1. Trang chủ
  2. » Kinh Doanh - Tiếp Thị

Measuring poverty and wellbeing in developing countries

372 68 0

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 372
Dung lượng 2,19 MB

Nội dung

Measuring Poverty and Wellbeing in Developing Countries United Nations University World Institute for Development Economics Research (UNU-WIDER) was established by the United Nations University as its first research and training centre and started work in Helsinki, Finland, in 1985 The mandate of the institute is to undertake applied research and policy analysis on structural changes affecting developing and transitional economies, to provide a forum for the advocacy of policies leading to robust, equitable, and environmentally sustainable growth, and to promote capacity strengthening and training in the field of economic and social policy-making Its work is carried out by staff researchers and visiting scholars in Helsinki and via networks of collaborating scholars and institutions around the world United Nations University World Institute for Development Economics Research (UNU-WIDER) Katajanokanlaituri 6B, 00160 Helsinki, Finland www.wider.unu.edu ‘This book makes accessible the recent advances in consumption and multidimensional poverty measurement The combination of literature review, computer code, and worked examples fill a major gap, making it possible for researchers in developing countries to estimate and analyse these metrics.’ John F Hoddinott, H.E Babcock Professor of Food and Nutrition Economics and Policy, Cornell University ‘This excellent volume combines theoretical discussion of the utility-consistent cost of basic needs poverty approach and first-order dominance multidimensional poverty analysis, empirical application, and practical tools in the form of user guides for estimation software essential reading for applied poverty researchers.’ Paul Shaffer, Department of International Development Studies, Trent University Measuring Poverty and Wellbeing in Developing Countries Edited by Channing Arndt and Finn Tarp A study prepared by the United Nations University World Institute for Development Economics Research (UNU-WIDER) Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © United Nations University World Institute for Development Economics Research (UNU-WIDER) 2017 The moral rights of the authors have been asserted First Edition published in 2017 Impression: Some rights reserved This is an open access publication Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-Non Commercial-Share Alike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO), a copy of which is available at https://creativecommons.org/licenses/by-nc-sa/3.0/igo/ It is permitted to reuse, share and adapt this work, subject to the following terms: Attribution - appropriate credit is given to the original work, the copyright holder and creator, and any changes made to the work are properly indicated Non-Commercial - the work, or any adaptation of the work, may not be used, distributed or reproduced in any format, by any means, for commercial purposes Share-Alike - the work, or any adaptation of the work is distributed under the same licence terms as the original, with a URL link provided to the licence Enquiries concerning use outside the terms of the Creative Commons licence should be sent to the Rights Department, Oxford University Press, at the above address or to academic.permissions@oup.com Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2016939850 ISBN 978–0–19–874480–1 (hbk.) 978–0–19–874481–8 (pbk.) Printed in Great Britain by Clays Ltd, St Ives plc Links to third party websites are provided by Oxford in good faith and for information only Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work Foreword Despite decades of research and advances in data and methodologies, measuring poverty and reconciling this with patterns of economic growth is a complex issue This contentiousness, and the fact that poverty remains widespread and persistent in sub-Saharan Africa (SSA) and in other parts of the globe, charged UNU-WIDER to launch in 2011 a major research project— Reconciling Africa’s Growth, Poverty, and Inequality Trends: Growth and Poverty Project (GAPP)—to re-examine growth, poverty, and inequality trends in SSA and in other developing regions Another key motivation for the GAPP project was that poverty analysis in developing countries remains, to a surprisingly high degree, an activity undertaken by technical assistance personnel and consultants based in developed countries This book was designed to enhance the transparency, replicability, and comparability of existing practice; and in so doing, it also aims to significantly lower the barriers to entry to the conduct of rigorous poverty measurement and increase the participation of analysts from developing countries in their own poverty assessment The book focuses on the measurement of absolute consumption poverty as well as a specific approach to multidimensional analysis of binary poverty indicators The intent is not to give the impression that these two domains alone are sufficient for rigorous poverty assessment On the contrary, the editors highlight that this book is designed to serve as a companion to the recently published volume entitled Growth and Poverty in Sub-Saharan Africa (Arndt, McKay, and Tarp 2016) That volume emphasizes repeatedly the desirability of the application of multiple approaches across multiple datasets combined with a concerted effort to triangulate results in order to develop a reasonably complete and coherent picture of living standards and their evolution as one moves across space or through time I hereby sincerely express my appreciation and admiration of the academic and analytical skills of the entire project team that made this volume possible and the detailed methodological expertise and knowledge of the case countries brought out so clearly It is my hope that the tools developed in this volume will be adopted by scholars and analysts in Africa, other developing Foreword regions, and beyond, in taking charge of the poverty analyses of developments in their respective countries The research project—Reconciling Africa’s Growth, Poverty, and Inequality Trends—was generously supported by the governments of Denmark, Finland, Sweden, and the United Kingdom, with a special project contribution additionally provided by the Finnish government UNU-WIDER gratefully acknowledges this vital research funding Finn Tarp Helsinki, October 2016 vi Acknowledgements UNU-WIDER’s Growth and Poverty Project (GAPP) brought together a highly qualified team of more than forty researchers from Africa and beyond Without their dedication and professional competence, this book and its less technical sibling would not have been possible We wish to express our sincere appreciation of all of the high-level academic input, together with the copious goodwill and patience—which were much needed when doing the original groundwork followed by numerous revisions and updates of the individual chapters A series of intensive planning meetings, involving many of the authors, helped shape the project, with the results presented at several UNU-WIDER development conferences and on many other occasions across African countries We are grateful to all of those who offered critique and most helpful comments They include Oxford University Press’s economics commissioning editor, Adam Swallow, and his team as well as three anonymous referees Their efforts were essential in helping to sharpen our research questions and approaches to analysing one of the most intricate challenges facing the development profession, the growth renaissance in developing countries and its impact on poverty reduction UNU-WIDER and its dedicated staff provided steady support, including research assistance, which goes far beyond the normal call of duty Particular thanks go to Dominik Etienne for excellent programming; Anne Ruohonen for consistent project assistance; Lorraine Telfer-Taivainen for all of the careful editorial and publication support on finalizing the book manuscript, including the many contacts with Oxford University Press; and the group of copy editors for helping to put out the numerous UNU-WIDER working papers produced during the course of the project Channing Arndt and Finn Tarp Helsinki, October 2016 OUP CORRECTED PROOF – FINAL, 21/11/2016, SPi Contents List of Figures List of Tables List of Boxes List of Abbreviations Notes on Contributors xi xiii xvii xix xxi Part I Principles and Choices Measuring Poverty and Wellbeing in Developing Countries: Motivation and Overview Channing Arndt and Finn Tarp Absolute Poverty Lines Channing Arndt, Kristi Mahrt, and Finn Tarp Multidimensional First-Order Dominance Comparisons of Population Wellbeing Nikolaj Siersbæk, Lars Peter Østerdal, and Channing Arndt Estimation in Practice Channing Arndt and Kristi Mahrt 10 24 40 Part II Country Applications Estimating Utility-Consistent Poverty in Ethiopia, 2000–11 David Stifel and Tassew Woldehanna 55 Estimating Utility-Consistent Poverty in Madagascar, 2001–10 David Stifel, Tiaray Razafimanantena, and Faly Rakotomanana 74 Methods Matter: The Sensitivity of Malawian Poverty Estimates to Definitions, Data, and Assumptions Ulrik Beck, Richard Mussa, and Karl Pauw A Review of Consumption Poverty Estimation for Mozambique Channing Arndt, Sam Jones, Kristi Mahrt, Vincenzo Salvucci, and Finn Tarp 88 108 Appendix B: User Guide to EFOD Table B5 Combination of welfare indicators, table_shares_1.csv Water 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 Sanit House Educ Info National 2004 National 2010 National_change 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 6.8726 6.1608 3.6046 5.6834 0.2131 0.6247 0.3519 2.0807 0.0000 0.0862 0.0000 0.0034 0.0452 0.0526 0.0032 0.5169 15.1621 14.9753 8.3381 13.1270 1.0865 3.4522 1.4996 10.1295 0.0353 0.2773 0.0353 0.1165 0.1068 0.6369 0.3204 4.4019 6.8797 5.7849 4.2612 6.5246 0.1921 0.8597 0.3719 2.4545 0.0167 0.0513 0.0318 0.0366 0.0023 0.0762 0.0682 1.1511 12.2475 12.1967 7.6366 12.8563 1.2374 2.8978 1.6262 10.2820 0.1238 0.1765 0.1065 0.5966 0.2910 0.8045 0.5954 7.5626 À13.2620 0.8490 À0.0889 3.4989 À0.1853 À0.1757 0.1551 1.3803 À0.0521 0.0141 0.0199 0.0194 À0.0078 0.0599 0.0292 0.9863 À14.6200 3.9844 À0.1913 7.3944 À0.2660 À1.2036 0.1844 3.7224 À0.1554 0.0974 0.0924 0.5338 0.2122 0.3230 0.4939 6.1580 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) Table B6 Number of deprivations, table_shares_1_num.csv num_ National National National_ dep 2004 2010 change 4.4019 11.7202 20.6382 31.2266 25.1405 6.8726 Rural 2004 Rural 2010 Rural_ change 7.5626 6.1580 0.8586 1.3157 1.0155 13.4295 6.0592 4.9436 7.3434 4.7315 20.5898 8.2659 19.1846 21.3244 12.9707 29.0360 6.8762 35.5143 34.2169 11.5780 22.5024 À14.0973 30.7850 27.2588 À13.8055 6.8797 À13.2620 8.7139 8.5409 À16.4902 Urban 2004 15.6553 33.2425 25.2549 17.6090 7.2138 1.0246 Urban 2010 Urban_ change 28.3526 23.1706 33.6845 10.0390 18.1451 À7.7582 11.7936 À8.7273 6.6730 À14.6567 1.3512 À2.0673 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) 333 Appendix B: User Guide to EFOD B.3.4 FOD 030_FOD_base.bat, 030_FOD_base.gms, 031_process1.bat, 031_process1.gms, 032_process2.bat, 032_process2.gms, 033_process3.bat, 033_process3.gms, 034_process4.bat, 034_process4.gms, 036_spatial.inc, 038_temporal.inc  Modify 030_FOD_base.bat and 030_FOD_base.gms FOD comparisons are conducted entirely in a linear program executed by GAMS The file 030_FOD_base.gms uses the dataset data_bs_100.csv and several include files to create required variables, equations, and parameters and save them to a base file A file for each processor, 031_process1.gms to 034_ process4.gms, then executes 036_spatial.inc and 038_temp.inc to conduct the FOD comparisons using the base file The user can execute the GAMS code in three ways First, FOD can be shelled directly from 000_master.do in Stata Second, the user can manually execute FOD in GAMS IDE Third, the user can execute FOD from a command window FOD involves a large number of comparisons that increases with the number of areas, survey years, population categories, and bootstrap iterations In order to reduce processing time, the FOD comparisons are divided by bootstrap iteration and executed using up to four processors It is possible to assign iterations to fewer processors depending on hardware capabilities The process time can be lengthy, even when taking advantage of four processors, and can vary from minutes to several hours B.3.5 FOD Tables 040_FOD_data, 042_Table_FODspat.do, 044_Table_FODtemp.do, 046_Table_Rank.do Depending on the number of processors utilized, GAMS saves up to four spatial (res_spat1.csv ) and four temporal (res_temp1.csv ) text files to the work directory The Stata do-file 040_FOD_data.do appends these files and creates two datasets (work/ res_spat.dta and work/res_temp.dta) From these datasets, three collections of tables are created that present temporal results, spatial results, and area rankings B.3.5.1 SPATIAL FOD TABLES The do-file 042_Table_FODspat.do creates spatial FOD tables for static and bootstrapped samples by area, category, and period FOD results are averaged across bootstrap iterations and are interpreted as the probability of domination A table is produced for static (Table B7) and bootstrap results (Table B8) Within each table, a blank cell indicates an indeterminate outcome between the row and column area In static tables, a ‘1’ indicates the row (column) area dominates (is dominated by) the column (row) area In bootstrap tables, values indicate the estimated probability that the row (column) area dominates (is dominated by) the column (row) area (probability is defined as total number of iterations where a domination outcome occurs divided by the total number of bootstrap iterations) The row (column) average yields the average number of times or the average probability that the row (column) area dominates (is dominated by) all other areas for the static and bootstrap cases, respectively Bootstrap sampling introduces variation to the results and therefore small values should be interpreted with caution For instance, Table B8 indicates that the nation 334 Table B7 Spatial FOD results (static), FOD_spat_1_1_static.csv Area National Rural Urban Central Eastern Lake Northern S_Highlands Southern Western Zanzibar Average National Rural Urban Central Eastern Lake Northern S_Highlands Southern Western Zanzibar Average 0.1 0.9 0.7 0.3 0 0.7 0.2455 1 1 1 1 0.5 1 1 0.3 1 1 1 1 0.3 0.1 0.4 0.1 0.3 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) 0.3 0.4 Table B8 Spatial FOD results (bootstrap), FOD_spat_1_1_boot.csv Area National Rural Urban Central Eastern Lake Northern S_Highlands Southern Western Zanzibar Average National Rural Urban 1 0.79 0.13 0.3 0.222 Eastern Lake 0.74 0.56 0.41 0.01 0.09 0.79 0.06 0.02 0.47 0.05 0.05 0.01 0.96 0.17 0.97 0.04 0.57 0.18 0.03 0.77 0.473 Central 0.03 0.42 0.211 0.06 0.08 0.02 0.42 0.326 Northern 0.95 0.22 0.06 0.123 S_Highlands Southern Western 0.26 0.04 0.33 0.04 0.86 0.01 0.3 0.25 0.91 0.01 0.07 0.03 0.99 0.1 0.69 0.83 0.314 0.4 0.289 0.01 0.38 0.286 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) Zanzibar 0.16 0.01 0.35 0.03 0.017 Average 0.209 0.002 0.88 0.065 0.58 0.006 0.195 0.031 0.009 0.364 0.2128 Appendix B: User Guide to EFOD dominates Central with a probability of 0.05, which is likely too small to conclude that the nation outperforms Central B.3.5.2 TEMPORAL FOD TABLES The do-file 044_Table_FODtemp.do creates temporal and net temporal FOD tables for each category Temporal tables present static and bootstrap results for all year combinations for each area As in spatial analysis, bootstrapped FOD results are averaged across iterations and are interpreted as the probability of domination Temporal results are presented in two ways First, FOD_temp_$cat.csv presents static and bootstrap temporal outcomes for both later years dominating earlier years and earlier years dominating later years (Table B9) Second, FOD_net_temp_$cat.csv, presents net temporal domination, which measures the difference in the probabilities of later years dominating earlier years and earlier years dominating later years (Table B10) In the case of no welfare regression, net results are equivalent to the results for later years dominating earlier years In static temporal columns, a ‘1’ indicates that a given year dominated the other year, while a blank cell indicates the given year did not dominate the other year When both years have a blank entry, FOD was indeterminate In the net temporal table, ‘1’ indicates the later year dominated the earlier year; a blank cell indicates FOD was indeterminate; and, ‘À1’ indicates the earlier year dominated the later year There is no difference in the amount of information in the static temporal and the static net temporal tables, rather the difference lies in the presentation In the bootstrap temporal columns, entries indicate the probability that a given year dominates the other year A blank indicates the year did not dominate in any iteration and ‘1’ indicates that year dominated in every iteration When both years have a blank entry, FOD was indeterminate in all cases In the net temporal table, positive probabilities indicate that the later year dominated in more iterations than the earlier year, and negative probabilities indicate that the earlier year dominated in more iterations than the later year A net result of 0.2 could mean that the later year dominated in 20 per cent of iterations, the earlier year never dominated, and 80 per cent of the iterations were indeterminate Or, for example, it could mean that the later year dominated in 60 per cent of the iterations, and the earlier year dominated in 40 per cent of the iterations The exact scenario should be determined by the user Similarly, a blank cell could indicate that the outcome was indeterminate in every iteration or that each year dominated with the same frequency Thus, in cases with frequent backsliding, the full temporal table provides a more complete story than the net temporal table For example in Table B9, both 1996 dominates 1992 and 1992 dominates 1996 with positive probabilities in Northern and Southern Highlands In Table B10, with the exception of Northern and Southern Highlands, static and bootstrap net domination results are the same as those in Table B9 Net domination is different in the case of Northern and Southern Highlands because there are small probabilities of 1992 dominating 1996 B.3.5.3 AREA RANKINGS Area ranking tables use spatial bootstrap FOD results to compare areas based on the net probability of domination, which measures the average probability that an area 337 Table B9 Temporal FOD results, FOD_temp_1.csv Area National Rural Urban Central Eastern Lake Northern S_Highlands Southern Western Zanzibar stat_ 1992 1996 stat_ 1992 2004 stat_ 1996 1992 1 1 1 stat_ 1996 2004 stat_ 2004 1992 stat_ 2004 1996 1 1 1 1 1 1 boot_ 1992 1996 boot_ 1992 2004 boot_ 1996 1992 0.3 0.28 0.22 0.14 0.35 0.05 0.01 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) 0.02 0.13 0.06 0.27 0.22 boot_ 1996 2004 boot_ 2004 1992 boot_ 2004 1996 0.99 0.71 0.11 0.14 0.42 0.62 0.68 0.67 0.55 0.23 0.99 0.95 0.45 0.03 0.13 0.18 0.17 0.85 0.46 0.69 0.12 0.86 Appendix B: User Guide to EFOD Table B10 Net temporal FOD results, FOD_net_temp_1.csv Area stat_ 1996 1992 stat_ 2004 1992 stat_ 2004 1996 boot_ 1996 1992 boot_ 2004 1992 boot_ 2004 1996 1 1 1 1 1 0.3 0.28 0.22 0.14 0.35 1 1 À0.03 0.12 0.06 0.27 0.22 0.99 0.71 0.11 0.14 0.42 0.62 0.68 0.67 0.55 0.23 0.99 0.95 0.45 0.03 0.13 0.18 0.17 0.85 0.46 0.69 0.12 0.86 National Rural Urban Central Eastern Lake Northern S_Highlands Southern Western Zanzibar 1 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) Table B11 FOD rankings, table_rank_1.csv Area Eastern Zanzibar S_Highlands Northern Lake Southern Western Central Net Domination 2004 PNet Domination 2004 Rank PNet Domination 2010 Rank 2004 Net Domination 2010 518 208 À154 38 À97 À62 À191 À260 0.74 0.2971429 À0.22 0.0542857 À0.1385714 À0.0885714 À0.2728571 À0.3714286 504 144 À17 À54 À80 À83 À143 À271 0.72 0.2057143 À0.0242857 À0.0771429 À0.1142857 À0.1185714 À0.2042857 À0.3871429 Change 2010 0 À2 À2 À2 Source: Based on calculations in Arndt et al (2014) using the 2004/5, 2010 TDHS (National Bureau of Statistics and Macro 2005, 2011) dominates all other areas minus the average probability of the same area being dominated by all other areas If the same areas are presented in the spatial and rank tables, the probability of net domination is equivalent to the spatial ranking row average minus the column average The do-file 046_Table_Rank.do produces separate tables for each category See Table B11 for an example of ranking outcomes Ranking results should be interpreted carefully Because bootstrapping results may vary from one execution of FOD to the next, rankings may be sensitive to small perturbations The difference in net domination scores is often insufficiently large to distinguish between differences in welfare outcomes and variability 339 Appendix B: User Guide to EFOD introduced through random bootstrapping For example, in Table B11, the difference in net domination between Lake and Southern in 2010 is extremely small However, even the difference between Northern and Lake may not be robust to bootstrap variation B.4 Alternative Specifications Thus far, the language in this description of the EFOD software has been geared towards welfare analysis of areas over time However, EFOD is flexible and can be applied to alternative specifications This section provides three examples of possible variations.6  To this point, the discussion has focused on analysis within a single country Alternatively, welfare comparisons can be made internationally With comparable indicators, areas could be specified as individual countries yielding spatial FOD comparisons between countries and temporal FOD comparisons for each country  Thus far, population groups have been discussed independently of each other However, FOD comparisons can be made between populations if the analyst defines the area parameters to specify population groups instead of areas See Chapter 14 where Mahrt and Masumbu specify FOD comparisons in Zambia by rural economic activity and urban housing cost areas One area variable would now classify the different population groups, similar to the category variable in the standard format If areas are still of interest, additional area variables can also be used to compare the population groups to aggregate areas such as urban/rural  In analysis focused on a single population group, say households, the category variable could be used to specify different sets of indicators In this context, the category variable would serve merely to signal each set of indicators rather than defining different populations For example, category one could include a set of health indicators, category two could contain a set of shelter indicators, and category three could contain a set of education indicators For a given set of indicators, spatial FOD comparisons would be made across areas and temporal analysis over time for each area FOD analysis within each indicator category would be conducted independently, thus highlighting the relative performance of areas for each set of indicators References Arndt C., R Distante, M A Hussain, L P Østerdal, P Huong, and M Ibraimo (2012) ‘Ordinal Welfare Comparisons with Multiple Discrete Indicators: A First-Order Note that the FOD code stream requires area to specify the entire population, and must continue to so 340 Appendix B: User Guide to EFOD Dominance Approach and Application to Child Poverty’, World Development, 40(11): 2290–301 Arndt, C., V Leyaro, and K Mahrt (2014) ‘Multi-dimensional Poverty Analysis for Tanzania: First-Order Dominance Approach with Discrete Indicators’, WIDER Working Paper 2014/146 Helsinki: UNU-WIDER National Bureau of Statistics (NBS) (Tanzania) and ICF Macro (2011) Tanzania Demographic and Health Survey 2010 Dar es Salaam: NBS and ICF Macro National Bureau of Statistics (NBS) (Tanzania) and ORC Macro (2005) Tanzania Demographic and Health Survey 2004–05 Dar es Salaam: NBS and ORC Macro 341 Index Note 1: Tables, figures and boxes are indicated by an italic t, f, or b following the page number Note 2: As most chapters pertain to a particular country, the sub-headings for each country have not been double-entered as main entries, so the reader is advised to locate main entries of interest under the country headings Note 3: The following abbreviations have been used in sub-headings: FOD: first-order dominance PLEASe: Poverty Line Estimation Analytical Software absolute poverty lines, estimation of 10–11, 22, 269–71 challenges of 269–70 consistency and specificity 12, 16–20, 152–3 consumption bundles 12 cost of basic needs (CBN) approach 13–16 data shortcomings 270 definition 11–12 food energy intake (FEI) approach 13 revealed preferences 19–20, 153 specific utility-consistent poverty lines 21–2 steps in 144–5 substitution effects 16–17 f unidimensional/multidimensional approaches 178–9 utility 11–13 Alkire, S 179 Alkire-Foster (AF) multidimensional index 217–18, 302 Appleton, S 143, 149 Arndt, C 21, 22, 56, 153, 276 Atkinson, A B 25 Bidani, B 123, 151, 152 bootstrapping 33–4 Bourguignon, F 25 child poverty, measurement of 179–80 see also Democratic Republic of Congo, estimating childhood poverty in; Ghana, estimating child poverty in; Tanzania, multidimensional assessment of child welfare climate, and energy requirements 20 consistency and specificity 12, 16–20, 152–3 consumption, and poverty estimation 41–2 consumption surveys avoiding excessive complexity 299–300 need for increased frequency of 298–9 cost of basic needs (CBN) approach 13–16, 144, 280 Daniels, L 144 Democratic Republic of Congo, estimating childhood poverty in 160–1, 175–6 children not deprived by welfare indicators 164–5 t, 166 definitions of child deprivation indicators 164 Demographic and Health Survey (DHS) 163 Enquête 1-2-3 163 first-order dominance (FOD) approach 160, 161 first-order dominance (FOD) indicators 163–6 indeterminate outcomes in FOD 161–2 f, 163 Multiple Indicator Cluster Survey (MICS) 163 results using FOD approach 166–75 spatial FOD bootstrap comparisons 169, 170–3 t spatial rankings 169–74 t, 175 t survey data 163 temporal net FOD comparisons 166 t, 167, 168 t Emwanu, T 140 Estimating First-Order Dominance (EFOD) software 4, 40, 48, 271, 297 Index Estimating First-Order Dominance (EFOD) software (cont.) implementation 50 indicators 48–50 output and interpretation 51 user guide 325–40 see also first-order dominance (FOD) Ethiopia, estimating poverty in 55–6, 65–72 application of PLEASe software 59–61 Central Statistics Agency (CSA) 56, 58 cumulative distributions of household per capita consumption 61 f data preparation 59–60 data sources 58–9 decline in poverty 61–2, 72 differences from official estimates 62–5 food poverty lines 57 household food consumption baskets by spatial domains 66–71 t Household Income, Consumption and Expenditure Surveys (HICES) 58 methodology 56–7 minimum calorie requirements 63–5 t PLEASe code preparation 60–1 poverty estimates 61–2 t, 63–4 t, 65, 72 Ferreira, F H G 37 first-order dominance (FOD) 4, 5, 26, 36–7, 271–3 alternative dominance criterion 35–6 assessment of approach 271–3 bootstrapping approach 33–4 checking multidimensional FOD 31–2 checking one-dimensional FOD 28–9 detecting in practice 32–3 faster solution algorithms 34–5 limitations 26, 33 mitigating limitations of 33–4 multidimensional FOD 29–33 notations and definitions 27–8, 29–31 one-dimensional FOD 27–9 theory and examples 27–34 see also Estimating First-Order Dominance (EFOD) software food energy intake (FEI) approach 13, 144 food poverty lines 42–4 Foster, Greer and Thorbecke (FGT) poverty measures 14, 45, 57, 76, 123, 145 General Algebraic Modelling Systems (GAMS) 40, 41, 48, 50, 183 Ghana, estimating child poverty in 191–2 approaches to measuring 179–80 bootstrap sampling 182, 183 Bristol (headcount) approach to 179 344 children by combination of welfare indicators 185, 187 t children not deprived by welfare indicators 185, 186 t comparison between deprivation, income and consumption expenditure poverty 189–90 t, 191 data sources 184 definitions of child deprivation indicators 182–3 Ghana Demography Health Survey 181 Ghana Living Standards Survey (GLSS) 180, 184 Ghana Statistical Service 184 income-based approach 183–4 multidimensional first-order dominance (FOD) approach 180, 181–3 Multiple Indicator Cluster Survey (MICS) 181 poverty reduction 178 spatial FOD comparisons 188–9 t studies of 181 temporal FOD comparisons 185–8 t Global Trade Analysis Project (GTAP) (Purdue University) Gordon, D 164, 179 inequality, measuring in developing countries 274–5, 292–3 approaches to measuring poverty 280–1 composition consumer price indices by country 285–7, 288 t composition effect 274, 276–8, 284–7, 292, 293 consumption aggregates 274 consumption bundles 276–8 consumption shares by consumption percentiles 285, 286 t data sources 275, 281–2 t, 283–4 deflated consumption aggregate 280 diversity in inequality across countries 283 diversity of country experiences 283 food and non-food consumer price indices 284 t Gini coefficients using alternative deflators 287–90 t, 291 inequality and poverty 287–92 poverty rates using different inequality measures 291 t, 292 quantity discounting effect 274, 278–80, 287, 289 t, 292, 293 SiMP methodology 275, 281 International Food Policy Research Institute (IFPRI) 7, 92 Levine, S 141, 143 Lokshin, M 20, 21, 153 Index Madagascar, estimating poverty in 74–5, 85–7 application of PLEASe software 77–80 consumption baskets 85, 86 t, 87 data preparation 77–9 data sources 76–7 dealing with extreme values 78 differences from INSTAT estimates 80–4, 85–7 Enquête Périodique auprès des Ménages (EPM) 76 food poverty lines 76, 84 Institut National de la Statistique (INSTAT) 75, 76 methodology 75–6 minimum calorie requirements 84, 85 t PLEASe code preparation 79–80 poverty estimates 80–1 t, 82–5 spatial domains 77–8, 81–2 t Mahrt, K 164 Malawi, estimating poverty in 88–9, 105–6 adjustments of the PLEASe methodology 91 b adjustment to PLEASe code 93 b baseline estimates 91–2 consumption aggregates 99–100 t differences from National Statistical Office estimates 88–9 food bundles 102–3 t, 104–5 food consumption conversion factors 92, 94 food poverty lines 94 Integrated Household Survey (IHS) 88, 105 methodological choices investigated 90 t methodological consistency with National Statistical Office approach 89–90 methodological differences with National Statistical Office approach 90–7 National Statistical Office (NSO) 88 poverty headcount rates under different methodological choices 100–1 t, 102 poverty lines under different methodological choices 97–8 t, 99 regional poverty lines 94, 98 temporal changes in food basket composition 95 temporal changes in non-food consumption 95–6 f, 97 using survey prices to update poverty lines 94–5, 99 utility consistency 94 Malik, S J 122, 127–8 Minot, N 144 Minujin, A 179 Mozambique, estimating poverty in 108–9, 118–19 challenges facing first national assessment 109 comparison of official and PLEASe estimates 111 t, 114 correlations between official and PLEASe estimates 115 t cost of basic needs (CBN) approach 110, 115 differences between official and PLEASe approaches 115–17 methodological choices in national surveys 109–14 national surveys 108 PLEASe estimates 114–18, 119 regional variations 110 revealed preference conditions 116 similarity of official and PLEASe approaches 115 trends in 119 Multidimensional Poverty Index (MPI) 180, 217 multidimensional welfare 180 comparing approaches to multidimensional analysis 218–19 measurement of 24–6, 216–19 Multiple Overlapping Deprivation Analysis (MODA) 180 Museveni, Y 156 Nanivazo, M 164 Nigeria, estimating poverty in 194–5, 213–14 bootstrap sampling 195–6 data sources 196 Demographic and Health Survey (DHS) 196 first-order dominance (FOD) approach 195–6 first-order dominance (FOD) welfare indicators 196–8 geographical zones of Nigeria 196, 197 f Multiple Indicator Cluster Survey (MICS) 198 National Bureau of Statistics (NBS) 194 regional inequalities 195, 213 spatial FOD comparisons 203–5 t spatial sensitivity analysis 205–6 t, 207 state-level FOD results 207, 208–12 f, 213 temporal net FOD comparisons 200–1 t, 202 t temporal sensitivity analysis 201–3 welfare indicator results 198–9 t, 200 World Bank estimates 194 Pakistan, estimating poverty in 133–4 calorie requirement calculation 130 cost of basic needs (CBN) approach 123–4 data preparation 126 data sources 124–6 evolution of estimates of 121–2 food energy intake (FEI) approach 123 Household Integrated Economic Survey (HIES) 122, 124–6 345 Index Pakistan, estimating poverty in (cont.) inflation adjustment 123 methodologies 123–4 modified PLEASe approach 129–31 national poverty headcounts 131 f Pakistan Bureau of Statistics 124 problems with using consumer price index (CPI) 122, 128 results using food energy intake (FEI) approach 128–9 t, 130 t, 132 t, 135–6 t results using official methodology 133 t results using PLEASe methodology 131–2 t, 133 t, 136–7 t revised estimation methodology 122 rural/urban differences 131 sample population 130 shortcomings in estimates of 122 spatial domains 126 trends in poverty indicators 127 t using official methodology 127–8 paternalism 15 Pinkovskiy, M 275, 281 poverty analysis, capability-building 298 avoiding excessive complexity 299–300 coming to grips with price trends 300–1 comparability issues 300–1 increased frequency of consumption surveys 298–9 using variety of methods 301–2 Poverty Line Estimation Analytical Software (PLEASe) 4, 5, 6, 40, 297 consumption 41–2 features 40–1 food poverty lines 42–4 non-food poverty lines 44–5 poverty measurement 45–6 user guide 305–23 utility consistency 46–7 Range, T M 35 Ravallion, M 10, 11, 12, 20, 21, 123, 151, 152, 153, 183, 302 relative poverty lines 11 Roach, J M 179 Russia 153 Sala-i-Martin, X 275, 281 Sen, A 24 Simler, K R 21, 22, 56, 153 Tanzania, multidimensional assessment of child welfare Alkire-Foster (AF) approach 217–18 Alkire-Foster (AF) approach results 233–4 t, 235 f, 236, 237 t bootstrap sampling 217 346 comparing approaches to multidimensional analysis 218–19, 238–9 comparing FOD and Alkire-Foster results 236–8 t data sources 219–20 first-order dominance (FOD) approach 216–17 multidimensional poverty measurement 216–19 poverty assessment studies 215–16 spatial FOD comparisons 228, 229–30 t, 231 spatial FOD rankings 231, 232–3 t Tanzania Demographic and Health Surveys (TDHS) 216, 219–20 temporal net FOD comparisons 225, 226–7 t, 228 trends in deprivation by welfare indicator 221–2 f, 223–4 t, 225 welfare indicators 220–1 t World Bank poverty assessment 215, 216, 300 Uganda, estimating poverty in 140–1, 156–7 accounting for local diets 150–1 aggregation 155–6 average calorie requirements by spatial domain 149–50 t constructing welfare indicator 145–9 cost of basic needs (CBN) approach 149–52 cost per kilo of staple crops 150, 152 f data sources 145 Demographic and Health Surveys (DHS) 144 density estimates for welfare indicators 148 f diverging views on levels of 140–1 diversity in diets 150–1 f estimated vs official poverty lines 154–5 t headcount poverty estimates 155–6 t increased inequality 143 official estimates 141–2 t, 143–4 poverty lines for each spatial domain 153–4 t poverty reduction 140 problems with official estimates 141, 143–4 regional variations 156 t revealed preference approach 153 spatial domains 152 steps in measuring poverty 144–5 Uganda Bureau of Statistics (UBOS) 141–2, 145 Uganda Census of Agriculture 147 Uganda National Household Survey 145 Uganda National Panel Survey (UNPS) 145 utility consistency 152–4 variation in poverty reduction rates 142–3 United Nations Children’s Fund (UNICEF) 179 Global Study on Child Poverty and Disparity 181 Index United Nations Development Programme (UNDP), Multidimensional Poverty Index (MPI) 180, 217 United Nations University World Institute for Development Economics Research (UNU-WIDER) Growth and African Poverty Project (GAPP) 4–5, 275, 281 World Income Inequality Database (WIID) 283 Verduzco-Gallo, I 92, 104, 105 wellbeing, measurement of 24–5, 36–7 Zambia, estimating poverty in 242–3, 263–4 agricultural support programmes 245 area rankings by probability of net domination 254, 255–6 t, 257 bootstrap spatial FOD comparisons (sanitation) 261–2 t Central Statistical Office (CSO) 246 consumption poverty headcount rates 244 t data sources 246–7 deprivation by sanitation indicator 258 t economic background 242, 263 Farmer Input Support Programme (FISP) 245 first-order dominance (FOD) approach 246 first-order dominance (FOD) welfare indicators 247 t, 248 Food Reserve Agency (FRA) 245 household stratum comparisons 257 indeterminate outcomes in FOD 261 levels of deprivation 249, 250 t Living Conditions Monitoring Surveys (LCMS) 243, 246–7 National Development Plans 243, 247 Poverty Reduction Strategy Paper (PRSP) 243 public service delivery 245 rural poverty context 243–5, 263 sanitation indicators 248 t, 257–61 sensitivity of FOD outcomes to indicator definition 257–61, 262–4 spatial FOD comparisons 251, 252–3 t, 254 spatial FOD comparisons by sanitation indicator 259–60 t, 261 temporal net FOD comparisons 249–51t temporal net FOD comparisons by sanitation indicator 259 t urban and rural poverty trends 244 f welfare definition 243 sterdal, L P 35 347 ... Principles and Choices Measuring Poverty and Wellbeing in Developing Countries: Motivation and Overview Channing Arndt and Finn Tarp Absolute Poverty Lines Channing Arndt, Kristi Mahrt, and Finn... primary research interests are food security, employment, child welfare and poverty, education, and health xxiv Part I Principles and Choices Measuring Poverty and Wellbeing in Developing Countries. .. project— Reconciling Africa’s Growth, Poverty, and Inequality Trends: Growth and Poverty Project (GAPP)—to re-examine growth, poverty, and inequality trends in SSA and in other developing regions

Ngày đăng: 03/03/2020, 10:44

TỪ KHÓA LIÊN QUAN