel (2009). All the life tables are reconstructed from MORTPAK software.

Results show that there is a convergence pattern observed for LE and LD. The convergence is more pronounced in later years of life i.e. 70-75 years and is delaying every year. The pattern of convergence of two curves happens slightly earlier for females than males. Maximum gap in sex differential between the two indicators is observed for age group 15-20 years and 20-25 years of age. Thus, there is a huge inequality in LE and LD in mid-adult age and decreases with increasing age.

Keywords: Life expectancy; Lifespan disparity; Lifespan variation; Remaining life expectancy; India


Vanessa Santos SÁNCHEZ, 1 Gabriele RUIU, 2 Marco BRESCHI, 3 Lucia POZZI 4

1,2,3,4 University of Sassari

Spatial analysis of municipal mortality in Italy (2012-2016)

Despite the increasing attention devoted in the literature to the study of the spatial distribution of mortality, this kind of analysis has been rarely applied to the Italian experience. This work aims to fill this gap, analyzing municipal mortality for the whole country in the period 2012-2016. A transversal study of small areas has been carried out for the 8,000 Italian municipalities. Death and population entries, broken down by municipality, year and sex were used as case source (ISTAT). Standardized mortality ratios (SMRs) were calculated as the ratio of observed to expected deaths. Smoothed municipal relative risks (RRs) with their corresponding 95% credibility intervals and posterior probabilities (PRPs) were calculated using the conditional autoregressive model of Besag, York and Mollie. The tool used for the Bayesian inference of the posterior marginal distributions was INLA. For men, RRs show a spatial pattern of mortality. The map depicts municipalities with higher mortality in the southwest of the Peninsula, in the regions of Campania and Latium. The province of Nuoro in Sardinia also stands out, as well as the southeast part of Sicily. This pattern is maintained for PRPs. The spatial mortality pattern of women is similar to that identified for men. Campania and Sicily remain the areas characterized by the highest mortality. The Apulia region emerges as an area with elevated mortality. All of these regions remain as over-mortality areas on the PRPs map. Knowledge of the spatial distribution of mortality could contribute to design specific health policies aimed at reducing mortality.

Keywords: Mortality; Spatial analysis; Italy


Sally Sonia SIMMONS 1

1 National Research University Higher School of Economics

The Contribution of Infectious and External Causes of Death to the Variations in Life Expectancy among Males and Females in South Africa from 2000 to 2005.

Background: From 2000 to 2005, South Africa experienced a sluggish increase in life expectancy due to high infectious diseases and external causes of mortality burden. There are gendered differences in the recorded rates of life expectancy. The study aimed to provide a clearer understanding of the contribution of infectious diseases and external causes of death to life expectancy in South Africa.

Methods: The trends in the age standardized infectious and external causes of death among South Africans for the period 2000 to 2005 were analysed. The contributions of infectious and external and the other causes of death to life expectancy were decomposed using the Andreev decomposition method, 1982.

Results: The recorded variations in life expectancy was due to the differences in the contributions of the infectious and external causes of mortality on the one hand and the increase in mortality from infectious and parasitic diseases and external causes on the other hand.

Conclusion: If infectious diseases and external causes of death remain the leading causes of death in the age group 15 to 34, the average South African young adult would be expected to live in the youthful ages with the perceived vulnerability of infectious diseases and external causes of death.

Keywords: Life expectancy; Mortality; Infectious diseases; External causes; South Africa


Sally Sonia SIMMONS 1

1 National Research University Higher School of Economics

Spatial variations in the prevalence of socio-demographic factors predicting visual impairment in Ghana: a bouquet of images

Background: In Ghana, age, sex, marital status, and residence predict visual impairment. The nation has ten administrative regions. However, poorer health conditions such as vision impairment are prevalent in some regions than others. These variations are influenced by sociodemographic factors. Yet, little has been documented about the spatial variations of visual impairment in the ten regions of Ghana. Hence, the present study acts in this direction. Method: The study used data from the World Health Organization Global Ageing and Adult Health Survey, Wave 1 for Ghana (2007-2008). Five thousand five hundred and seventy-three adults aged 18 years and above were included in the study. Age, sex, residence and marital status within the regions under study were used for the spatial analysis. The variable categories for sex, residence, and marital status were allotted numeric identities to ascertain the prevalence of the phenomenon of interest. Results: Vision impairment was less prevalent among the aged, residents, females and males and the currently married who identified with the regions in the southern part than regions in the other parts in Ghana. Conclusion: Implications of the results of the current study are relevant for public health policies in Ghana in streamlining eye health policies and programmes on the prevention, treatment, and management of visual impairment.

Keywords: Socio-demographic factors; Prevalence; Visual impairment; Regions; Ghana




1 Max Planck Institute for Demographic Research

Hypertension in Argentina. Transition and Differential Mortality. An indirect estimate based on cross-sectional data (2009-2013)

Because of the epidemiological transition in Argentina, the weight of infectious and parasitic causes of death has decreased, in favor of those related to cardiovascular diseases, cancer and malignant tumors and external causes (Grushka, 2015). Deaths from cardiovascular diseases represent the leading cause of death worldwide, being 34% in Argentina (Rubinstein et al., 2009). Arterial Hypertension (MSAL, 2011) is the main risk factor for mortality due to cardiovascular diseases.

This work aims to provide information on Arterial Hypertension (AH) in Argentina during the period 2009-2013 in the adult population, estimating indirectly the probability of transition to AH, the associated differential mortality, and life expectancy with and without the medical condition, which can be an input for decision making in prevention policies. An indirect estimation was made from two cross-sectional surveys, adapting the “intercensal method” described in Guillot and Yu (2009).

The main results indicates that the probability of being hypertensive increases more than 5 times when adulthood begins, with almost 2% of adults between the ages of 30 and 70 entering AH every year, according to estimates for the 2009- 2013 period. Life expectancy at age 30, temporary until age 70, was 8.6% lower in people with hypertension. The percentage of years with AH that a person of age 30 expected to live up to the age of 70, regardless of their initial status, was 36%.

Keywords: Health; Mortality; Risk factor; Incidence; Indirect estimation


Wirginia Anna WOJTCZAK,1 Helena M. REPCZYŃSKA 2

1,2 Adam Mickiewicz University, Poznan

Body size of adult Polish men born before, during and after WWI

Changes in height and weight among population which occur under specific socio-economic circumstances are generally accepted as valid indicators of conditions in society. The purpose of this study was to estimate the influence of birth date during the years before, during and after WWI on body size (height, weight and BMI) of adult men in Poland. Information on height and weight was obtained from the files of Poznań Penitentiary Registers deposited in Poznań State Archive. Data on 298 men were collected, who were measured at the day of admission. The sample was divided into four groups by birth year with reference to WWI, i.e. men born before WWI (1885-1899 and 1900-1913), during the war (1914-1918) and afterwards (1919-1930). Age differed significantly among groups (pre>during>post WWI). Body mass index (BMI) was calculated based on individual height and weight data. Differences in height, weight and BMI values of men in the four birth cohorts were evaluated using the analysis of variance (ANOVA). The impact of SES on height, weight and BMI values was also assessed (ANOVA). The results reflected the influence of poor social and living conditions at the time of birth and in early childhood during WWI and in the years after achieving Independence in 1918, and during the economic crisis of the 1920s, on body size.

Keywords: War time; SES conditions; Growth status; Height; Weight; BMI; Birth cohorts


Aleksandr ZUBKO,1 Tamara SABGAYDA 2

1,2  FRIHOI of MoH of the RF

Mortality from surgical vascular diseases as indicator of prevention effectiveness for circulatory diseases

Indicator mortality from surgical vascular diseases is very important part all mortality for circulatory diseases.Mortality in rubric ICD-10 «Atherosclerosis» (I70.0, I70.1 and I70.2) composed 42.2% among men and 26.9% among women in  in 2010, and 70.0% and 44, 2% respectively in 2016. In rubric “Diseases of arteries, arterioles and capillaries” determined almost completely by diseases requiring surgical treatment (I71, I72, I73.9, I74, I77.0, I77.1, I77.2, I77.4, I77.5 and I78.0). In 2010, the analyzed causes accounted 98.1% of men deaths and 97.9% of women deaths in 2016 99.0% and 98.9% respectively. In rubric "Diseases of veins, lymphatic vessels and lymph nodes, not elsewhere classified" proportion increased to 48.3% among men and to 72.4% among women. Proportion of surgical vascular diseases (I80.1, I80.2, I82.1, I82.2, I82.3, I82.8, I82.9, I83.9) also increased from 64.1% among men and 69.9% among women to 72.9% and 81.4% respectively. Sum mortality from surgical vascular diseases is increasing: from 16.8 per 100 000 among men and 12.3 among women to 21.5 and 15.9, respectively. Surgical vascular diseases left unattended of today programs to reduce circulatory diseases. Practice using mortality from surgical vascular diseases as the indicator could objectively judge death level and proportion of early detection of this diseases.

Keywords: Mortality from surgical vascular diseases; Early detection; Structure circulatory diseases



1 Department of Demography and Geodemography, Faculty of Science, Charles University

Researches dealing with the influence of education on fertility in Czechia after 1989

The reproductive behavior of the Czech population began to change together with the change of political regime in Czechia after 1989. The most significant changes that took place in Czechia since 1989 include among other things a substantial drop in the number of births and significant fertility postponement. It has been proven that education is one of the factors that have a crucial influence on fertility timing in the present. The aim of this paper is to summarize the current state of knowledge documenting the influence of the educational attainment level on the fertility of Czech women from the early 1990s to the present. It turns out that women are most often starting families after completing education and stabilization on the labour market. Women with university degree postpone the birth of children to the later age and the proportion of university educated women in the population is still growing. At the same time, there is a group of women in the Czech population who are starting families and continue in their studies. This is especially concerns mothers studying a doctoral degree.

Keywords: Fertility; Education; Fertility postponement; Czechia



List of abstracts – online poster


Alexander RUSANOV 1

1 Lomonosov Moscow State University

Socio-economic and demographic characteristics of suburbanization in Moscow Region

Post-socialist suburbanization changed greatly both nearest and remote suburbs in Moscow Region due to revival of the land market.

Development of the nearest suburbs changes age structure of their population. Comparison of the cities from the near suburban area (Balashikha, Mytishchi, Odintsovo) and from more remote (Elektrostal, Kolomna) shows that in 2004 all these cities had a similar age structure close to the average in the region. Mass construction of new multi-storey blocks in the nearest suburbs differentiated the age structure of the population: data for 2016 shows that the proportion of the population below the working age in near suburban area is higher than the average in the region, and in remote suburbs, on the contrary, is lower.

Main distinctive feature of remote suburbs is widespread distribution of informal seasonal settlements which appeared mainly in the late soviet period and predetermined modern mobility, when significant part of Moscow population has two residences: apartment in the city and house in suburbs.

The influence of suburban areas on socio-economic development at the regional level is manifested through the development of various modes of transport: informal seasonal settlements mainly contribute to the development of public transport (commuter trains and buses), cottage settlements – to modernization of the road network. Construction of new areas of mass multi-apartments affects the emergence of new metro stations outside the Moscow ring road. At the local level, all types of settlements contribute to the development of trade and construction.

Keywords: Suburbanization; Second home; Allotment garden; Suburbs; Moscow Region


Alexander SUBBOTIN 1

1 Lomonosov Moscow State University, Russia

Contemporary Brain Drain in Russia: Problems and Prospects

At the present moment brain drain in Russia does not simply continue to exist, but its scale has increased over the recent years, as well as the scale of emigration on the whole. Non-return emigration of highly skilled specialists is aggravated by the present demographic situation, which can be described as a demographic crisis. It is first and foremost connected with not only qualitative, but also structural and quantitative negative changes in the population of Russia.

Based on the available data, the current situation is compared to that a quarter century ago. It is revealed that in recent years the level of emigration of people with higher education has reached the level of the mid-1990s.

A fairly stable trend of the outflow of Russian highly qualified civil aviation pilots to the countries of Southeast Asia is considered as a revealing example. In 2015 50 first-class pilots left Russia, whereas in 2016 and the first half of 2017 the number exceeded 250 pilots. Of course, it should be noted that not all aircraft captains have left Russia forever, but some of them will definitely stay abroad. Thus, if each separate field is considered in detail, it can be seen that Russia has suffered considerable losses therein – somewhere more, somewhere less.

Keywords: Brain drain; Demographic crisis; Intellectual migration; Migration of highly skilled specialists; Negative qualitative changes; Non-return migration



Photos from the conference.



We are grateful to:




Add comment

Overview of comments

There have not been any comments added yet.