As part of the development of the Situation Analysis of Children, the UNICEF Mozambique Country Office undertook a comprehensive analysis of child poverty, using both consumption and deprivation based measures. This was the first attempt to assess the prevalence of child poverty in Mozambique, as a concept distinct from adult poverty, using measures specifically adapted for the national context. The aim of the initiative was to increase strategic information about the situation of children in Mozambique, highlighting both the extent of child poverty and disparities in its distribution, which could be used to inform programme and policy development and to advocate for increased investment to address child poverty.
The strategy consisted of adapting international standards for measuring child poverty in its economic and non-economic dimensions to the Mozambique context. Two databases of recent national household surveys conducted by the National Institute of Statistics were used. The national Household Income and Expenditure Survey (IAF 2002/2003) database was used for measuring child poverty in its economic dimension while the Demographic and Health Survey (DHS 2003) database was used for measuring child poverty in its non-economic dimension. Once measures were calculated, a consultation process with various partners was conducted to validate findings, including the Ministry of Planning and Development, bi-lateral partners and the World Bank.
The economic dimension of child poverty: analysis using the consumption approach
The purpose was to identify the proportion of Mozambican children living below the poverty line by using the cost of basic needs (CBN) approach This approach consists of the development of a food basket, comprised of items consumed by the poor, that meets calorie needs. The cost of this basket comprises the food poverty line. A non-food poverty line is obtained by examining the share of total expenditure allocated to non-food by households living near the food poverty line. The poverty line is simply the sum of the food and non-food poverty lines.. The consumption measure is an important aspect of poverty that is well grounded in economic theory, although it does not provide the full picture of child poverty as a multi-dimensional phenomenon. In addition, the IAF survey did not measure the allocation of resources within the household directly and therefore measures apply only to an "average" household member. These measures consider as poor every child that lives in a household categorised as poor. As scope exists for families to allocate fewer or more resources to children, making children better or worse off than the average household member, the level of child poverty rates under this approach can best be viewed as the percentage of children living in households that are categorised as poor.
The non-economic dimension of child poverty: analysis using the deprivation approach
The purpose was to identify the proportion of children living in absolute poverty using the deprivation approach proposed by the report ‘The Distribution of Child Poverty in the Developing World’ (2003) and known as the Bristol indicators Research funded by UNICEF and undertaken by the Bristol University and the London School of Economics.. The study was intended to provide a firm conceptual foundation for defining and measuring child poverty and its dimensions in developing regions of the world. The Bristol indicators are based on the deprivation approach to poverty, in line with the definition of absolute poverty agreed by 117 countries at the World Summit for Social Development, as “a condition characterised by severe deprivation of basic human needs” (United Nations, 1995). The Bristol indicators look at seven types of deprivation and define the proportion of children living in absolute poverty as those children facing two or more types of severe deprivation. During the analysis, the team felt that the indicators needed to be adapted to reflect more appropriately the Mozambican context. An important challenge was therefore achieving consensus on the indicators to be used in the analysis of the DHS data, while being mindful of the need to avoid significantly altering the international standards, for comparison purposes. As a result of the consultation process, it was agreed that some of the indicators would be adapted as outlined below. The remaining indicators were felt to be suitable.
Severe Food Deprivation: children whose heights and weights for their age are more than -3 standard deviations below the median of the international reference population, i.e. severe anthropometric failure. Modification: The name of this indicator was changed to reflect the fact that malnutrition in children is not only related to the lack of food but also to poor health. The name "severe nutrition deprivation" was adopted.
Severe Water Deprivation: children who only have access to surface water (e.g. rivers) for drinking or who live in households where the nearest source of water is more than 15 minutes away (indicators of severe deprivation of water quality or quantity). Modification: The team felt that it was not appropriate to define 15 minutes walk (i.e. 30 minutes round trip) as 'severe' deprivation, particularly in a context where most households do not have access to safe water. The indicator was therefore modified to 30 minutes (i.e. one hour round trip). The final figure was not significantly different from the figure obtained using 15 minutes, but it was felt that it carried more weight in terms of severity.
Severe Deprivation of Sanitation Facilities: children who have no access to a toilet of any kind in the vicinity of their dwelling, including communal toilets or latrines. No modification.
Severe Health Deprivation: children who have not been immunised against any diseases or young children who have had a recent illness involving diarrhea and did not receive any medical advice or treatment. Modification: It was felt that the treatment of diarrhea was not a good indicator, as many diarrhea cases can be treated at home with ORT (there is therefore no need to seek ‘medical advice’). It was therefore agreed that the treatment of acute respiratory infection (ARI) should be used instead, particularly as ARI is a greater cause of death among children in Mozambique. A composite indicator was agreed, of children that had never been immunised or those that had suffered from a severe episode of ARI that was not treated.
Severe Shelter Deprivation: children living in dwellings with more than five people per room (severe overcrowding) or with no flooring material (e.g. a mud floor). Modification: A major issue of debate was the component of this indicator relating to the household flooring material. The DHS indicates that the percentage of children living in overcrowded housing (i.e. more than five people per room) is 6%, while those in housing with a mud floor is 72%. It was felt that it is not an indication of severe deprivation to have a mud floor, as most people in Mozambique live in houses with mud floors. It was therefore agreed that the ‘mud floor’ component should be removed from the indicator, leaving only the overcrowding aspect. As the DHS provides little information on other housing characteristics (i.e. the existence of a roof) it was not possible to agree on a more appropriate measure.
Severe Education Deprivation: children aged between 7 and 18 who have never been to school and are not currently attending school (no professional education of any kind). No modification.
Severe Information Deprivation: children aged between 3 and 18 with no possession of and access to radio, television, telephone or newspapers at home. Modification: It was felt that telephone is more a means of communication than information. This component was therefore removed from the definition. The age range was also modified, from 3-17 to 5-17, as it was felt that age 3 was too low.
The Bristol indicators classify any children or households with children who suffer from two or more different types of severe deprivation of basic human need as living in absolute poverty. Discussions were held around whether, in a context where 54% of the population lives below the poverty line (consumption approach), two or more deprivations constituted absolute poverty, or whether the measure should be three or more deprivations. It was eventually agreed that two or more deprivations should be used. This also led to debate on the feasibility of weighting particular deprivations considered to be more important than others, although it was agreed that, in light of the inter-dependence of human rights, no weighting should be allocated.
One result achieved is the development of an adapted tool for the analysis of child poverty in Mozambique, which can be used in future analyses to assess trends in the prevalence of child poverty. The main results, however, are the findings of the analyses, which not only provide a baseline for monitoring purposes, but also strategic information that can be used in developing programmes and policies, advocating for increased commitment and leveraging funds for the realisation of children’s rights. The findings of each of the analyses are outlined briefly below:
The economic dimension of child poverty: the analysis showed that the level of child poverty in 2002/3 was significantly higher than the level of adult poverty in the same year (58% to 49% respectively). In addition, while both adult poverty and child poverty decreased by 22% between 1996/7 to 2002/3, the gap between the two groups did not decrease. Considerable disparities between provinces were noted, with child poverty increasing over the period from 1996/7 to 2002/2003 in Cabo Delgado and Maputo provinces. The analysis also highlighted the phenomenon of increased urban child poverty, with the proportion of children living below the poverty line dramatically increasing from 52% in 1996/7 to 60% in 2002/3.
The non-economic dimension of child poverty: the analysis indicated that 49% of children face severe water deprivation, 47% severe sanitation deprivation, 6% severe shelter deprivation, 24% severe education deprivation, 17% severe health deprivation, 20% severe nutrition deprivation and 39% severe information deprivation. The proportion of Mozambican children living in absolute poverty in 2003 was established at 49%, with significant disparities: 20% in urban areas versus 63% in rural areas, 90% among children from the poorest households (first wealth index quintile) versus 4% among the best-off quintile (fifth wealth index quintile). In terms of provincial disparities, Zambezia was found to be by far the most affected province (75%), followed by Sofala province (59%).
A comparison of the two different analyses, both using data from the same years, led to interesting findings. While both studies found higher levels of poverty in rural areas than in urban areas, there were interesting differences in terms of the provincial distribution of poverty. For example, Zambezia province was the most affected using the modified Bristol Indicators measure, while the consumption measure showed Zambezia to be one of the better off provinces, and showed an decrease from 75% to 49% child poverty between 1996/7 and 2002/3.
Severe deprivation and absolute poverty among Mozambican children, 2003
Absolute poverty among children by province
The findings of the analysis support the argument that improvements in the level of poverty based on consumption measures do not automatically lead to improvements against key indicators of child well-being.
This initiative can provide an example to other countries on the adaptation of international child poverty assessment tools to suit local contexts. The modified version of the Bristol indicators can also be used in the region, where many countries present similar characteristics.
The high levels of poverty identified and the disparities between provinces provide strategic information to assist in the targeting of interventions, and can be used to advocate for increased commitment and investment for children.
This initiative indicates the need to enrich the “household characteristics” module of future household surveys, such as DHS and MICS, with additional questions related to living conditions, such as house roofing type and whether household members sleep on the floor or not.
Challenges and Future Activities
As with any measure of poverty, there were limitations in the two methodologies adopted. The adaptation of the Bristol indicators to the local context made it less statistically relevant in terms of international comparison; the added value of developing new measures of poverty, when they already exist, to some extent, in the Millennium Development Goals, was questioned; and although analysis of child poverty using the Bristol indicators broadens analysis from the traditional income or consumption-based definitions, it still fails to capture any aspect of child protection. Another limitation of the process of analysis is that the relative complexity of such analysis makes it a largely academic process and therefore the adaptation of the Bristol Indicators was not discussed within communities or with children themselves.
The publication and dissemination in 2006 of the Situation Analysis of Children, which will contain the analysis of child poverty, will be an opportunity to increase awareness on the concept and prevalence of child poverty in the country and should pave the way for increased debate around this issue.