2.4.1 Population Inequalities by Poverty

Inequalities in population behaviour and trends refer to three aspects of demographic change, namely i) risk of early mortality (the lower the socio–economic level of individuals and communities, the higher early mortality), ii) final fertility intensity, i.e. the number of children women have (higher fertility at lower socio–economic levels) and iii) timing of fertility (high fertility at younger ages at lower socio–economic levels). These three types of disparity reflect the systemic pattern inherent in the population dynamics of poverty and as such require special attention. Since they tend to worsen or put pressure on the situation of lower socio-economic groups, they feed back into the pattern, and thus exacerbate existing social inequalities.

1.1. Inequality in Population Behaviours and Trends

Facts/messages: Poverty is often associated with population behaviours and trends whose main features are: avoidable mortality, unintended pregnancies and adolescent births, early marriage and reproduction and as a result an age structure (at the group and household level) that is burdened by heavier child-rearing responsibilities and more rapid growth as a group. This pattern does not necessarily reflect the desires of poor people, suggesting that there may be underlying phenomena of social exclusion and constraints on the exercise of human rights. The structure of the disadvantage generated by social inequality limits demographic convergence.

Individual behaviors and practices have implications at the aggregate level, on population size and growth, the composition of a population by age and sex, and its spatial distribution. Education plays an important role in this context. Aggregate population dynamics reflect the changes in individual demographic processes. Slight changes in fertility and mortality can produce large changes in the size of the total population, e.g. low fertility levels have major implications for the overall population size, where a decrease is expected. Population ageing accelerates as fertility levels remain low.

Methodology: National population and housing censuses provide information that allows indirect estimates of mortality and fertility by using standardized demographic procedures. Indicators include the Total Fertility Rate, the infant mortality rate, the percentage of mothers (and those pregnant for the first time, if available) among adolescents, the dependency ratio, and the average number of children per household by poverty status (extreme poverty, poor and not poor). The DHS surveys allow direct estimation of fertility and infant and child mortality. In addition, the Household Wealth Index allows a stratification of the population into five strata (quintiles), using the standard DHS criterion. Although this stratification does not allow the identification of poverty as such, it provides at least a reasonable measure of the relative socio-economic position of population groups, such as women, with respect to each other. An alternative is to use completed years of education, which can be done with census as well as DHS data. DHS surveys are crucial for analyzing aspects that are not covered in the census, in particular those regarding intermediate variables of fertility and mortality and issues such as reproductive ideals, which cannot be investigated based on census information. If household surveys of the Living Standards Measurement Study (LSMS) type are available that contain a demographic module that allows the calculation of demographic indicators, these can be classified by poverty strata, in the strict sense of the word. If only censuses are available, the index of socio-economic stratification can be calculated and the percentages of poverty in the most recent survey can be used to define the extremely poor, the poor and the non-poor. Vital statistics can be used to compare and validate indirect estimates of fertility and infant mortality with census data, but the irregular quality of vital statistics and their limited socio–economic information make it difficult to use them in a systematic fashion to calculate demographic rates.

INTERACTIONS OF POPULATION WITH POVERTY AND VULNERABILITY

Generate demographic tabulations of persons and households by socio-economic strata. Additional indicators are the age at the first union, sexual initiation (taking into account adolescents that have not yet initiated sexual activities), percentage of the use of modern contraceptives, desired fertility (Westoff or Bongaarts variant), number of children at the onset of contraceptive use, percentage of individuals who received professional antenatal care, access to drinking water and sanitation, malnutrition (of children), immunization (of children), high-risk sexual relationships and the prevalence of HIV, also by socio-economic strata.

Primary Sources:

  • Household surveys of the Living Standards Measurement Study (LSMS) type;
  • DHS and other specialized surveys household surveys;
  • International Reproductive Health Surveys (IRHS);
  • Population censuses;
  • Vital statistics.

Secondary Sources:

  • DHS. STATCOMPILER;
  • IADB. Website with MDG indicators disaggregated by social groups;
  • ECLAC. Social Panorama. Chapter 3: “Demographic Inequalities and Social Inequality. Recent Trends, Associated Factors, and Policy Lessons”;
  • ELDIS website on poverty and PRSPs: http://www.eldis.org/poverty/prsp.htm;
  • World Bank Poverty Net site: http://www.worldbank.org/prsp.

Tools:

  • Lamlenn B. Samson (2008). Guidance note for the in-depth analysis of data from a Population and Housing Census. Dakar, CST: sections on Housing and Living Conditions and Analysis of Non-Monetary Poverty;
  • The Resource Guide for Youth and Poverty Reduction (UNFPA, 2011) contains an example of the analysis of PRS and other social programmes, in terms of their benefits to poor people, specifically for the case of Honduras (Case Example 36).

1.2. Trends in Reproductive Inequality

Facts/messages: Reproductive inequalities persist despite the fact that the demographic transition is proceeding across-the-board and affects all social groups and regions (depending on the country concerned). Inequities are as relevant as mere inequalities, because they operate in the same direction, punishing people labouring under unfavourable socio-economic conditions. In addition to the statistical analysis carried out in the previous section, it is important to analyze how inequalities / inequities have developed over time.

Methodology: The indicators for reproductive inequality are the same as those used to examine population dynamics and SRH in the previous section (with the exception of those that relate to the natural and total growth of the population). Censuses, household surveys of the Living Standards Measurement Study (LSMS) type and specialized surveys (DHS and others) can be processed directly. In principle, the three sources are useful for quantifying regional and ethnic inequalities, although they provide different options based on the number of indicators for “inequality”.

Census data offer the possibility to control for the distortions in composition that can occur in survey–based convergence analysis. Distortion in composition arises, for instance, from using socio–economic groups whose representation changes over time, which can have quantitative and substantive implications. This limitation can be controlled by using socio–economic groups that maintain their relative representation over time, such as socio–economic quantiles that are specific to urban and rural areas, e.g. autonomous urban and rural quintiles. Assign an equal weight to the two dimensions considered, in order to obtain the socio–economic index from the simple average of the two sub–indices. Note that the procedure should be applied to rural and urban areas separately, so that the weighting factors are used specific to each area. By doing so the different quantiles are particular to urban and rural areas. Construct this initial segmentation to allow for inter-temporal monitoring exercises, controlling for compositional effects.

Inequality can be measured using various procedures and measures, from comparisons between extreme groups to heterogeneity measures (coefficient of variation) and synthetic indicators such as the index of concentration. In general, the measures of heterogeneity are preferable with regard to territorial inequalities while composite indices are appropriate when assessing socio-economic inequalities. Unlike inequality, inequity is not a strictly statistical concept. In the case of health, it is defined as disparities in health that are a result of systemic, avoidable and unjust social and economic policies and practices that create barriers to opportunity. Note that this calls for an operational definition of what is considered systemic, avoidable and unjust. This is not always obvious. One may argue, for instance, that inequality in the distance to the nearest service facility is inequitable if it is due to the fact that such facilities tend to be constructed in better-off neighborhoods, thus making them less accesible to the poor. But if it is due to low demographic density, it can be justified on operational grounds. Finally, it may be purely random: even in the most equitable health system, some families will live closer to the nearest facility than others.

When dealing with aspects such as, i) intermediate variables of fertility and mortality, ii) micro–modeling and iii) unwanted fertility surveys have advantages over censuses and are therefore to be preferred. When using household surveys in the case of socio-economic segmentation, quintiles should be constructed with the income variable (the same used for estimating poverty).

Primary Sources:

  • Household surveys of the Living Standards Measurement Study (LSMS) type;
  • DHS and other specialized surveys, household surveys;
  • International Reproductive Health Surveys (IRHS);
  • Population censuses;
  • Vital statistics.

Secondary Source:

  • ECLAC. Social Panorama. Chapter 3: “Demographic Inequalities and Social Inequality. Recent Trends, Associated Factors, and Policy Lessons”.

1.3. Inequalities in Mortality and Morbidity

Facts/messages: Early mortality has declined strongly in recent decades in most countries of the world. As a result, there have been significant gains in life expectancy. This progress was not halted on account of economic recessions or political crises, but there are still marked disparities between and within countries. In fact, an overall downtrend co-exists with growing heterogeneity. Compliance with the MDGs will only be possible if the future reduction in infant mortality is concentrated in the most disadvantaged groups.

Public policies should be focused on reducing these inequalities in survival, between population strata, ethnic groups and geographic areas. In the case of infant survival, it is of utmost importance to dismantle the disadvantage structures associated with high fertility. Further, the intermediate determinants of infant mortality are related to some characteristics of mothers (extreme age, high parity, pre and post natal controls, immunization and nutrition) as well as exposure to pathogenic factors associated with the habitat, as confirmed by DHS surveys.

These linkages make it technically possible to achieve a decrease in demographic inequalities (reproductive and survival), which is to some extent disconnected from the reduction of socio-economic inequalities. One should also take into account differentials in specific mortality with regard to age and sex (young people and older adults) and, to the extent possible, differentials in morbidity, measured in terms of Disability Adjusted Life Years (DALYs), among different social strata. Differential morbidity due to HIV/AIDS should be taken into account.

Methodology: For the purposes of this chapter, demographic inequality refers to the risk of early mortality that is larger at lower socio–economic level of individuals and communities. The measurement of the inequalities in survival can be structured into three groups of indicators: i) absolute and relative differences in infant mortality rates; ii) measures of disparities among population groups and geographic areas (convergence or divergence); and iii) measures of the effect or impact of socio-economic conditions on the level of mortality and regarding access to health, in order to determine the degree of concentration of inequality (differences between extreme quintiles). For morbidity differentials, one depends to a considerable extent on a number of studies based on household surveys; there are not many studies that have estimated the burden of disease in terms of DALYs by socio-economic stratum.

Primary Sources:

  • Population censuses;
  • DHS.

Secondary Sources:

  • United Nations General Assembly Special Session (UNGASS): National reports on the situation of the HIV/AIDS epidemic;
  • World Bank: Health, Nutrition and Population (HNP) reports;
  • ECLAC: Social Panorama. Chapter 3: “Demographic Inequalities and Social Inequality. Recent Trends, Associated Factors, and Policy Lessons”.

71  The mechanisms whereby these dynamics feed back into the reproduction of poverty will be formally presented, conceptualized and illustrated with ad-hoc indicators and procedures in Chapter V (e.g. in the absence of social and geographic mobility the greater natural population growth among poor people implies that their weight within the total population will gradually expand).
72 If it is not possible to have access to the databases, the online processing of DHS STATCOMPILER can be used.
73   ECLAC (2005). Social Panorama. Chapter 3. Demographic Inequalities and Social Inequalities.