Child multidimensional poverty in latin america: are data restrictions insuperable for the development of a child mpi for the region?

González-Uribe, Catalina; Pinilla-Roncancio, Mónica; García Jaramillo, Sandra; Ritterbusch, Amy; Carrero, Lorena; Sánchez- Franco, Sharon (2018). 'Child multidimensional poverty in Latin America: Are data restrictions insuperable for the development of a Child MPI for the region?' Paper presented at the annual conference of the HDCA, Buenos Aires, Argentina 2018.


Poverty is often measured using unidimensional measures such as income or consumption (Ravallion, 2016). No single indicator alone can capture the multiple aspects that constitute poverty and in the case of child poverty, indirect measures do not capture the complexity of this phenomenon (Delamonica, 2014; Singh & Sarkar, 2014). Beyond measuring child poverty, action on the determinants of poverty is urgent considering the backdrop of the Sustainable Development Goals (SDGs). Target 1.2. of the SDGs agenda established that countries must reduce at least by half the proportion of men, women and children of all ages living in poverty across all dimensions according to national definitions (Inter-Agency and Expert Group on Sustainable Development Goal Indicators, 2016). To achieve action, poverty monitoring is an undeniable task of all national governments, particularly in Latin America (LA), where inequality levels have increased in the past decades, with catastrophic consequences for the most vulnerable.

A multidimensional poverty measure can incorporate a range of dimensions and indicators capturing the complexity of poverty and deprivation that individuals face (Roelens, et al., 2012). Only one study in LA conducted by UNICEF and ECLAC (2010) has analysed the levels of multidimensional poverty of children in the region using a comparative perspective. Countries such as Colombia, Honduras, El Salvador, Guatemala, Mexico and Uruguay have developed proposals to measure child poverty. However, neither of these exercises has had a practical impact on how to measure children´s multidimensional poverty in LA. Also, there is still a lack of information related to the levels of multidimensional poverty for children in the region and no study to date has proposed a comparable Child Multidimensional Poverty Index (Child MPI) for countries in the region. The main objective of this study is to develop a comparable Child MPI for countries in LA using the Alkire-Foster Method (Alkire, et al., 2015). We employed a mixed-method approach. We conducted semi-structured interviews with experts worldwide (n=27) to identify their views about the relevance of having a child poverty measure, the main technical characteristics of the measure (e.g., dimensions and indicators) and sources of data that can be used to compute a Child MPI. In parallel, we conducted an in-depth review of the literature on child poverty and revised more than 30 available surveys for the 24 countries in the region (including Household Survey, Demographic Household Surveys and Multiple Indicator Cluster Surveys), looking for availability and comparability of indicators across datasets.

The results of the interviews and the literature review revealed that there is a wide consensus about the importance of measuring child poverty from a multidimensional perspective and the relevance for policy analysis. Also, considering that children have different needs in comparison to adults, it is important to design and calculate a measure considering child deprivations and using children as the unit of identification. Based on the results from the interviews and the literature review, the research team proposed an Ideal Child MPI that includes the following dimensions: health, nutrition, education, housing, water and sanitation, protective environments, social inclusion and economic security. Each dimension was defined according to the literature, and the final list of indicators were based on the results from the qualitative expert consultation process, and information available from previous studies in the region. Subsequently, the Ideal Child MPI was contrasted with the data available from each of the countries to determine feasibility to compute the Child MPI.

The process of identifying dimensions and indicators in existing surveys revealed four main data restrictions. Firstly, there is a critical data gap across all dimensions (except education) for children aged 5 to 15 years. Secondly, there is a lack of individual level child indicators, especially for child dimensions such as recreation, care and protection. Thirdly, there is a worrisome gap of indicators about individual health outcomes, which is more concerning for children older than 4 years. Finally, there is no comparable data to calculate the proposed Ideal Child MPI in LA and no single country had available information for all the dimensions and indicators identified as fundamental to measure child poverty.

Based on the identified data restrictions, three scenarios were proposed for computing a real Child MPI. These scenarios consider that we found no comparable data to calculate a comprehensive Child MPI in the Latin American region and no single country had available information for all the indicators. Calculation of these scenarios is our present work in progress, we expect preliminary results including sensitivity analyses to conclude by April 2018.

In conclusion, there is a lack of data sources including information for children in all age groups. Surveys in LA countries are not designed to capture child specific deprivations, aspect that limits the analysis of the levels of multidimensional poverty of children in the region, increasing the invisibility of children´s needs and vulnerabilities. Policy implications of our findings could focus on individual level data measurement and availability, especially regarding important and relevant indicators aspects such as health, child care, protection and recreation.

scroll to top