![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
This paper focuses on changes in multidimensional poverty, as measured by the Global Multidimensional Poverty Index (Global MPI) in Sub-Saharan Africa. Using data for 35 countries, we describe the changes in the level, intensity and composition of multidimensional poverty at the national level. For a subset of countries we discuss results at the sub-national level and provide a brief comparison to changes in income poverty. Our findings suggest that 30 countries, home to 92% of the population in our sample, significantly reduced multidimensional poverty as measured by the Global MPI for at least one comparison and significantly reduced the share of poor people. Looking within countries, we find different patterns of poverty reductions, with some countries reducing poverty for the poorest regions, while poorer regions in other countries do not seem to benefit from the general reduction in poverty to the same extent. When comparing trends in income and multidimensional poverty reduction we find significant differences, indicating that a holistic approach to poverty reduction should look at both, multidimensional and income poverty.
Notes
1 Data can be accessed online at http://www.ophi.org.uk/multidimensional-poverty-index/global-mpi-2016/.
2 Weights are usually assumed to sum up to one. In this case, the weighted sum of deprivations is the share of weighted deprivations a person experiences.
3 Population data are from the 2012 revision of the United Nations World Population Prospect Citation2015.
4 If the comparison is ambiguous because there are several for one country, we additionally report the time periods which we refer to.
5 See Alkire et al. (Citation2017) and Alkire, Jindra, Roche, and Vaz (Citation2016) for a detailed description of the harmonisation process. Results for countries with more recent data can additionally differ from Alkire et al. (Citation2015) because countries have then been strictly harmonised across three surveys. Thus, although most of the surveys are published as part of the Global MPI, results here can differ from previously published results due to the harmonisation.
6 Denote as the achievement matrix in
and
as the achievement matrix in
. The annualised absolute rate of change is the absolute rate of change divided by the difference between the two years
. The annualised relative rate of change is the compound rate of reduction in MPI between the starting and the end period
. The formulas apply to each of the partial indices as well.
7 The statistical analysis was done using Stata Version 13.1 (StataCorp, Citation2013a).
8 We use the dataset provided by the Quality of Government Institute (Teorell, Dahlberg, Holmberg, Rothstein, Khomenko, & Svenson, Citation2016) to merge GDP and income poverty data to the multidimensional poverty data. The world development indicators in this dataset were downloaded on the 02.11.2015 from http://go.worldbank.org/2EAGGLRZ40.
9 In case the headcount ratio H (or poverty gap) is missing, the value in year t is calculated by finding the two closest points and
where
and
for which we have observed
and
and
is then interpolated using
. Extrapolation uses the two closest points on the same side of t and the same formula.
10 If not mentioned otherwise, we use .
11 The timing of the surveys seems to matter for whether or not we find a significant reduction in . Countries with significant reductions have on average surveys which are 6.8 years apart, while countries with non-significant reductions have surveys which are on average 4.5 years apart. The difference is statistically significant
,
).
12 Again based on .
13 Assuming we have m groups and the population share of group l is given by , we can express MPI at the national level, given a specific achievement matrix X, as the population share weighted subgroup poverty levels:
.
14 The survey design does not allow to decompose by region for Zambia, South Africa, and the Comoros.
15 The results are descriptive as the data do not allow significance tests for the annualised changes.
16 We only extra- or interpolate the income headcount ratio if we have information on income poverty available that is less than four years apart from the years of the surveys used for estimating multidimensional poverty.
17 In case of the relationship between the multidimensional H and growth in GDP per capita we get . For income poverty we get
, t-value
.
![Supercharge Your Next Research Paper: Research and write your next paper with Jenni AI.]({"type":"image" , "src" : "/sda/112443/Skyscraper-socialsciences.png"})