An investigation of circumstantial and durational urban poverty by developing a new multidimensional measure, in the case of Harare province

Abstract

This study investigated circumstantial and durational urban poverty by adapting the UNDP MPI and developing a new multidimensional measure in Harare province. The study divided Harare province into 4 locations which are Harare rural, high density, medium density and low-density locations. The data used in this study was collected using a questionnaire employing a stratified simple random sampling across all 46 wards in all locations in Harare in order to investigate circumstantial and durational poverty in Harare. The study strove to objectively adapt the UNDP MPI and then generate a multidimensional poverty measure specifically for Harare province that can determine circumstantial and durational poverty for the city. Such a study has never done for Zimbabwe and Harare in particular and UNDP in general has not included durational aspect in its MPI. This makes the study unique in its contribution to both theory and practice. The study operationalised two broad objectives: theoretical and empirical objectives. The theoretical objectives were designed to critically review theories of poverty and their applicability in the context of Zimbabwe, to review and critique the existing measures of poverty and their applicability to Zimbabwe, to review theories on trends of poverty and how they contradict each other based on the different measures, to review and critique the existing durational poverty measures and their applicability in Zimbabwe and to review and critique literature on circumstantial poverty. The empirical objectives of the study were crafted to adapt the UNDP MPI and develop a multidimensional measure of circumstantial poverty, to determine the durational urban poverty for Harare province, to compare poverty among the locations in Harare using the developed multi-dimensional measure, to determine the perception of urban poverty in Harare province, to investigate the drivers of circumstantial urban poverty in Harare and ultimately offer recommendations on policies designed to reduce urban poverty in Zimbabwe. Quantitative research design was used and the data were collected using stratified simple random sampling from household heads or their spouses. To enable robustness of the study, a pilot study was administered at 11.4% of the sample target of 400 households. Open Data Kit (ODK) was used to collect data from the sample to the domain server of the study in real time. The study used UNDP MPI methodology to develop a new multidimensional poverty measure for Harare province. The multidimensional poverty measure that was developed for Harare uses household as a unit of identification. The study used 7 dimensions and 46 indicators to generate the multidimensional poverty measure for Harare. The dimensions and indicators used in this study were based on United Nations Sustainable Development Goals 2030 and various ongoing An investigation of circumstantial and durational urban poverty by developing a new multidimensional measure, the case of Harare province iii academic discourses on poverty measurement. The dimensions used are education, health, standards of living, economic security, empowerment and participation, environment and public infrastructure and food adaptation. The dimensions and indicators were allocated equal weights on the researcher’s strong convictions that all dimensions and indicators are equally important. Alkire and Foster (2011) methodology was used to arrive at the poverty head count and poverty intensity which are the seminal constructs of multidimensional poverty measure Descriptive statistics and logistic regression were used to ascertain results for the empirical objective of the study. The study first performed multicollinearity test using Variance of Inflation factor (VIF) values. Hosmer and Lemeshow test were subsequently performed to determine goodness-of-fit of the developed model of the study. This was done in order to determine whether the observed events produced by the model matched expected events. Perceptions of poverty were analysed using 14 indicators within 3 dimensions which are structuralist, individualistic and fatalistic. The overall Harare multidimensional poverty index developed for the study MPIoH was found to be 0.37, implying that 37% of people in Harare are multidimensionally poor. The multidimensional poverty index calculated for each location was Harare rural 0.50, high density 0.42%, medium density 0.25% and low density 0.16%. The results indicate that the highest multidimensional poverty is in Harare rural (ward 1) followed by wards in Harare high density locations, wards in Harare medium density and lastly Harare low density locations. When cut-off was changed (k=2, k=3, k=4, k=6, k=7), the multidimensional index produced the following results: 0.40, 0.40, 0.38, 0.37, 0.31, 0.19, 0.07 and 0.01 respectively. Analysis of multidimensional poverty when cut-offs are increased shows that multidimensional poverty decreases. The study reports that 73.8% of households that are currently not poor, are vulnerable to poverty at cut-off above 33% while 24% are vulnerable to poverty at cut- off ranging from 20% to 33%. The logistic regression found education, health, economic security, environmental and public infrastructure and food adaptability statistically significant in explaining circumstantial poverty in Harare province. Standards of living and empowerment and participation were found to be statistically insignificant. Increase in education was found to decrease probability of poverty by 0.3002 while increase in health was found to decrease likelihood of poverty by 2.019. Economic security improvement was found to reduce chances of poverty by 4.451. Environment and Public Infrastructure improvement would decrease the probability of poverty by 2.629 while Food security and adaptation increase the probability of reducing poverty by 3.139. An investigation of circumstantial and durational urban poverty by developing a new multidimensional measure, the case of Harare province

Indicators that were found to be significant in explaining poverty are 5-16 years not attending school, school materials/needs affordability, distance (km) from nearest health facility, house size, primary source of cooking fuel, employment status of household head, employment of other household members, household asset ownership, land ownership, nature of roads infrastructure, pollution, skipping meals and borrowing food, or relying on help from friends. The overall poverty for education dimension was found to be 33.8% in Harare province, Harare rural was the most deprived of education at 57.4%. High density was found to be 32.2% education poor while medium density and low density showed intriguing results of 25.5% and 26.2% respectively. Durational poverty was found to be increasing in Harare province for the period from 2010 to 2020 when poverty head count (P0), poverty gap (P1) and severity of poverty (P2) was analysed for the same period. The highest poverty head count was recorded in Harare rural and lowest in low density. Individualistic, structuralist and fatalistic perceptions were significant in explaining poverty in Harare province. The significant indicators for explaining poverty perceptions for Harare province are: poverty is inevitable and it is part of life; the society lacks social justice; the poor live-in places with no opportunities; the poor are poor because they experienced bad luck and finally households are poor because they have encountered misfortunes. The drivers of poverty in Harare province were found to be dropout from school by children aged 5-16 years, non-affordability of school materials/needs, long distance (km) to the nearest health facility, small household house infrastructure, poor primary source of cooking fuel, unemployment of household heads and other household members, non-land ownership, poor roads infrastructure, pollution exposure and insufficient food hence skipping meals and borrowing food, or relying on help from friends. The empirical objectives of adapting the UNDP MPI and developing a multidimensional measure of circumstantial poverty is relevant as new dimensions and indicators were used in this study. Furthermore, the second objective which determined the durational urban poverty for Harare province is rare and has not been used in Zimbabwe and Harare in particular for measuring multidimensional poverty. Thirdly, comparison of poverty among the locations in Harare using the developed multi-dimensional measure provides an informed understanding of circumstantial poverty in the city. The principal objective of determining the perception of urban poverty in Harare province helped to comprehend what the residents of Harare think poverty is, and this is helpful in informing the An investigation of circumstantial and durational urban poverty by developing a new multidimensional measure, the case of Harare province

right prescriptions required to fight poverty. The study found that lack of opportunities in poor locations, lack of social justice in the society, bad luck and misfortunate experiences formed perceptions of poverty in Harare province. The objective of investigating the drivers of circumstantial urban poverty in Harare generates information on what should be addressed in Harare in order to reduce poverty. Recommendations on policies that could be used to reduce urban poverty in Zimbabwe could allow policy makers to fine-tune overarching concerns and tame poverty. Academics and researchers’ policy recommendations gain additional epistemic horizons on poverty measurement.