Latin America & the Caribbean sessions abstracts


Latin America & the Caribbean. Tuesday 13 September 16:45pm 

Clustering age-sex structures to monitor their development over time: Latin America and the Caribbean sub-national areas 1960-2011
Simona Korenjak-Černe 1, Ludi Simpson 2, 1 University of Ljubljana, Slovenia, 2 University of Manchester 

We analyse the development of age-sex pyramids based on the weighted agglomerative hierarchical clustering of age-sex structures (Korenjak-Černe et al., 2014). We clustered age-sex structures of sub-national areas of Latin America and the Caribbean from national censuses 1960-2011 (downloaded from IPUMS), into 4 main clusters and 9 more detailed clusters. We focus on three research questions: 1. Can the clusters be associated with socio-demographic indicators that are usually used to describe social and economic development? 2. Is there a strict order of the age-sex structures over the observed time? 3. Are the dissimilarities between areas smaller inside the clusters than inside the countries? Do the dissimilarities diminish over time? We described the clusters with indicators including industrial structure and women’s economic activity to be able to link them with the classical development characteristics. To answer the second research question we investigate the transitions of sub-national areas from one cluster to another across time and connect these observations with the answers to the first question. In order to answer the last research question, we intend to observe dissimilarities among sub-national areas inside countries and compare them with the dissimilarities inside clusters. 

How prepared are we for population ageing? Current Status of Quality of Life of the Elderly in Countries of Latin America
Javiera Fanta Garrido, National University of Córdoba-UNC & National Council of Scientific Research-CONICET    

The countries of Latin America and the Caribbean are experiencing an abrupt rise in the share of older population. Currently, there are nearly 70 million people aged 60+ in the region, representing 11.1% of the total population. In the light of the priorities defined by the Regional Strategy for the implementation of the Madrid International Plan of Action on Ageing, the assessment of the various aspects related to ageing, protection and welfare of the elderly have become paramount topics in the regional agenda. This study contributes to create a composed index for assessing the Quality of Life of the Elderly (QLE) in Latin America. The methodological approach for the construction of the QLE Index was grounded on the judgment of experts, recommendations from the Latin American and Caribbean Demographic Centre (CELADE), and literature on Active Ageing Index methodology. Population censuses from the 2010 round and national surveys from 19 countries were analysed. After examining available sources and excluding low-quality data, an index of 20 variables-distributed in three domains-was developed. This measurement tool was used to analyse and compare the state of the quality of life of older aged in six countries of the region (Argentina, Bolivia, Chile, Ecuador, Guatemala and Peru). Results show that life quality of the elderly has a better performance among countries that are currently placed in advanced stages of the demographic transition. However, even the top performing countries must aim for further improvements, as the overall index comprises baseline standards of quality of life. 

Comparative sub-national demographic development in Latin America
Ludi Simpson 1, Leandro Gonzalez 2, 1 University of Manchester, 2 CIECS Argentina    

How predictable is sub-national demographic change in Latin America and the Caribbean? The presentation is based on a collaborative project which has prepared and is analysing databases of demographic indicators from national censuses and published vital statistics, about the largest sub-national administrative areas in each country. The project’s key questions are: a) Has the variation between areas decreased? In other words, is there evidence of demographic convergence? b) Is similarity between areas stronger within a country or between similar areas of different countries? What evidence exists of demographic trends characterised by geography, for example climate or economic development or other? c) What can we learn from this evidence to make more reliable assumptions to underlie projections of future population? The presentation will discuss how members of the project are variously operationalising ‘convergence’ in demographic indicators, using cluster analysis and multivariate regression. It will present initial results of fertility, mortality, headship rates, and economic activity. The collaborators are based in Brazil, Venezuela, Costa Rica, Argentina, and Ecuador, and are planning independent analyses to learn from each other while following different approaches and motivations. Detailed results will be possible relating to headship rates for which the author is primarily responsible. The project’s collaboration is funded by a British Academy award until February 2018. The analyses are being prepared for this year’s Latin America Population Congress (ALAP), and will bring a Latin American partner to BSPS in 2017. 

Latin America & the Caribbean. Wednesday 14 September 11:30am 

Unintended Fertility and Unmet Need for Family Planning in Countries of Latin America and the Caribbean
Javiera Fanta Garrido 1, Roberto Ariel Abeldaño 2, 1 National University of Cordoba, IPMA-CONICET (Buenos Aires, Argentina), 2 School of Public Health, Faculty of Medicine, National University of Cordoba-CONICET (Cordoba, Argentina)   

Today, the proportion of unmet need for family planning in Latin America and the Caribbean is below the worldwide average of 12%. Yet, a critical gap continues to exist between actual and desired family size, resulting in unintended childbearing in many countries of the region. Unmet need for family planning and unintended fertility are basic measures of women’s autonomy and capacity for self-determination in the area of reproductive decisions. However, much of what is known today on the subject comes from industrialized countries and developing countries in Africa, and little is known about this topic in LAC. This investigation examines the changes over time in unmet need and unintended fertility in six countries of LAC (Bolivia, Colombia, Dominican Republic, Haiti, Honduras and Peru). We analyse the differential effects of women’s age, place of residence and level of education in both trends. Data was derived from Demographic and Health Surveys conducted between 1985 and 2013. The sample includes married women and women in union aged 15-49. Results show that, except for Guatemala, the proportion of unmet need decreases over time, while unintended fertility trends exhibit much higher variability. In almost all countries, over half of the total pregnancies and births were mistimed or unwanted. The levels of unmet need and unintended childbearing differ sharply when determinants of fertility are introduced into the analysis. As educational level increases, unmet need and unintended fertility levels decrease. The place of residence continues to be a differential factor among fertility results, reflecting higher rates of unmet need and unintended childbearing in rural areas. Additionally, young women (15-19) show higher outcomes in both indicators, irrespective of the year. 

Effects of external causes of death in life expectancy and population structure in Latin America, 1970-2010
Jenny Garcia, Institut National d'Etudes Demographiques INED & Université Paris 1- Pantheon Sorbonne   

In recent decades Latin America has experienced a significant increase in the proportion of deaths from external causes, especially those triggered by accidents or violence (mainly homicide). For this study, there have been selected twelve countries in the region, six of which are at the top of the lists of world's most violent countries: El Salvador, Colombia, Venezuela, Guatemala, Honduras and Brazil (Waiselfisz, 2008). The aim is to analyze the differential impact that deaths from external causes have had on life expectancy of men and women of all ages during the period 1970-2010 and its implications in 2010's population structures for each country. In this sense, Latin American Mortality Database (LAMBDA) and National Census data were used to construct Multiple Decrement Life Tables in the absence of external causes, allowing us to capture both differential: (1) in life expectancy for all ages, as (2) in the age distribution of deaths. In addition, survival functions are used to present a counterfactual analysis of population structure for 2010. The results show that there are two dimensions to consider in the differential in life expectancy introduced by external causes for both sexes: the amplitude of the age range in which the deaths occur and the proportion reached for each age group. While in men, the age range of incidence of external causes is concentrated in the youngest ages (losing about 5% of life expectancy in all countries throughout the period considered), women offer the greatest differential after 70 years old. 

Living as a ‘couple’ in a context of partnership instability: gender, conflict and partner violence in the French Caribbean
Stephanie Condon, Carole Beaugendre, INED   

Union formation and couple relationships between men and women in the Caribbean continue to be in the spotlight in discussions (academic or policy-related) on topics such as school drop-out rates, teenage pregnancy and family violence. Sixty years after the first studies highlighting the specificities of the Caribbean context, research and public debate still tend to make links between the supposedly ‘disorganized’ nature of Caribbean family life and a wide range of social problems. Since the focus of research has tended to be placed on ‘fathering’ and ‘mothering’ roles within this family relational system, and in particular on the importance of single mother or ‘matrifocal’ families (in the Caribbean and within Caribbean-origin populations in migration destinations), less attention has been paid to partner relationships. Whilst census and survey data confirm the instability of such relationships, the data also show that a considerable number of people report, at any one time, ‘living as a couple’. At the same time, this experience remains structured by gendered perceptions and expectations. Using data from a recent survey conducted in the French Caribbean islands of Guadeloupe and Martinique (Ined-Insee, 2010), as well as other survey materials, we investigate the dynamics of such relationships, the quality of relations and also the role of gender-based violence. Social inequalities, insularity and the migration context are also examined as factors in outcomes for partner relationships.

Educational Homogamy and Descending Economic Inequality in Ecuador
Adriana Robles, independent consultant 

Evidence indicates that in a context of growing economic inequality, educational homogamy increases. Whereas preferences for the potential couple tend to be similar between men and women, greater inequality could increase the importance of socio-economic aspects in choosing a partner. Also, facing greater economic inequality, there is more residential segregation as well as other areas of socialization, which can increase the assortative mating by income. However, what happens when inequality is reduced? From the case study of Ecuador for the period 2000-2015, the objective of this study is to analyze the changes in the levels and patterns of homogamy by level of education of the spouses in the context of reducing income inequality considering the change of the national educational structure. Log-linear models are used to estimate the changing association between couples’ educational levels while they control the shifts of the educative structure. Preliminary results indicate that there is a reduction of the propensity of a homogamous union in relation to a heterogamous one, although homogamy remains as a major feature of a union. The pattern of homogamy remained equal between 2000 and 2015. There is a major strength of the barriers at the lower and upper end of the educational distribution. Nonetheless, changes were observed in the propensity of homogamous unions throughout the educational hierarchy. The reduction in the odds of homogamy occurred with more strength in the lower end of educational distribution.