External publications

Multidimensional poverty trends and horizontal inequalities: new insights from the G-CSPI database

Burchi, Francesco / Daniele Malerba / Nicole Rippin
External Publications (One Pager 434)

Brasilia: The International Policy Centre for Inclusive Growth (One Pager 434)

Volltext/Document

While poverty alleviation has always been at the heart of international development efforts, the 2030 Agenda introduced two significant shifts. First, poverty is no longer defined strictly as a lack of sufficient income, but rather as deprivation in several life domains. Second, the Agenda looks beyond national averages, and calls for a reduction of poverty for different social groups. This requires broadening the perspective through which poverty is analysed; and tools that provide detailed information on the multiple deprivations suffered by men and women, rural and urban people, as well as other groups.

This One-Pager presents the main findings of a study conducted by the German Development Institute (DIE) (Burchi et al., 2019). The paper uses a new indicator of multidimensional poverty, the Global Correlation Sensitive Poverty Index (G-CSPI), which includes three dimensions: education, (decent) employment and health (Burchi et al., 2018). This indicator has various advantages compared with existing measures. Among them, it can be decomposed into the three poverty components incidence, intensity, and inequality. The well-known MPI, instead, is decomposable only into the first two. In addition, the G-CSPI is an individual-level, rather than household-level, measure of poverty: therefore, no assumption has to be made with regard to intra-household distribution of resources. The latter feature is crucial for a sound gender-disaggregated analysis.

The study of Burchi et al. (2019) had three main objectives: (1) to analyse aggregate and country-level trends in multidimensional poverty and compare them with those in income poverty; (2) to explore the evolution of rural-urban differences in poverty over time; and (3) to assess gender disparities in poverty. The authors focus on the period starting from the establishment of the MDGs. Their analysis relies on a total sample of 60 low- and middle-income countries, for which data for multiple years are available. While the countries included in the study are located in different world regions, the largest coverage concerns sub-Saharan African, and Latin America and the Caribbean.

As for the first objective, both income poverty and multidimensional poverty, in aggregate terms, fell between 2000 and 2012. However, the decline in income poverty, in percentage terms, was twice as large as the decline in multidimensional poverty (-32% vs. -15%). There is significant regional heterogeneity. Multidimensional poverty declined the most in Asia, which has been converging towards the relatively low levels of poverty witnessed in Latin America and Europe. In contrast, sub-Saharan Africa experienced a very slow progress, which further distanced this region from the others. These findings point to the existence of poverty traps.

As for the second objective, the paper confirms that poverty is predominantly a rural phenomenon: in all countries rural poverty is higher than urban poverty. In aggregate terms, the rural G-CSPI is steadily more than four times the urban G-CSPI. This shows that the urban bias theorized in the 1970s by Lipton, according to which public resources are allocated disproportionally more in favour of urban compared to rural areas for political economy reasons, is still a major issue.

As for the third objective, no gender bias in 2000 at the global level is found. This contrasts with the claim made at the 1995 United Nations Fourth World Conference on Women in Beijing that 70% of the poor were women. However, since 2000 multidimensional poverty declined more among men (-18.5%) than women (-15%), indicating a minimal process of feminisation of poverty. This was triggered by the decline in employment poverty, which was much slower among women. Given that existing studies focusing on high-income or upper middle-income countries concluded that there was no evidence for a feminisation of poverty, this finding provides a substantial contribution to the literature.

In summary, the study of Burchi et al. (2019) has shown the potential of the G-CSPI database in tracking country, regional and global progress towards the achievement of SDG 1. Furthermore, the findings have relevant policy implications. First, the progress towards poverty eradication has not been as remarkable as suggested: poverty is still a big problem, especially in sub-Saharan Africa. This calls for renewed efforts in tackling the different forms of poverty, as policies successfully alleviating income poverty are not necessarily as effective in tackling multidimensional poverty. Second, the findings point to the need for new labour market policies in Africa, which could improve the quantity and the quality of employment, especially for women. Third, the study shows that most of the poor still live in rural areas. Despite the current emphasis on urbanisation, a considerable part of poverty-alleviating efforts should still focus on improving the living conditions of rural households.