JhqEqWEG

1%\documentclass[english]{article}
2\documentclass[12pt, a4paper, oneside, english]{article}
3\usepackage[utf8x]{inputenc} 
4\usepackage[T1]{fontenc}
5%\usepackage{lmodern}llllllll
6\usepackage{babel} % English language/hyphenation
7\usepackage{textcomp}
8\usepackage{xfrac}
9\usepackage{xcolor}
10\usepackage{lipsum}
11\usepackage{sectsty} % Allows customizing section commands
12\usepackage{blkarray, tabularx, longtable, multirow,threeparttable, makecell, booktabs, hhline, caption}
13\theadset{\renewcommand{\arraystretch}{0.6}}
14\renewcommand\theadfont{\normalsize}
15\newcolumntype{C}{>{\centering\arraybackslash}X} % centered version of "X" column type
16% \newcolumntype{Y}{>{\centering\arraybackslash}X}
17\renewcommand\arraystretch{1.5}
18\setlength{\tabcolsep}{5pt} % default is 6pt
19\usepackage{siunitx}
20\newcommand*{\head}[1]{\textbf{#1}}% to make heads of table in BOLD, it is used with booktabs package and \head
21%\usepackage{url} \makeatletter \g@addto@macro{\UrlBreaks}{\UrlOrds} \makeatother
22\usepackage[justification=centering]{caption}
23\usepackage{csvsimple}
24\usepackage{pdflscape}
25\usepackage{amsmath, amsfonts, amsthm} % needed for command eqref or Math packages
26\usepackage{bigstrut} % needed for the command \bigstrut
27\usepackage{enumerate} % needed for some options in enumerate
28\usepackage{makeidx} % needed for creating an index
29\usepackage{blindtext} % needed for creating dummy text passage
30\usepackage{fancyhdr}
31\usepackage{rotating}
32\usepackage{multicol}
33\captionsetup[figure]{font=footnotesize,
34                    labelsep=period}
35\captionsetup[table]{font={footnotesize, sc},
36                      labelsep=newline,
37                      justification=centering}
38
39\usepackage{tikz, pgfplots, pgfplotstable} %to plot figure
40\makeatletter
41\long\def\ifnodedefined#1#2#3{%
42    \@ifundefined{pgf@sh@ns@#1}{#3}{#2}%
43}
44
45\pgfplotsset{
46    discontinuous/.style={
47    scatter,
48    scatter/@pre marker code/.code={
49        \ifnodedefined{marker}{
50            \pgfpointdiff{\pgfpointanchor{marker}{center}}%
51             {\pgfpoint{0}{0}}%
52             \ifdim\pgf@y>0pt
53                \tikzset{options/.style={mark=*, fill=white}}
54                \draw [densely dashed,blue] (marker-|0,0) -- (0,0);
55                \draw plot [mark=*] coordinates {(marker-|0,0)};
56             \else
57                \tikzset{options/.style={mark=none}}
58             \fi
59        }{
60            \tikzset{options/.style={mark=none}}
61        }
62        \coordinate (marker) at (0,0);
63        \begin{scope}[options]
64    },
65    scatter/@post marker code/.code={\end{scope}}
66    }
67}
68
69\makeatother
70
71%% page settings%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
72\usepackage[a4paper, top=2cm, bottom=1.8cm, left=2.5cm, right=2.5cm]{geometry} % needed for page border stteings
73\parindent=0cm % for spece of first line of new text block
74\sloppy % for writing with hypenless justification (tries to)
75%\usepackage[round, authoryear]{natbib} % needed for cite and abbrvnat bibliograph style
76
77%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
78\usepackage[
79    colorlinks=true
80   ,breaklinks
81   %, ngerman
82    ] {hyperref} % needed for creating hyperlinks in the docuent, the option colorlink=true gets rid of the awful boxes, breaklinks breaks lonk links (list of figures), and the ngerman sets everything for german s default hyperlink languages.
83\definecolor {c1}{rgb}{0,0, 1} % blue
84\definecolor {c2}{rgb}{0.3,0.3, 0.9} % light blue
85\definecolor {c3}{rgb}{0.3,0.3, 0.9} % red blue
86\hypersetup{
87   linkcolor={c1}, % internal links
88    citecolor={c2}, % citations
89    urlcolor={c3} % external links/urls
90}
91\usepackage{cleveref}%% to be loaded after hyperref
92%\parindent 0ex
93\pagestyle{fancy}
94\fancyhf{}
95\fancyhead {}
96\frenchspacing
97\fancyfoot{}
98\fancyhead[L]{\slshape \MakeUppercase {Multidimensional Poverty Index (MPI) in Ethiopia}}
99\fancyfoot[C]{\thepage}
100%\renewcommand{\headrulewidth}{0pt}
101\renewcommand{\footrulewidth}{0pt}
102\renewcommand{\baselinestretch}{1.5}%This affects the in spacing of my paragraphs
103%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
104\theadset{\renewcommand{\arraystretch}{0.6}}
105\renewcommand\theadfont{\small}
106\renewcommand\theadalign{bc}
107%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
108
109%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
110\begin{document}
111%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
112%\input{Section_1}
113%\input{Section_2}
114%\input{Section_3}
115%\input{Section_4}
116%\input{Section_5}
117
118%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
119\graphicspath{ {images/} }
120\begin{titlepage}
121\begin{center}
122\vspace*{5cm}
123\pagestyle{empty}
124\pagestyle{plain}
125\Large {\textbf{ MULTIDIMENSIONAL POVERTY INDEX\\ IN ETHIOPIA (2011/12): ROBUSTNESS TEST,\\ DECOMPOSITION, AND MAPPING }}\\
126%\Large{\textbf{Multidimensional Poverty Index (MPI) in Ethiopia}}\\
127
128\vfill
129\line(1,0){400}\\ [1mm]%automatically putting vetical space btween the two elments to fill the page
130\huge{\textbf{A Ph.D. Thesis:\\ University of KwaZulu-Natal\\ School of Accounting, Economics and Finance\\ Department of Economics }}\\[3mm]
131%\lipsum[1]
132%{\setstretch{1.0}\color{blue}
133%\lipsum[2]
134%}
135%\lipsum[3]
136%\Large{\textbf{-This is a sample subtitle-}}\\[1mm]
137\line(1,0){400}\\
138\vfill    %Now it is centered at the page
139By WELDESLASSIE HAILAI\\
140Candidate\#:{214585063}\\
141Supervisor Dr. Claire Vermaak\\
142%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
143\date {\today}
144%\maketitle
145\end{center}
146\end{titlepage}
147\tableofcontents
148\thispagestyle{empty} %This helps you to remove the page # in the table of contents part
149\clearpage
150%\chapter{Background and Justification}
151\pagestyle{fancy}
152%\fancyhf{}
153%\fancyhead[OC]{\leftmark}
154%\fancyhead[EC]{\rightmark}
155%\renewcommand{\footrulewidth}{1pt} % needed to add a line at the bottom
156\cfoot{\thepage}
157\setcounter{page}{1} %Now the f
158%\chapter{doc one}
159%\input{document_1_body}
160%\chapter{doc two}
161%\input{document_2_body}
162%\chapter{doc three}
163%\input{document_3_body}
164\section{Background and Justifications}
165
166\subsection{Introduction:Why  Multidimensional Poverty Index(MPI)?}  
167There is a synchronization of thoughts in broad literature in the study of well-being and/or poverty measurement study and analysis practice that well-being is multidimensional and the overlapping deprivation and destitution that a poor person suffer should be studied in a multidimensional perspective(Alikre and Foster, 2011; Alkire et al.,2015; Sen, 1985; Atkinson, 2003; Ferreira and Lugo, 2013.\\
168According to Decancq et al.,(2014),even those that are suitably deflated prices for temporal and geographical differences in the uni-dimensional poverty measurement approach using income or consumption do not qualify as good proxies for two main reasons. They argue that individuals may suffer from rationing since private goods market as well as labor markets may fail to be competitive. Moreover, some essential social and public goods such as education, health and security services may not be private and marketable. \\
169
170 
171Henceforth, MPI is a holistic and big picture view of overlapping  multifaceted human deprivations and poverty. Forthwith, it synthesis the wide ranging deprivation and poverty indicators in to meaningful and comprehensive  measures that can  be decomposed and broken down by subgroups, dimensions, gender, area, religion, ethnic group and other sub-subsections to efficaciously establish amenable national and global measures that enable us to comprehend how the various parts of the whole intertwine.\\
172
173Alternatively, MPI can be comprehended as a higher resolution and evidence based measurement of  multidimensional overlapping deprivations and poverty that goes beyond the conventional uni-dimensional measurements  to encompass horizons of multifaceted  and overlapping deprivation domains such as lack of access to basic human necessities: lack of economic measures (such as food, clothing, shelter/housing, safe drinking water, improved access to electricity and sanitation, measures of income, or measure of wealth, entitlements,tangible and intangible assets,\dots); lack of social measures and civil rights (health services, education,   access to information, education, health care, or political power, transport, cultural goods, clean environment\dots); lack of social capital(networks, bonding,credit and saving services, feeling of trust and safety, participation, self-esteem,sense of belonging,of being heard, symbolism, peace, decent/quality of life, \dots); lack of psychological dimensions (powerlessness, voicelessness, dependency, shame, humiliation, \dots) at global, regional and national level that can be further broken down by dimensions;  decomposed by social subgroups  upto a household and an individual. This multifaceted well-being measure enables us to track poverty dynamics and accurately target the poor while framing antipoverty intervention programs. Besides, it enables us to identify and aggregate levels of multiple deprivations at the individual and  household level.\\
174
175Therefore, MPI is a blend of  two key pieces of vital data to measure acute poverty: the incidence of poverty, or the proportion of people (within a given population) who experience multiple deprivations, and the intensity of their deprivation - the average proportion of (weighted) deprivations they experience.
176Both the incidence and the intensity of these deprivations are highly relevant pieces of information for poverty measurement.\\
177 
178The MPI also has further benefits. Because of its parsimonious functional form and direct measures of acute deprivation, it enables us to make comparisons across countries or regions of the world, as well as within-country comparisons between regions, ethnic groups, rural and urban areas, and other key household and community characteristics, categories and subgroups. Additionally, it enables to make an analysis of patterns of poverty: how much each indicator and each dimension contributes to overall poverty.\\
179
180In this study, MPI consists of  three dimensions made up of sixteen indicators. Associated with each indicator is a minimum level of satisfaction, which is based on literature, normative decisions, contextualized to Ethiopian practical facts on the ground, leading plans developing on consensus like the Sustainable Development Goals (SDGs 2016-2030). Based on the Santos E.and Alkire1, (2011)the minimum level of satisfaction/achievement is designated as a deprivation cut-off. Two steps are then followed to calculate the national MPI:
181\begin{enumerate}
182 \item[Step 1]: Each person is assessed based on household achievements to determine if he/she is below the deprivation cut-off in each indicator. People below the cut-off are considered deprived in that indicator.
183
184 \item[Step 2]: The deprivation of each person is weighted by the indicator's weight.
185\end{enumerate}
186If the sum of the weighted deprivations is one-third or more of possible deprivations, the person is considered to be multidimensionally poor.\\
187
188\subsection{Justifications for Using the Multidimensional Alkire-Foster Approach}
189This section highlights  the importance of the topic and the approached employed, presents brief weakness of the main stream method solving poverty related problems using monetary approach, reveals the gap of the one dimensional measurement of poverty and formulate the key questions we sought to answer. \\
190The various types and dimensions of poverty as agreed up on by many actors in the Sustainable Development Goals (SDGs) including poor sanitation, malnutrition, gender discrimination, quality of work, exposure to violence and so on; it is vital to study the multiple facets of overlapping deprivations and poverty since it matters when  poor people describe their experience, they refer to those dimensions that are components of the dimensions and indicators of multidimensional poverty. \\
191
192Format years, any destitution and poverty measurement and analysis practice has been using the dominant approach of uni-dimensional measurements of proxy welfare indicators of monetary poverty measures as the forerunners  argue that there exist  strong correlation between monetary measure and others direct measure of poverty. 
193It is thus necessary to better understand the connections and links between different poverty dimensions (Razafindrakoto and Roubaud, 2005; Ferreira 2011). Obliquely, this method was already defended by proponents when they argued that one of the major revolutions in policy-making was the recognition that policies that combine different objectives were more effective in reducing poverty (Kanbur and Squire 2001). But for them, this implicit hypothesis was that monetary poverty is strongly correlated with other dimensions of poverty. Thus, policies which fight monetary poverty ought to have a positive impact on the different dimensions of non-monetary poverty measures. Nevertheless, their proposition was criticized by other research findings (Razafindrakoto and Roubaud, 2005). The empirical evidence they gathered proposes that the correlation between the different dimensions of poverty is limited.\\
194
195Comprehending the existing connections and relations between poverty dimensions requires reconsidering poverty within a broader outlook in which diverse forms of poverty related to different poverty dimensions are recognized. But poverty should be distinguished not only according scopes but also according to the chronic/transient condition, the compounded effects of multiple deprivation and its relations to social exclusion, the factors allied to inter-generational poverty transmission not essentially allied to deprivations but also to given initial unfavorable settings. Social connections and transformation processes transiting through markets, public decisions and so on may also play a key role behind insufficient outcomes. This may be particularly the case in the presence of discrimination (in its many dimensions) and negative neighborhood effects.\\
196
197The aggregate MPI is the most useful for providing a headline to global , regional, and national multidimensional poverty within any population. But also is very essential in showing in which sense people are poor, where the poorest people live (by region and social group), as well as the intensity of the deprivations. Another advantage to study MPI, in lieu of applying the money metric approach, is that the more policy pertinent information on deprivations and poverty is made available on the public domain, the better the policy makers are empowered and equipped   to tackle it. Where most people lack access to improved sanitation, for example, a antipoverty intervention strategies, policies and programs  must obviously be different from where people are deprived of access to education.\\
198
199This research is also encouraged by the fact that, in sustainable development goals (SDGs), it is recognized that poverty has multifaceted dimensions, settings and circumstances and needs to be measured and assessed in multiple approaches since the much needed progress is being advanced in defining and measuring poverty in its spectrum of multiple. To shed more light on this, poor people encounter multiple facets of intertwined deprivations and poverty simultaneously such as poor health services, lack of education, insecurity, powerlessness, voicelessness, lack of decent living and so forth which are contemporaneous with lack of money. For this reason, the Multidimensional Poverty Peer Network (MPPN) have been calling for meshed measures of the AF approach to be considered in the sustainable development goals (SDGs) in measuring their achievements on poverty reduction to complement monetary measures of poverty. Besides, the Global Multidimensional Poverty Index woven with SDGs pinches various types of disadvantages that each deprived individual or households  encounter simultaneously. This includes poor sanitation, malnutrition, gender discrimination, poo quality of work, or violence coupled with providing headline measure of multidimensional poverty within a population.\\ 
200
201Indexes like MPI, broken down to district level, age group, by faiths etc help to reveal those area that we are facing them and that were hidden under the average buses. This will help us to target and bring on to the surface those area that are  today facing extreme poverty, however, have been  hidden from the lens of the policy/decision makers and were not  stated and surgical/specific antipoverty intervention programs were not developed in the policy design landscape in the traditional poverty measurement and analysis using the overall measures. It further highlights the gaps in the development process, the area/sectors that are doing well and it draws the attention to the area regions and communities that are left behind and craft policies to address their problems.\\
202
203More importantly, The monetary poverty measures of includes very limited indicators of human and natural capital (stocks) which is the major components of wealth of nations and gives much emphasis on the income and expenditure (flows). Contrary, the Alkire-Foster Approach of Multidimensional Poverty Measurement gives more focus to human capital by considering the education and health dimensions and the indicators thereof and gives doe attention planetary economic factors and the deprivations of these factors directly or indirectly. This again is another advantage of Multidimensional Poverty measurement over uni-dimensional poverty measurement. 
204 \subsubsection{Why Multidimensional Dimensional (MD) Poverty Analysis?} 
205 
206This subsection shall shed light on the key reasons why multidimensional measures of poverty (and well-being) are on the upsurge and why we employed these method in lieu of the conventional one-dimensional income or consumption approach.\\
207
208Improving the well-being of households and poverty reduction has long been a focal point of policymakers, development experts, and socioeconomic researchers in developing countries general and in Ethiopia in particular.
209For the last two decades, Ethiopian households have experienced progress in well-being mainly owing to structural Economic Reforms,  investments in infrastructure and in socioeconomic development programs, fast economic growth, and relatively political stability and security (World Bank, 2015).\\
210
211Although carefully constructed, this kind of hybrid index shares a common limitation with separate analysis of individual attributes: it does not consider the dependence among various attributes.\\
212In the MPI framework, one needs to account for the dependence among various attributes to arrive at reliable welfare assessments. Regardless of the choice of welfare index,general inferences should be based on the joint distribution of the multiple attributes in
213question. 
214
215According Alkire (2016), in addition to moral or ethical motivations, the following three vital requirements should be considered in justifying why we should use MPI in lieu of multidimensional money metric poverty assessment and profiling:
216\begin{enumerate}
217\item Technical – they can be constructed
218\item Empirical – they add information and value
219\item Policy – they meet policy demands\\
220Furthermore, MPIs are forward looking, and they will be the main tools for assessing and  measuring the achievements  in the Sustainable Development Goals(SDGs) by breaking down national and global silos to strengthen antipoverty intervention policies. Furthermore, the MPI supports these SDGs priorities:
221 \begin{itemize}
222 \item Integrated, coordinated policy (break Silos)
223 \item Inclusiveness (dis-aggregation by groups, dimensions, and other indicators)
224 \item Universality (acute and moderate poverty)
225 \item Data Revolution (do-able, adds value)
226 \item Global Monitoring in order to complement $\$1.90$
227 \end{itemize}
228To put it differently, by breaking down the existing silos and facilitating work flow, MPI enables broad policy design, coordination, implementation, communication, governance and resources integration greed towards fostering inclusive development and ameliorating fundamentalist components of human dignity such as cultural values, health services, social support developmental programs, freedom to life choice, security, environmental safety, empowerment of women, equality of autonomy, opportunity and outcomes. 
229\end{enumerate}
230Moreover, while coordinated and strenuous endeavors are exerted, development agents, leaders, academia, philanthropists  are  working to see the world free of poverty, multidimensional poverty measures are upsurging for their plus prose, attractive features, gaining wider acceptance by those in charge of developing public policies since various dimensions and indicators can be selected to generate specific measures for particular contexts.
231To substantiate this claim, a recent broad based survey study that cover:
232$\bullet$ 88 national consultations\\
233$\bullet$ 11 thematic consultations
234My World survey\\
235$\bullet$ 9.7 Million Voices (today)\\
236$\bullet$ survey asking you to vote for 6 out of 16 topics.\\
237$\bullet$ People clearly said that the fundamental areas covered by the MDGs education, health, water and sanitation, and gender equality – remain critically important, and not only for people living in poorer countries but also for people in the developed world.\\
238$\bullet$ At the same time, there is a call to strengthen ambition and urgency so as to reach the remainder of the world's people who are still living with many unacceptable expressions of poverty (Alkire, 2016).
239From this survey explained above, top priorities were identified and ranked as follows: 
240\begin{enumerate}
241\item A Good Education
242\item Better Health care
243\item Better Job Opportunities
244\item An honest and responsive government
245\item Affordable and nutritious food.
246\end{enumerate}
247  \begin{enumerate}
248 \item although poverty is rarely about the lack of one thing, the bottom line is that it is fundamentally lack of food. 
249\item poverty has important psychological dimensions such as powerlessness, voicelessness, dependency, shame, and humiliation
250 \item poor people lack access to basic infrastructure road mobility, and clean water. 
251 \item poor people realize that education offers an escape from poverty.
252 \item poor health and illness are dreaded almost everywhere as a source of destitution.  Finally, 
253\item the poor people rarely speak of income, but focus instead on managing assets, physical, human, social, and environmental as a way to cope with their vulnerability.
254 \end{enumerate}
255%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
256%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
257\subsubsection{Research Rational and Motivation for Using MPI}
258Article 1:\\
259The traditional and conventional-the cost of basic needs, consumption expenditure or income and the associated direct and indirect indicators thereof have been allied to measures well-being. In so doing, a basket of goods and services regarded as the minimum necessity to live a non-impoverished life is valued at the current prices. Households who have insufficient income to procure that basket are believed to be poor. Even if this methodology used to furnish vital information up till now; yet, has sever draw backs.\\ Nevertheless, poor people classify themselves and their deprivation much more widely to embrace deprivations in education, health care services, housing, improved electricity and running water, improved sanitation, empowerment, employment, personal security etc.\\
260\begin{quotation}
261Advocates for these new indices correctly point out that
262command over market goods is not all that matters to peoples' well-being, and that other factors need to be considered when quantifying the extent of poverty and informing policy making for fighting poverty (Ravallion, 2011). 
263\end{quotation}
264Therefore, a single average or composition of indicators, such as consumption/expenditure, income, are uniquely unable to capture the multiple aspects, setting,and matters that contribute to poverty.\\
265However, academic and research works that estimates and analyses poverty from a multidimensional perspective is severely missing in Ethiopia and to our knowledge no  studies have documented/profiled the multifaceted overlapping deprivations broken down by dimensions, age categories, administrative regions, gender, areas, trends,  dynamics and determinants of multidimensional poverty longitudinal time span using panel data sets which are the pillar focuses of this study. The contribution of this study therefore will be calculating, mapping individual, household, and district deprivations on the basis of multidimensional poverty measures, conducting robustness analysis and testing and developing a hypothesis testing framework  for poverty comparison among among regions, areas, households and so forth. In so doing, this study  will thus apply the class of multidimensional poverty measures developed by Alkireet al (2014), Santos (2014) and Lavine et al. (2011), as the dynamic MPI has been insufficiently documented in Ethiopia. \\
266
267Furthermore, the one dimensional main stream poverty measurements and analysis approach, which deals with only the monetary aspect, does not reveal the whole picture of households' deprivations for the following reasons: First, the pattern of consumption behavior may not be uniform, so that attaining the poverty line level of income, consumption or the subjective measure do not guarantee that a person will meet his or her minimum needs. Second, different people may face different prices, reducing the accuracy of the poverty line. Third, the ability to convert a given amount of income into certain functioning varies across age, gender, health, location, climate and conditions such as disability i.e. peoples' conversion factors differ. Fourth, affordable quality services, such as water, health and education, and other non-traded good and services are frequently not provided through the market mechanism, and failing to take into account government provision of such services may overstate or under state poverty. Fifth, employing the indirect measure of poverty gives no way to verify the intra-household distribution of income of the households. Sixth, participatory studies indicate that people who experience poverty describe their state as comprising deprivations in addition to low income. Finally, from a conceptual point of view, income is a general purpose means to valuable ends.\\
268
269...xxx(More example from Butan, India and Europe, for deatil see Alkire 2016)\\
270
271Measurement exercises should not ignore the space of valuable ends (Sen, 1979 cited in Alkire and Santos, 2013).
272In order to mitigate the disadvantages of the one-dimensional poverty measure, a number of approaches have been proposed to measure or analyze deprivations in more than one dimension. The assessment of multidimensional occurrence poverty like development goes back to the influential works of Amartya Sen (1979, 1985 and 1987). More recently, the multidimensional poverty index (MPI) measures are gaining ground as the canonical measures of poverty, as absolute and relative monetary indicators, such as the income, expenditure or consumption approaches, may not only give a poor measure of the actual experience of poverty but may also lead to wrong policy implications (Alkire and Foster, 2007 and 2011a and Alkire and Santos, 2010).\\
273
274The Oxford Poverty and Human Development Initiative (OPHI) in collaboration with the United Nations Development Program's Human Development developed the conceptual genesis of the newly emerging MPI as a measure of acute global poverty (detailed information can be obtained on, Alkire and Santos 2010, 2014; Alkire et al. 2011, 2013; UNDP 2010). Moreover, Alkire and Foster (2007, 2011) developed the class of measures that belong to the multidimensional measure of destitution. While calculating the multidimensional deprivations, households are identified as multidimensional destitute if their deprivation score exceeds a cross-dimensional poverty cutoff. The number of deprived people and their deficiency score (i.e. the intensity of poverty or percentage of simultaneous deprivations they experience) become part of the final poverty measure. A more formal explanation of the methodology is presented in Alkire and Santos (2014), Alkire and Foster (2011) and Alkireet al (2014).\\
275
276On one hand, the novelty and strength of the MPI with respect to the SDGs is that it offers a mechanism for identifying people with joint disadvantages-what has been called by Ravallion (2011) a dashboard approach. The MPI employs a class of a family of deprivation measures developed by Alkire and Foster (2007, 2011a; AF henceforth), the Adjusted Headcount Ratio or  measure. The AF measures belong to a new generation of poverty measures that have renewed interest in the direct measure of deprivations by using solid aggregation methodologies based on axiomatic frameworks analogous to those which enabled the advancements in income poverty measurement in the ‘70s and ‘80s. \\
277
278The UNDP-MPI evaluates poverty based on a household’s deprivation in three basic dimensions: education, health, and living standards. This MPI incorporates the Human Development Index (HDI) measures combined with deeper and broader indicators the quality and levels of household living standards. They have been chosen as there is broad agreement that any multidimensional poverty measure should at least include these three dimensions; for the simplicity of interpret-ability; and finally for reasons of data availability. Notwithstanding the arguments to include additional dimensions such as powerlessness, deprivations of rights, violence, shame, and time use, among others, there is often no data available and there is divergence about which dimensions are apposite (Alkire and Santos, 2013).\\
279
280Here we have the view that the conventional-one-dimensional Ethiopian monetary measurement of poverty progress report remains incomplete since it has traditionally been measured in one dimension, usually expenditure/consumption or income and there is an urgent need to introduce a holistic multidimensional measurements of poverty that brings together the economic, social and environmental priorities and helps to accurately target, be able to better, track and  predict poverty conditions and design policies that enable the poor escape the multidimensional poverty trap. 
281
282%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
283\subsection{Concise Review of Global Poverty Perspectives}
284Although it may be true that employing the conventional monetary measurement of  the incidence of poverty, commonly the less than $\$$1.25 a day in the developing world has significantly brought down poverty, over the last 15 years, from being 42 percent in the 1990 to around 12  percent currently (World Development Report, 2017). The challenge of global poverty especially in developing countries is more urgent than ever as the absolute number of extreme poor still stands at an alarm rate although it is down to below one billion from 1.8 billion in the 1990. In the contrary, in Sub-Saharan Africa, the number of extremely poor have increased by 100 million(FAO, 2015). Using the conventional  monetary poverty measurement, it was indicated that close to half the world lives count on less than $\$$2.50 a day.\\
285
286Roughly 2.6 billion people in the developing world make their living on less than $\$$2 a day; of these people, close to 1.4 billion people are chronically poor  each barely surviving on less than $\$$ 1.25 per day. According to the United Nations Food and Agricultural Organization recent reports, it is estimated that close to 800 million people of the 7.3 billion people in the globe, in other words, one in nine, were found to be suffering from chronic undernourishment in 2014-2016. Over 780 million of the hungry people live in developing countries. This represents 13$\%$, or one in eight, of the population of developing countries.  \\
287
288Furthermore, the World Bank Global Monitoring Report(GMR, 2015/16) shows that since 1990, the share of the world population living below the dollar-per-day poverty line has come down from nealrly 40 percent to nearly 12 percent today. Nonetheless, more than 700 million poor people are still estimated to live below that line. This number is expected to drop to between 340 and 480 million by 2030 if the SDGs, the targets thereof and the initiatives are successfully accomplished. The number of the poorest of the poor  (chronic/extremely) people would then go down to some where between 4.2 and 5.7 percent of the global population. This is close to the     target by the World Bank which is below 3 percent. Also, employing the AF global Multidimensional  Poverty Index (MPI), recent literature indicate that  more than 1.5  billion people are multidimensional poor (Alkire, 2015).\\
289xxx\\
290
291Like wise, according to UNICEF report (2015), 22,000 children die each day due to poverty. Equally important, it is believed that more than 1 billion children worldwide are living in poverty. Moreover, more than 805 million people worldwide do not have enough food to eat. As well as this report asserts that more than 2.4 billion people do not have access to clean water and/or to basic  improved sanitation facilities and more than 750 million people lack adequate access to clean drinking water. Besides, one in four humanity lives don't have access to improved electricity. Over a billion people are unable to read a book or sign their names. Being that every 3.5 seconds, a child dies from poverty.\\
292
293To say nothing of, diarrhea caused by inadequate drinking water, sanitation, and hand hygiene kills an estimated 842,000 people every year globally, or approximately 2,300 people per day (UNDP, 2015).\\
294
295In the like manner, wealth of literature and reports such as World Health Organization (WHO) and UNICEF Joint Monitoring Program (JMP) (2015), contended that avertible diseases like diarrhea and pneumonia take the lives of over 2 million children a year that are destitute to afford proper treatment. Again, the same report states that more than 20 million children under 1 year of age worldwide had not received any of the three recommended doses of vaccine against diphtheria, tetanus and whooping cough. As well as, close to 1.5 billion people, in other words, 25 percent of all human lives are without access to electricity and roughly  more than 160 million children under the age 5 were stunted (reduced rate of growth and development) due to chronic malnutrition.  In effect, nearly 30,000 children die each day-about 11 million per year because they're too necessitous to survive.\\
296
297With such a toll, addressing poverty in new and more effective ways must be a priority for the global policy agenda. Fortunately, a variety of new actors are bringing new perspectives, holistic and all encompassing new approaches/methods, new energy to the new and dynamic challenges of human kind with particular emphasis to welfare, multidimensional poverty and inequality of opportunities, outcomes and autonomy.
298
299\subsubsection{Ethiopia's Global Competitiveness Rank}
300It is argued that Economic activity requires a streamlined regulatory environment and effectual policies that are transparent and accessible to all. Under the theme of 'Equal Opportunity for All,' a World Bank Group flagship publication has produced the 14th in a series of annual reports measuring the regulations that enhance business activity and those that constrain it. This report summarizes "Doing Business Index", presents quantitative indicators on business regulations and the protection of property rights that can be compared across 190 economizer and its impact to economic growth (World Bank, 2017; website accessed on June 20, 2017).\\
301
302Wealth of literature argue that "Doing Business" measures regulations affecting 11 areas of the life of a business. Ten of these areas are included in this year's ranking on the ease of doing business: starting a business, dealing with construction permits, getting electricity, registering property, getting credit, protecting minority investors, paying taxes, trading across borders, enforcing contracts and resolving insolvency. Doing Business also measures labor market regulation, which is not included in this year's ranking.\\
303
304Generally, countries with easy of Doing Business attract investors, eliminated inefficiency and has positive impact on economic growth and development-hence by implication it has a high impact on poverty reduction, creating employments and reducing inequality among the citizens.\\
305
306Compared to other economies, Ethiopia has the lowest/unattractive regulatory performance. Scoring 3.8 points out of 7, Ethiopia is the 109 most competitive nation in the world (over all index) out of 138 countries ranked  in the 2016-2017 edition of the Global Competitiveness Report published by the (World Economic Forum, 2016).\\
307Like wise, this is very low as compared to some selective Sub-Saharan Africa countries as shown below. \\
308
309\begin{landscape}
310\centering\sisetup{table-alignment=center, table-column-width =2.4cm}
311\setlength\doublerulesep{4pt}
312\begin{table}
313\caption{Comparing Ethiopian Regulatory Performance with selected SSA countries}
314\label{table1}
315\begin{tabularx}{\linewidth}{ |X||*{3}{S[table-format=3.0]| >{}S[table-format=1.2]||}S[table-format=3.0]| >{}S[table-format=1.2]|}
316\multicolumn{1}{c}{} & \multicolumn{8}{c}{Economy}\\
317\hhline{~||--||--||--||--|}
318\multicolumn{1}{c!{\phantom{\vrule}\vrule}}{} &
319 \multicolumn{2}{c||}{Over all Ranking} &
320\multicolumn{2}{c||}{Basic Requirement} &
321 \multicolumn{2}{c||}{Efficiency Enablers} &
322 \multicolumn{2}{c|}{Innov. and Sophis. Factors} \\
323\hhline{|~||--||--||--||--|}
324\addlinespace[4pt]
325\hhline{|-||--||--||--||--|}
326Country& {Rank} & {Score} & {Rank} & {Score} & {Rank }& {Score} & {Rank} & {Score}
327 \\
328\hhline{|-||--||--||--||--|}
329 Tunisia & 95 & 3.92 & 79 & 4.41 & 103 & 3.65 & 104 & 3.32\\
330\hhline{|-||--||--||--||--|}
331 Kenya & 96 & 3.9 & 115 & 3.81 & 75 & 4.03 & 40 & 4.03\\
332\hhline{|-||--||--||--||--|}
333 Ethiopia & 109 & 3.77 & 106 & 3.96 & 117 & 3.47 & 74 & 3.53 \\
334\hhline{|-||--||--||--||--|}
335 Cape Verde & 110 & 3.76 & 89 & 4.32 & 121 & 3.40 & 105 & 3.32 \\
336\hhline{|-||--||--||--||--|}
337\end{tabularx}
338\end{table}
339\end{landscape}
340
341\subsection{Review of Ethiopia's Well-being Indicators}
342
343Ethiopia, officially known as the Federal Democratic Republic of Ethiopia(FDRE); located in the Horn of Africa, is a rugged, landlocked country split by the Great Rift Valley. With archaeological findings dating back more than 3 million years.  Ethiopia is a place of ancient culture, civilization, multiplicity, multilingual, multicultural and multi-religious nation that constitutes nine member states and two city administrations. Ethiopia's total land area is 1.13 million square km (436, 295 sq. miles).\\
344
345According to Worldometers(March, 2017) and Global Finance on Ethiopia GDP and Economic Data (Country Report, 2017), having more than 103, 694,504 (as of March 31, 2017, based on the latest United Nations estimate) million inhabitants, Ethiopian population triples in every 50 years (for example it was 18,128,034 in 1950, 66,443,603 in 2000, and expected to be over 190,000,000 in 2050)
346Source: Worldometers (www.Worldometers.info) accessed March 31, 2017\\
347  
348According to Global Finance Country Report (2017), Ethiopian is the second most populous country in the African continent and 11th of the top 15 poorest countries in the world. Ethiopia's population is equivalent to 1.40 percent of the total world population ranked number 12 in the list of countries and dependencies by population. Out of the total population, 20.5(21, 174, 205, 2017 est.) percent of the population is urban while the rest is rural, the median age in Ethiopia is 18.9 years, yearly population growth rate is 2.45 per annum, the population density is 104 per $KM^2$ (270 people per $mi^2$). \\
349
350Besides, this report indicates that Ethiopian Gross Domestic Product (GPD) at market value by 2017, (based on the IMF World Economics Outlook 2016)is estimated to be USD 76.9 billion (7.5 percent real GDP growth per year), GDP per capita at current prices (nominal) is extrapolated to be roughly USD 830, GDP per Capita-Purchasing Power parity(PPP) International is $\$$ 2,071, GDP(PPP), Ethiopia Economy share of the world total is estimated to be approximately 0.2 percent, average inflation rate is 8.2 percent while public debt (General government gross debt as a \% of GDP is 60.3 percent. \\
351
352Furthermore, GDP(Purchasing Power Party) international is approximately 191.9 billion $\$$ public deficit (General government lending/borrowing as percentage of GDP) is -3.2 percent. Correspondingly, the International Monetary Fund(IMF) World Economic Outlook Report (IMF-WEO Report, 2017)shows that Ethiopia's Real GDP Growth Rate (annual percent change) was 10.14 on average in the study period (2011-2015), whereas Consumers' Price (annual percent change) for the same period was on average 16.58 (for details see IMF-WEO Report 2017: pages 266 and 271). \\
353
354The recent reports regarding Ethiopia's Economy demonstrate its remarkable economic and public service growth as compared to the 2015 Report across similar indicators as presented below. Not to mention, the Gross Domestic Product (GDP) in Ethiopia was worth 62 billion U.S. $\$$ in 2015 only representing 0.10 percent of the world economy. According to the World Bank National Account Data set, and the Organization for Economic Cooperation and Development (OECD)National Accounts Data files (2012), Ethiopians GDP per Capita\footnote{GPD per capita is gross domestic product divided by midyear population. GDP is the sum of gross value added by all resident produced in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. Data are in constant 2010 U.S. $\$$.} (constant 2010 U.S. $\$$)\footnote{GDP in Ethiopia averaged 17.01 U.S. $\$$ billion from 1981 until 2015, reaching an all-time high of 62 US $\$$ Billion in 2015 from a record low of 6.93 U.S. $\$$ Billion in 1994. The lowest was record 163.7 U.S. $\$$ in 1992)}.\\
355
356Ethiopia GDP Annual Growth Rate: Gross Domestic Product (GDP) in Ethiopia expanded 9.60 percent in 2015 from the previous year. In a related issues, Ethiopia recorded a trade deficit of 3254.50 U.S. $\$$ Million in the second quarter of 2016 was 392.2 in 2012.\\
357
358Moreover, there is a noticeable progress and leapfrog in improving life expectancy in Ethiopia. According to the latest WHO data published in 2015, life expectancy in Ethiopia is presented as follows: Male 62.8, female 66.8 and total life expectancy is 64.8 which gives Ethiopia a World Life Expectancy ranking of 138.\\
359As it can be seen from figure 1 below, among several Sub-Saharan African countries under consideration, Ethiopia's GDP per Capita (constant 2010 US $\$$ in 2012 was one of the lowest only above Burundi of the sample set in the domain. Looking at figure 1 below, we can argue that Ethiopia's GDP per capita (constant 2010 US Dollar) in the year 2012 was one of the lowest even when it is compared among the low in come Sub-Saharan African countries. In fact, of these countries considered for comparison, Ethiopia's per capita is only better than Burundi's GDP  per capita.  \\ 
360
361What is more, according to the World Bank, International Comparison Program database, Ethiopia's GNI per capita (constant 2011 international U.S.$\$$), 1,231.1 was  the lowest even among the Eastern African Economic groups.\\
362
363As can be observed from figure 2 above, Ethiopia's GNI per capita, PPP (constant 2011 international $\$$) by the year 2012 was very low. It is produced using StatPlanet based on the World Bank International Comparison Program data base. \\
364
365Employing the Stat Plane Trend analysis we analyzed the trend of Ethiopia's PPP GNI per capita\footnote{{Note: GNI per capita based on purchasing power parity (PPP). PPP GNI is gross national income (GNI) converted to international dollars using purchasing power parity rates. An international dollar has the same purchasing power GNI as a U.S. $\$$ has in the United States. GNI is the sum of value added by all resident produces plus any product taxes (less subsidies) not included in the valuation of output plus net receipts of primary income (compensation of employees and property income) from abroad. Data are in constant 2011 international dollars}}, though on the ascending trend, remains the lowest and only above Burundi in the year 2011-2015.  \\
366\begin{landscape}
367\begin{figure}
368\caption{ Comparing GDP per capita Among Selected SSA Countries}
369\;\includegraphics [scale=0.50]{EthGDP1perCapitagraph.png}
370\label{figure1: Bar graph}
371\end{figure}
372\end{landscape}
373
374\begin{landscape}
375\begin{figure}
376\caption{GNI per capita, PPP (constant 2011 international $\$$) Comparison among selected SSA Countries}
377\label{Figure2}
378\includegraphics [scale=0.45]{ETHGNIPPP1graph.png}
379\end{figure}
380\end{landscape}
381
382As it can be seen from the time series trend analysis scatter diagram figure 3 below, Ethiopia's GNI per capita, PPP (constant 2011 international US $\$$ represented in green is again the lowest as compared with selected Eastern Africa Economic block save for Malawi, Mozambique and Burundi that are the bottom low income countries on earth. Gross national Income (GNI) per capita (constant 2011 international U.S. $\$$) (2012).\\
383 
384Uniquely, Ethiopians GNI per capita \footnote{"PPP GNI is gross national income (GNI) converted to international dollars using the purchasing power party rates. An international dollar has the same purchasing power over GNI as a U.S. $\$$ has in the United States.
385GNI is the sum of values added by all residents plus any product taxes (less subsidies) not included in the valuation of output plus net receipts of primary income (compensation of employee and property income) from abroad.  Data are constant 2011 international dollars."
386Sources: World Bank (2012), International Comparison Program database.} based  on purchasing power parity (PPP) remains very low (1,231.1) even by the Sub-Saharan Arfican (SSA) Countries standard and is the 9th poorest countries of the 33 SSA countries with the required database.
387
388
389\newpage
390\begin{figure}
391\caption{GNI per capita, PPP (constant 2011 international $\$$)  Time Series graph Analysis among selected Eastern African Countries Economic block}
392\label{figure3}
393\includegraphics [scale=0.45]{EthGNI2perCapitagraph.png}
394\end{figure}
395\subsubsection{Lack of  Houses $\&$ Unsafe Homes } 
396The vast majority of Ethiopians live in poorly built, shabby and ramshackle houses which lack even the basic facilities, such as toilets and sanitation materials. Even in the seat of the AU, UN Organizations, Diplomatic Community, and the capital city Addis Ababa, reports indicate that 80$\%$ of the houses are in poor condition and below standard. Houses in slum areas are old and dilapidated and too narrow to accommodate families, where the health and dignity of families is compromised.\\
397According to Habitat for Humanity website report (2017, accessed on April 21, 2017); most families who live in dilapidated homes in slum areas share toilets that are also in very poor condition. Over 24$\%$ of the households do not have any form of toilet facility at all and 63$\%$ use shared pit latrines. 25$\%$ of the solid waste generated from the city is left unattended. Poor families do not have toilets at all or use bad toilets that are nearly abandoned.
398
399\subsection{Review of Ethiopia's Poverty Status}
400
401During the inception of the Millennium Development Goals (MDGs, 2000); Ethiopia had one of the highest poverty rates in  the world, with more than 56 percent of the population living below the international poverty line (or US $\$$ 1.25 PPP a day)and  44 percent of the population below the national poverty line (World Bank, 2015). It is further revealed that in 2011, less than 30 percent of the population live below the national poverty line and 31 percent on less than U.S. $\$$ 1.25 PPP per day (World Bank, 2015).\\
402
403In addition, according to the World Bank Report (2015), commendable progress has been registered in education, health and socioeconomic programs by promoting pro-poor spending on basic social services and social protection. Moreover, recent reports from the World Bank (World Bank Report, 2015) pointed out that by 2012, women antenatal check-up has enhanced and is more than one in three now while it was one in five some 10 years ago.  Alike, the incidence of stunting was reduced from 58 percent in 2000 to 44 percent in 2012.\\
404
405Illiteracy rate also declined considerably from 70 percent to less than 50 percent in the same year. At last, the figure of households with better living standards measured by electricity, piped water, and water in residence doubled (from 12 to 23' from 17 to 34 respectively)percent from 2000 to 2012. In the same year, less than 30 percent of the population lived below the national poverty line and 35 percent lived on less than US$\$$1.25 PPP a day. 
406\begin{quotation}
407%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
408On average, household in Ethiopia also has better health, education and living standards today as compared with the situation a decade ago in which percentage of population (people without education declined 70 to 50, with electricity increased from 12 to 23, access to piped water augmented from 17 to 34). There are major steps forward in life expectancy and advancement towards the realization of the Sustainable Development Goals (SDGs), mostly in gender equality in primary education, child mortality, HIV/ AIDS, and malaria and so on (World Bank Report, 2015).
409%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
410
411%\addcontents{toc}{table5}
412\end{quotation}
413% \begin{landscape}
414 \begin{center}
415 \begin{table}
416 \caption{Ethiopia's Progress in Well-being \\Indicators from the Year 2000 to 2011}
417 \label{table2}
418  \begin{tabular}{ |p{8cm}||p{3cm}|p{3cm}|}
419\hline
420\multicolumn{3}{|c|}{Ethiopia then and now: a decade of progress from 200 to 2011 }\\
421\hline
422 \hfill                                 & 2000  &   2011\\
423\hline
424Percentage of population   \hfill       &       &       \\
425\hline
426Living below the national poverty line \hfill  & 44   &      30 \\
427\hline
428Living on Less than  US$\$$1.25 PPP a day \hfill & 56 &      31\\
429\hline
430Without education       \hfill                & 70 &     50 \\
431\hline
432With electricity      \hfill                    & 12 &      23 \\
433\hline
434Piped water          \hfill               & 17  &      34\\
435\hline
436Percentage of children under 5 years that are stunted \hfill & 58 & 44\\
437\hline
438
439Percentage of rural women receiving antenatal checkup \hfill & 22 & 37\\
440\hline
441Life expectancy   (years)     \hfill                & 52  & 63\\
442\hline
443\end{tabular}\\
444
445Source: Ethiopia Demographic and Health Survey, Household Income and Consumption\\ Expenditure Survey, World Development Indicators, Carranza and Gallegos (2011),\\ Cannin et al. 2014.\\
446\end{table}
447\end{center}
448%\end{landscape}
449
450Nonetheless, progress has not been consistent across the country, since households that are dependent on agriculture remain vulnerable. \\
451Food and nutrition insecurity have long been one of the most serious issues in Ethiopia. What is more, Ethiopia has been structurally in food shortfall since at least 1980s. The modern insight on nutrition insecurity in Ethiopia, according to researchers and other experts in the field, can be simply summarized as follows: 
452
453\begin{enumerate}
454 \item  Population ascending trim down landholdings more and made places intolerable stress on an already flimsy natural resource base;
455 \item Soil fertility is going down owing to intensive cultivation and limited application of yield-augmenting inputs;
456 \item  Land ownership are too small and uneven, although they are unusually evenly distributed, to permit the majority farming households to accomplish food production self-sufficiency;
457\item Persistent famines exacerbate food production shocks to unusually low yields;
458 \item  Inadequate off-farm employment opportunities put a ceiling on diversification and migration options, leaving people trapped in more and more nonviable agriculture. 
459 \item Despite its miscellaneous landscape and diverse climate, owing to global climatic changes, rainfall is becoming ever more unpredictable, arriving later in the season, and droughts happening in short cycles that it was.  
460 \item Natural catastrophe, financial disasters, and other economic distresses can have major negative impacts for unprotected households.
461\item Its shock on small-scale farmers who make up 80.3 percent of the Ethiopian population, are likely to put up with the burden of climate variation provoked drought in Ethiopia, resulting in water scantiness and food insecurity (Sources: Devereux, (2000); Skoufias, (2003); Block et al. (2004); FAO, (2014) and World Bank, (2015)).
462\end{enumerate} 
463%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
464Article 2: Education Deprivation In Ethiopia
465\subsubsection{Overview of Socioeconomic Domains (Health and Education) in Ethiopia} 
466\begin{quotation}
467Education is the most powerful weapon which you can use to change the world. Nelson Mandela
468\end{quotation}
469To begin with a general perspective, according to the UNESCO Institute for Statistics database (UIS, 2016), 758 million adults (15 years old and over) in the world cannot read and write. At the same time, close to 500 million (2/3) women are unable to read prescription, fill out some basic forms or send text messages. Significantly, currently there are 123 million illiterate youth (15-24 years old). Out of this, over 76 million (2/3) are female. Another key point, 57 million out-of-school children, half of them (1 in 2) live in Sub-Saharan Africa. Strikingly, half of those children who live in Sub-Saharan Africa  never enter a class room  while the other half have dropped out or entered late. 54 million of these live in only nine developing countries where Ethiopia takes the forth place. One of the low income countries Ethiopia is not an exception to these facts \\
470
471Although enormous progress have been made under the watch of the incumbent Government of  Ethiopia (GoE) in  putting students into classrooms that resulted in the  primary level enrollment surge in which general enrollment rate (GER) being over 90 percent and net enrollment rate (NER) of  85 percent (World Bank, 2015); education quality deficits, low literacy rates, low reading performance ( 1 in 3 pupil of second graders were found to be non-readers and close to 50 percent scored zero on a comprehensive test); deterioration in the reading mean score, and very low literacy and numeracy indicators (USAID Ethiopia, 2010); substantially low, as compared to the respective Low-Middle Income Countries ( LMIC) averages of 72 percent and 45 percent,  lower and upper secondary (preparatory) GERs for Ethiopia is approximately  [38 percent and 8 percent, respectively (World Bank, 2015)], massive dropout and repetitions rates and below average national matriculation and university entrance results (MoE, 2016)   have been particularly severe and critical bottlenecks in Ethiopia (Tassew et. al., 2012, UNESCO, 2016, Dibaba, 2015). \\
472
473In like manner, the UNDP Human Development Report (2016) reveals that although Ethiopian Index of Human Development (HDI) showed a positive trend (increased from 0.283 to 0.448) with an increase of 58.2 percent. Ethiopia's 2015 HDI of 0.448 is below the average of 0.497 for countries in the low income development category and below the average of 0.523 for the Sub-Saharan African countries. \\
474%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
475\begin{landscape}
476\begin{table}
477\caption{Comparing Ethiopia' HDI and Other Development Indicators with selected SSA countries }
478\label{3}
479\begin{tabular}{|p{4cm}||p{3cm}|p{3cm}|p{3cm}|p{3cm}|p{3cm}|p{3cm}|}
480\hline
481\multicolumn{7}{|c|}{HDI and component indicators for 2015 relative to selected countries and groups }\\
482\hline
483country name or list & HDI value & HDI rank & Life expectancy at birth &Mean years of schooling & Expected years of schooling & GNI per capita (PPP USD)\\
484\hline
485Ethiopia & 0.448 & 174 & 64.6 & 8.4 & 2.6 & 1,523\\
486\hline
487Rwanda & 0.498 & 159 & 64.7 & 10.8 & 3.8 & 1,617\\
488\hline
489Uganda & 0.493 & 163 & 59.2 & 10.0 & 5.7 & 1,670\\
490\hline
491Sub-Saharan Africa& 
4920.523 & - & 58.9 & 9.7 & 5.4 & 3,383\\
493\hline
494Low HDI & 0.497 & - & 59.3 & 9.3 & 4.6 & 2,649\\
495\hline
496\end{tabular}\\
497\end{table}
498Source: UNDP Human Development Report (2016)\\
499\end{landscape}
500With all its success stories, a measurement of human development, HDI has draw back as it is an average measure of basic human development achievements since average measure overlook crucial spots hidden under the average analysis particularly in the uni-dimensional measure of poverty methods. 
501Considering this shortcoming, a distribution sensitive measure, Inequality-adjusted HDI (IHDI) that combines  country's average achievements in health, education, and income and how those achievements are distributed among the population by discounting each domain mean value according to its level of inequality developed. \\
502
503In other words, the IHDI is the discounted HDI for inequality.  
504According to the IHDI measure, that is when the value of HDI which was 0.448 is discounted for inequality, it falls to 0.33, nearly 26 percent owing to the inequality in the distribution of the HDI dimension indices. Rwanda and Uganda show losses due to owing to inequality of 31.9 percent ad 30.9 percent respectively (UNDP, 2016).\\
505
506\begin{landscape}
507\begin{table}
508\caption{Comparing Ethiopia' IHDI and Other Development Indicators with selected SSA countries }
509\label{table 4}
510\begin{tabular}{ |p{2.5cm}||p{2.5cm}|p{3.5cm}|p{3.5cm}|p{3.5cm}|p{3.5cm}|p{3.5cm}| }
511\hline
512\multicolumn{7}{|c|}{Ethiopia's IHDI for 2015 relative to selected countries and groups }\\
513\hline
514country name or list & IHDI value & Overall loss($\%$) & Human inequality coefficient($\%$) & Inequality in life expectancy ($\%$) & Inequality in education ($\%$) & Inequality in income ($\%$)\\
515\hline
516Ethiopia & 0.330 & 26.3 & 25.5 & 30.3 & 36.6 & 9.5\\
517\hline
518Rwanda & 0.339 & 31.9 & 31.8 & 29.8 & 29.3 & 36.4\\
519\hline
520Uganda & .341 & 30.9 & 30.8 & 35.7 & 29.4 & 27.3\\
521\hline
522Sub-Saharan Africa& 
5230.355 & 32.2 & 32.1 & 34.9 & 34.0 & 27.4\\
524\hline
525Low HDI & 0.337 & 32.3 & 32.0 & 35.1 & 37.1 & 23.9\\
526\hline
527
528\end{tabular}\\
529 \end{table}
530Source: UNDP Human Development Report (2016)
531\end{landscape}
532
533The incumbent Government of Ethiopia (GoE) has been working with stakeholders, exerting strenuous endeavor to further increase and enhance access to education in undeserved areas and improves the overall quality of education (World Bank, 2015, 2013).  Uniquely, Ethiopia's Ministry of Education (MoE) has developed Education Sector Development Program (ESDP V) (MoE, 2016), General Education Quality Improvement Program II (World Bank 2016); a Nationwide Quality Education Frameworks aiming at improving general education quality. \\
534However, diversifying delivery strategies to shift from the conventional approach of chalk and talk, reading and exams, teacher dominated and one sided transmission of information to a wide spectrum of delivery strategies by blending the conventional practices with greater use of cooperative teaching-learning, utilization of innovative digital platforms and solutions to increase access and/or facilitates interactive learning, to improve access, quality, and relevance to deliver secure content, facilitate student exposure to alternative skills and knowledge sources, and more importantly provide access to standard central resources hubs and quality teachers were not given much attention. \\
535Most compelling evidences, youth (15-24 years old) literacy rate (in percentage) between 2008-2012 in Ethiopian was recorded 63 and 47 for male and female respectively. Various reports revealed that pre-primary school participation gross enrollment ratio (in percentage) was 5.6 and 5.3 for male anf female respectively. Besides, percentage of children of primary school age (7-14 years old) out of school was  43 and 31; whereas percentage of school secondary school age (15-18 years old) was 42 and 44 for male and female respectively (UIS, 2013). \\
536
537A noble and fairly representative survey of Ethiopian Socioeconomic Survey (ESS, 2011-2012, 2013-2014 and 2015-2016)- Living Standard Measurement Study-Integrated Survey of Agricultural  (LSMS-ISA) reveals that the national literacy rate is approximately close to 59 percent. Expressly, decomposing this by region, literacy rate ranges from 53 $\%$ in the Amhara Regional State to 67 $\%$ in the Tigrai Regional State (CSA, 2016). \\
538
539Considering Ethiopian adults (age 15 and over) those who can read and write, Ethiopia is  one of the 10 Countries with the worst literacy rates in the World.  Among the total population, the national literacy rate is only 49.1 $\%$; out of this, the proportion of male population literacy rate 57.2 $\%$ that is much higher as compared female counter parts which is  41.1$\%$ (UIS, 2016). To put it differently, Ethiopia has one of the highest literacy rate in the world. \\
540Sources: United Nations Educational, Scientific, and Cultural Organization (UNESCO) Institute for Statistics (UNESCO-UIS, 2016).\\
541
542Yet, those who need education the most, children living in poverty are the least likely to attend and complete school. Building academic and physical resources, such as schools, labs, libraries, research centers, resources centers, science  centers, sport and recreation and so forth in poorest countries like Ethiopia is building the future to break trough the poverty trap, illiteracy, backwardness and low expectations through service and education. An educated society is like an oiled machine and is the best weapon to free, emancipate, and transform the society; to empower women, reduce inequality, improve decent living, foster self esteem, improve family planning and investment on children, foster health service, better security and environment sensitivity.\\  
543"For millions of children in developing countries, another missed school year is another missed opportunity to lift themselves out of the cycle of illiteracy and poverty" BuildOn website accessed on March 25, 2017.\\
544
545\subsubsection{Life Expectancy in Ethiopia}
546According to the latest WHO data published in 2015 life expectancy in Ethiopia is presented as follows: Male 62.8, female 66.8 and total life expectancy is 64.8 which gives Ethiopia a World Life Expectancy ranking of 138 (WORLD HEALTH RANKINGS: LIVE LONGER LIVE BETTER 2017, accessed on April 21, 2017; for details refer to www.worldlifeexpectancy.com/ethiopia-life-expectancy).
547%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
548Article 3: Health Deprivation in Ethiopia
549\subsubsection{Brief Review of Situations of Drinking Water; Sanitation and Hygiene in Ethiopia}
550Water, Sanitation and Hygiene (WASH)are fundamental and integral part of decent human health and survival. It's also a vital component in freeing people from multiple deprivations and poverty. According to WASH and UN-Water Day report(2016), globally, 663 million people or 1 in 10 people do not have access to improved and safe sources of drinking and use water, more than one third of the global population (33$\%$) or 2.4 billion people  does not have access to adequate hygiene and  sanitation, and 97$\%$the world water is salty or otherwise undrinkable. 
551 
552\begin{quotation}
553Today, 1.8 billion people use a source of drinking water contaminated with faeces, putting them at risk of contracting cholera, dysentery, typhoid and polio (Charity:water, 2017; accessed April 14, 2017).
554The Sustainable Development Goals, launched in 2015, include a target to ensure everyone has access to safe water by 2030, making water a key issue in the fight to eradicate extreme poverty (WASH/OHorizons, 2016).
555\end{quotation}
556
557Accesses to improved water and sanitation have the power to transform peoples' lives; water is essential to human survival, the environment, and the global economy. Safe working conditions and a living wage can provide workers with sustainable income and pave the way for broader social and economic advancements. So is sanitation. The lack of access to both leads to miserable life, environmental deterioration that will have an adverse impact on the global economy, hampers sustainable growth, conducive  work climate, peace, security and development. \\
558
559Additionally, provisions of clean water, basic toilets and good hygiene practices are crucial for the continuity and development of children. At present, close to 2.4 billion people  are deprived of improved sanitation, and 663 million  do not have access to improved water sources and 1 billion people of the global population defecate in the open air. In other words, 7 out of 10 people are without access to improved sanitation and 9 out of 10 who have to go open live in rural area   (UNICEF, 2015).\\
560   
561Globally, it is estimated that women and children spend 125 million hours each day collecting water(WASH/OHorizons, 2016). Time is money and it adds up. this time could instead be spent generating income or doing other activities. By alleviating this back breaking and time consuming chore, women can gain more hours in their daily life and be empowered to consume it where it deems fit. Increasing access to drink clean drinking water promotes right to live a healthy, productive and dignified life. Kids miss school if they sick or spending time collecting water from far away water sources; and families cannot save additional money they spend on hospital,clinics and other health services to cure/prevent water-related illness. \\
562
563Globally, it is estimated that women and children spend 125 million hours each day collecting/fetching water (WHO, 2016). Time is money and it adds up. This time could instead be spent generating income or doing other activities. By alleviating this time consuming chore, we're giving women more hours back in their day and empowering them to use their time as they see fit.\\
564
565Improved Sanitation facilities, urban (percentage of urban population with access (2011). Access to improved sanitation facilities refers to the percentage of the population using improved sanitation facilities. Improved sanitation facilities are likely to ensure hygienic separation of human extract from human contact. These include flush/pour flush (to piped sewer system, septic tank, pit latrine), ventilated improved pit (VIP) latrine, pit latrine with slab, and composting toilet (WHO/UNICEF Joint Monitoring Program(JMP) for Water Supply and Sanitation, 2011). As it can be seen in the graph below, of the 20 urban Ethiopia, slightly higher 26 percent have access to improved sanitation and it is the lowest as compared with other Sub-Saharan countries.\\
566 
567Nation wide, over 42 million Ethiopian lack access to safe and improved running water, over 70 percent lack access to improved sanitation, 71 percent of the total Ethiopian population live on less than US$\$$3.10 per day,  women and children walk over 3 hours to collect water, often from shallow wells or unprotected ponds they share with animals and over 30 million practice open defecation (Water.Org, 2017) accessed January, 2017. \\
568
569Women are hit the hardest when water, sanitation and hygiene (WASH) infrastructure is insufficient. When the drinking water sources is not on the premises, an adult female household members bears the burden. In the ESS panel data 2011-2012, 2013-2014, and 2015-2016, it is reported that in rural areas, protected well, unprotected well and river/lake are the main sources of drinking water receptively 37 percent, 19 percent and 20 percent of the households. Close to 69 of large towns and 33 percent of small town area households have access to pipped water. \\
570 
571Community water points can require women to walk far distances and wait in long lines. In the time use data that were collected during February-April, 2014 which is the post harvest season for  in Ethiopia; collecting water and fuel wood is considered as females job. In the ESS (2014), it was reported that about 54 percent of female household members spend some time collecting fuel wood or water to the household on daily basis. Whereas, only 21 percent of male household membranes reported spending time on fuel and water collection for the family (CSA, 2015).\\
572
573Besides, Ethiopia's water supply, sanitation, and hygiene (WASH) is again one of the lowest in the world; just 49 $\%$ and 21 $\%$ of the population has access to safe water and sanitation respectively. As it can be deduced from the graph below, as compared with selected Sub-Saharan African countries whose data are available in Stata Planet, improved sanitation service in Ethiopian is very low (slightly higher than 25$\%$). Thus, considering an other important indicator of living standard of Ethiopian households, it's at its lowest level and this reinforces our finding of very high MPI level. //
574As can be seen in figure 4 below, access to improved sanitation facilities in urban area of Ethiopia was very low (Just around 25$\%$ even among Sub Saharan African Countries.
575\begin{landscape}
576\begin{figure}
577\caption{Delivery of Improved Sanitation Facilities in Urban Ethiopia \\(percentage of urban population with access to improved facilities) in 2010 }
578\label{figure4}
579\includegraphics [scale=0.62]{EthSani1Graph.png}
580\end{figure}
581\end{landscape}
582Improved water source, urban (percentage of population with access)(2010), access to improved water sources refers  to the percentage of population using an improved drinking water sources. The improved drinking water sources includes piped water on premises (piped household water connections located inside the user's dwelling, plot or yard), and other improved drinking water sources (public taps or standpipes, tube wells or borehole, protected dug walls, protected spring, and rainwater collection); (WHO/UNICEF Join Monitoring Program (JMP) for Water Supply and Sanitation, 2015).\\
583
584The graph below portraits improved water sources in urban area of some selected Sub-Saharan African countries. As can be seen from the graph below, Ethiopian access to improved water (percentage of urban population with access to improved water in the urban residents in in 2010) is 91.1
585\begin{landscape}
586\begin{figure}
587\caption{Improved water supply in Urban Ethiopia\\(percentage of population with access to improved water supply by 2010)}
588\label{Figure5}
589\;\includegraphics [scale=0.53]{EthWater2Ugraph.png}
590\end{figure}
591\end{landscape}
592As it can be deduced from the graph below, access to improved and potable water delivery in rural residents in Ethiopia is by far the lowest even as compared with some selected least developed and low income Sub-Saharan-African countries. This is consistent to our analysis and  finding in improved water supply as one key MPI indicator deprivation among Ethiopian households.  \\
593
594Despite Ethiopia bounty of nature and has immense water resources, owns some of the biggest rivers that cross boundaries in African, water deliver and access to improved water in the rural areas of the country is one of the among Sub-Saharan Low Income countries. Moreover,modern irrigation development, fair and sustainable utilization of water sources and hydro power generation and development are at the bottom despite some positive progress recently. \\
595
596Looking at figure 6 below, access to improved water delivery in the country side is very limited that need  a big-push endeavor as development in the water resources and energy can ignite other engines of development and have much multiplicative effects in other sectors. 
597
598\begin{landscape}
599\begin{figure}
600\caption{Access to Improved Water Resources in Rural Ethiopia\\ (Percentage of Rural population with access) by 2011}
601\label{figue6}
602\includegraphics [scale=0.52]{EthWater1Rgraph.png}
603\end{figure}
604\end{landscape}
605%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
606\subsection{Purpose Statement of the Ethiopian Multidimensional  Poverty Measurement and Analysis}
607
608The Ethiopian national multidimensional overlapping deprivations and poverty measurement aims at estimating the Ethiopian national population-wide progress in capability framework of poverty reduction at national level, at regional state level, in urban-rural area, in various ethnic and religious groups, different age groups, gender wise and so forth in course of actions that are believed to be the most legitimate, realistic, accurate and comprehensive in the lenses of citizens. Moreover, the core mission is  developing, assessing and measuring  acute multidimensional poverty that is the proportion of people who experience multiple deprivations and the intensity of such deprivations by  using forms of the Alkire-Foster method upon which the MPI is based to better address local realities, needs and the data available at our disposal.\\
609
610Furthermore, by expanding the base, domains and angles from which poverty is analyzed and taking the broad based and measurable responsibility, we presented the status of official measures that show the level and composition of multidimensional poverty decomposed by subgroups and broken down  by indicators and domains contribution to the overall aggregate multidimensional poverty that is amenable to updating regularly across the board throughout the population-wide progress in capability poverty reduction every two years from 2011-2016(i.e. 2011-2012, 2013-2-14, and 2015-2016)at national/Federal level, Regional States, rural-urban area,  ethnic and religious groups, age group, gender in ways that are regarded as legitimate and accurate by the citizenry.\\
611 
612As well as, to compares the situation of regional states with respect to acute poverty across various groups so as to enable policy makers precisely target the poorest more efficiently; measure and monitor national progress while enacting antipoverty intervention programs. The measure shall be disseminated across the public sector, NGOs, and academic institutions among others. Results will be communicated widely to Ethiopian government official, academics, research institutions, citizen and social groups.  \\
613
614\subsection{Sources and Governance (Data, Authority, and Procedures)}   
615For measuring multidimensional overlapping deprivations and poverty, we employed three  waves of panel data sets(secondary), commonly known as  the Ethiopia Socioeconomic Survey (ESS), 2011-2012, 2013-2014, and 2015-2016.  Ethiopian Socioeconomic Survey(ESS) is a joint project managed by the World Bank Living Standards Measurement Study-Integrated Surveys of Agriculture team (LSMS-ISA) and Central Statistics Agency of Ethiopia (CSA) that is  a newly-designed and noble survey, that was fielded every two years. The World Bank (WB) and the Ethiopian Central Statistical Agency (CSA) have the authority to implement the survey, construct the measure, and release it as an official statistic. These institutions can propose to update the methodology roughly once per decade. A cross-institutional working group can be constituted to propose changes to the joint Statistical Advisory Council for approval.\\
616
617The ESS contains information on economic activities, access to services and resources, household income, well-being, socioeconomic characteristics of households in the rural area and small towns; and household farming activities along with other information on individuals and other households indicator variables like human capital access to services and resources, household income, well-being, socioeconomic characteristics of households in the rural area and small towns; and household farming activities along with other information  and so on.\\
618
619The Ethiopia Socioeconomic Survey (ESS) collects detailed information on agricultural practices and labor activities; multiple dimensions of well-being, including health, wealth, and education; as well as data on food consumption, food security, socioeconomic characteristics of households in the rural area and small towns; and household farming activities along with other information on individuals and other households indicator variables like human capital access to services and resources, household income, socioeconomic characteristics of households in the rural area and small towns; and household farming activities along with other information  and so on. The first wave of data was collected in 2011/12 from about 4,000 households in rural and small towns. These same households were re-interviewed in 2013/14, along with an additional 1,500 households in urban areas(total 5,262) and in 2015-2016 along with 4,954 households in rural, urban and small towns.
620 \begin{quotation}
621Because the survey is designed to follow the same households over time, we are better able to learn what factors, policies, and programs help people to improve their productivity and wellbeing, \end{quotation} said Ato Amare Legesse, Deputy Director General of the Central Statistical Agency Report (CSA Report, 2015). 
622%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
623\subsection{General Objective of the Study}
624The overarching goal of this study is to analyze the settings and matters of overlapping  Multidimensional Poverty Index and economic well-being(ill-being) of Ethiopia's Households using the AF approach and utilizing the three waves of Ethiopian Socioeconomic Survey (ESS) (2011-2012, 2013-2014 and 2015-2016). Plus, to examine the status of MPI poverty profile of Ethiopian households.
625\subsubsection{Specific Objectives of the Study}
626The specific objectives are:
627\begin{enumerate}
628\item To determine Ethiopia's Official Multidimensional Poverty Index  (MPI) profile at national/federal level, regional level, identify its deprivations among the poor  Ethiopian households and household members.
629\item To breakdown and decompose MPI by dimensions and subgroups.
630\item  To compare and assess multidimensional poverty levels/status among four  major regional  states (namely, Amhara, Oromia, SNNP and Tigrai) of Ethiopia and test whether  multidimensional overlapping deprivation and  MPI indices/components are the same these regional states. 
631\item To examine the decomposition of  MPI by subgroups (region, gender, age-group, area...) and its decomposition by dimensions and indicators.
632\item To explore the spatial distribution of MPI  poverty by Mapping that facilitates and enables targeting the deprived and the poorest more effectively. 
633\end{enumerate}
634%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
635%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
636\subsection{Research Contribution to the Existing Knowledge and Learning }
637%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
638While poverty has been studied employing the conventional uni-dimensional monetary poverty measurement approach, there exist wealth of literature on this realm and has been well documented in Ethiopia to a great extent, the available literature on poverty seldom provides a multidimensional perspective of poverty since money alone is incomplete measure of poverty and imperfect proxy indicator of human development.   
639As a matter of fact, monetary poverty is only one among several measures of deprivation. Decent living conditions, access to basic services, a minimum level of educational attainment and adequate safety nets are equally important measures of human well-being.\\
640
641Not only the over-reliance on the monetary dimension of poverty  risks under-estimating the real extent of poverty in the developing world context, where living conditions can be misrepresented by dis-economies of agglomeration, such as income, consumption/expenditure and aggregated HDI indicators and poverty measures. But also, in reality, a good understanding of the education, health, living standard, and other pertinent socioeconomic dimensions of poverty can provide policy-makers with more entry points for anti-poverty interventions. As an illustration, monetary poverty could better be addressed by tackling non-monetary aspects of poverty, such as improvements in housing quality, fostering inclusive growth, creating job opportunities by investing in the pro-poor programs, better access to basic services and so forth.\\
642
643Moreover, MPI increases the importance of policy coordination and multi-sectoral programs and thinking beyond money is becoming more important for addressing the puzzles of poverty. We have the vie that the use of Alkire-Foster multidimensional measurement of poverty provides valuable insights to studying  in the multidimensional perspective employing a multidimensional approach that will enable the Government of Ethiopia to create national measures of poverty and well-being. \\
644
645This measure has flexibility to include various indicators and dimensions  that evolve through time. As a result, it was broken down and decomposed by subgroups that can better reflect realities on the ground, facilitate policy actions and targeting the deprived, it contributes to in the break though thinking to unlock the thinking trap in the mono-dimensional/conventional monetary measures of poverty and the research findings will contribute to perspective of practitioners, government officers, academia, scholars, researchers in solving the multidimensional puzzles of poverty. \\
646In this study, we estimate and compile  Ethiopia's MPI employing the Alkire-Foster method, decomposed by subgroups and broken down by three domains and fifteen indicators and conceptually embedded  within the capability approach (Sen, 1985, 1992, 1999;  UNDP, 2010; Alkire and Santos, 2014; and Suppa, 2016). The normative judgment of capability approach is a basis for the ope-rationalization of MPI indexes, national and regional assessments of MPI poverty measures.\\
647 
648As a matter of fact, numerous studies such as (Wolrd Bank 2014, 2015, 2016) have documented poverty in Ethiopia using the one-dimensional monetary approach. Save for limited reports like  Global MPI Report (UNDP, 2010 and OPHI, 2015). To the authors knowledge, limited study has been conducted to analyses  MPI in Ethiopia employing the AF approach and hence our intention is to analyze MPI Poverty nationally, decomposed it by sub-groups and break it down by MPI indicators and domains in order to be able to track, profile and target Ethiopian MPI poverty measurement using the noble panel data set of ESS 2011-2012, 2013-2014, and, 2015-2016 that discuss poverty break down by dimensions and decomposition by region, age group, gender, area, religion, and race. Therefore, this research will address this gap and give new insight in to Ethiopian MPI deprivation and poverty measurements using parsimonious measures.\\
649
650The Ethiopian multidimensional poverty index is compiled using the Alkire-Foster method and conceptually nested within the capability approach (Sen, 1985, 1992, 1999). The capability approach helps to call  set the indexes while the MPI deals with the technical aspects of identifying and aggregating  multidimensionally deprived and poor Ethiopian households.\\
651  
652We contend that that a well-specified multidimensional poverty indexes represent human well-being better than conventional resource-based approaches where vital aspects of poverty are shielded under the buses of average composite poverty analysis.\\
653 
654We find Ethiopian MIP profile is much higher that the official reports under the monetary approach. Considering gender, MPI for women headed households are higher than men headed Ethiopian households. As one would  generally expect MPI poverty in the rural area is higher than urban area. However, when we decomposed poverty by regions, the most resources reach and hub of all Ethiopian business activities, the Oromia region, is the most MIP poor, it's  significantly and substantially higher as compared with the other four major regions. \\
655In addition, we also documented MPI poverty by different age groups, MPI indicators and domains.\\
656Nevertheless, we also documented a substantially higher MPI poverty as compared with the Ethiopian poverty reports in the World Bank, MPI and other international organizations. Again, we could not find a substantive supportive evidence to claim that Ethiopian poverty in the Global MPI was higher since lowers poverty cut-offs was used as it was claimed in the World Bank Report regarding Ethiopia's Economy (World Bank report, 2015). We used a very high poverty cut-offs (K=40$\%$), our estimation result are closer to the global MPI in the UNDP and OPHI reports.\\
657 
658Finally, the findings of this research will contribute to reforming curriculum and modules that deal with development economics in  general and multidimensional measurements of poverty and well-being  and evaluations  of antipoverty programs outlined and implemented by Regional Governments and Federal Governments  of Ethiopian that are currently being taught in the Ethiopian  Higher Educations Institutions. 
659%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
660%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
661\subsection{Research Dissemination Mechanisms/Implementations Considerations}   
662
663We understand that dissemination and extension of the research work are strenuous and challenging. Effective dissemination here is contextualized as a phenomena that engages the recipient in a process whether it is one of increased awareness, understanding or commitment and action.\\
664
665Our first endeavor towards materializing and achieving the pillars of dissemination explained above is establishment of a steering committee of key informants, eminent and key figures, academia and research institutions in the fields of multidimensional poverty measurement, development economics, and sustainable development etc.  We engaged these critical mass to have common goals and create consensuses and we plan to reach out to the community.  After we assessed the impact at this stage and enriched the existing information, knowledge, and practices, the scaling up and scaling out phases will follow suit.  \\
666
667On the other hand, we  communicate and influences to other projects that have been carried out in similar areas in order to avoid the danger of overlapping or conflicting activities. Since we strongly believe that it is extremely useful to make contact with other projects and think about how we might maximize resources, such as hosting a national conference on a theme jointly that much better use of staff time and resources.\\
668
669We considered all other dissemination windows of opportunities including in informal setting. Finally, the final report shall be reproduced in CD-ROM, paper materials, online resources, mass media, radio, tv, magazines and most importantly their articles will be published in the British Journal of Environment and Climate Change, Journal of African Economies, African Journal of Economic and Management Studies,  African Journal of Economic Review, African Journals Online: Economics and Development, and Journal of Special Issues/OPHI. Besides convenient types of dissemination media will be considered using, mailing lists, newsletters, briefings, conferences, e-mail, reports, and workshops, one-to-one dialogues/conversations, emailing base lists, websites, and other Medias. 
670
671%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
672\subsection{Limitation of the Study}
673The pillar limitation of this study is that it does not address the conventional monetary approach, nor does it compare the monetary and multidimensional measurement of poverty as this subject will be addressed in a separate chapter (chapter 2) as it provides critical additional insight. Additionally, it cannot be corroborated to resolve the data related problems prevalent in developing countries in general and in Ethiopian in particular. \\
674
675The rearrangement properties hold true given the basic assumption under the rearrangement properties is that either all the dimensions are substitutes or all dimensions are complements that is highly constrained and simplistic assumptions. However,  further research is required on how rearrangement properties behave if the fundamental assumptions are relaxed/violated and on the practical aspect of  situations when some some poverty indicators are substitutes in one dimension but complements in the other dimension and or both.\\
676 
677Although we have employed a noble research method that covers almost all major indicators of overlapping multiple deprivation and poverty of Ethiopian households and is the best representative data in terms of coverage and depth, accuracy, legitimacy and precision, timeliness and relevance, completeness and comprehensiveness, availability and accessibility and granularity and uniqueness; and so forth. However, these noble data do not include information on relevant to economic opportunities, economic autonomy,  social capital(networks, bonding, feeling of trust and safety, participation, self-esteem, sense of belonging,of being heard, symbolism, peace, decent/quality of life, \dots); lack of psychological dimensions such as powerlessness, voicelessness, dependency, shame, humiliation and what have you. \\
678
679
680MPI Critic\\
681hhhhh\\
682zzzzz\\
683wwwwww\\
684aaaa\\
685(more discussions here)\\
686
687\subsection{Organization of the Study}
688The study is organized as follows: section one is the broad back ground, introduction, justification, overview of the multidimensional poverty in broad ways and pertinent to Ethiopia; purpose of the study, general and specific objectives and limitation of the study. Section two deals with Literature Review in general consisting of types and sources of literature review subsection that in turn includes  integrated literature review, Overview of Ethiopian Economic Strategies and Review of Socioeconomic Factors, Ethiopia's policy related, data related and development challenges and so forth. Besides, Theoretical and conceptual literature review, empirical and systematic literature review, methodology related literature review, and historical/chronological literature review. \\
689
690Section three by and lager presents research methods and approaches that comprises subsections research design, data sources, research questions and hypothesis. Equally important, the results and finding section comprises subsections multidimensional poverty index in Ethiopia, comparing Ethiopia's MPI with Other One-dimensional monetary measures, Ethiopia's MPI decomposed by dimensions and indicators, mapping Ethiopia's national and regional MPI indices, and analysis of dominance tests. More over, this section includes, Estimation Results and Interpretations, and discussions, implications of contribution of each indicator to Ethiopia's MPI;  examining the difference between MPI incidences of poverty and monetary absolute poverty measurement in Ethiopia and Ethiopia's MPI poverty indexes. Finally, conclusions and implications for Policy and/or Further Research section which in turn consists of conclusions, implications for policy and further research and recommendations. 
691
692%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
693\newpage
694 \section{Literature Review}
695 
696 \subsection{Understanding Multidimensional Poverty Index}
697 This section briefly describes what the research topic is why this topic, highlights the importance of the topic, succinctly makes general statements regarding the topic, and presents an overview of the currents status of research subject.\\
698In order to create a general understanding of the topic, we conflated spectrum of views to provide operational definition of Multidimensional Poverty Index (MPI) that fits to our purpose. Modifying and expanding the definition given by Alkire et al. (2015), the Multidimensional Poverty Index (MPI)can  broadly and roughly be defined as the state or condition of a person or community of being homelessness, landlessness, joblessness, lack of access to basic human necessities, lack of economic outcomes, economic opportunities, economic autonomy, and economic resources ( such as wealth, employment, infrastructures,\dots); lack of social resources (health services, clean water, improved electricity, education, transport, sanitation services, credit, information, cultural goods, \dots); lack of social capital(networks, bonding, feeling of trust and safety, participation, self-esteem, sense of belonging,of being heard, symbolism, peace, decent/quality of life, \dots); lack of psychological dimensions such as powerlessness, voicelessness, dependency, shame, humiliation, \dots). It identifies multiple deprivations at the society, household and individual level in health, education, standard of living and so on. It uses micro data from household surveys; unlike other multiple indicators of poverty measurements (for example the  Inequality-adjusted Human Development Index where all the indicators needed to be constructed from the measure that must come from the same survey). Each person in a given household is classified as poor or non-poor depending on the number of deprivations his or her household experiences. These data are then aggregated into the national measure of poverty. The MPI reflects both the prevalence of  overlapping multidimensional deprivation, and its intensity how many deprivations people experience at the same time (UNDP, 2015). \\
699%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
700\subsection{Types and Sources of Literature Review}
701%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
702By and large, we relied highly on primary and  secondary sources as the sources of data is secondary noble and well representative data. Research summaries reported in the text book forms, journals (printed and open access), articles (in news papers and scholarly journals),scholarly literature review and review articles, book reviews,  newspapers, online websites, visual and audio materials and so forth were among the main sources for our literature review part and the entire study.  \\
703
704Moreover, serial publications (journals, magazines, and news papers) consisting of primary sources; books( both functioning as as primary and secondary sources); visual and audio material(visual materials such as maps, photographs, prints, graphic arts, ... and digital recordings, documentaries, tv news broadcasts) were also reviewed as potential sources to some extent. \\
705
706Furthermore, archival materials  (manuscripts, archives, diaries, journals, photographs, interactive maps,) as primary sources; government documents (statistical compilations of economics, demographic and scientific data) that provide evidences of activities, functions, and polices were used as primary source. Finally, we also reviewed tertiary sources (guide books, indexes, abstracts, manuals and textbooks)\\
707
708The following section succinctly presents different types of literature review pertinent to the study. Brief systematic analysis and synthesis of wealth of literature relevant to the study are considered, integrated and interpreted aiming at giving the general picture and overview of matters and stetting around resource based monetary approach of poverty measurement and the associated draw backs.  The strength of MPI are highlighted so that it enables one to grasp the value added contribution of this study to the exiting knowledge.
709%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
710\subsection{ Empirical and Systematic  Literature Review}
711
712Scholars and researchers have divergent views and are becoming skeptical over the use of one-dimensional poverty analysis particularly in the previous two decades. There is a debate on what a comprehensive perspective of poverty is, how it can be measured in the lens of  multidimensional perspectives, settings and matters, how it can be compared with the conventional monetary measure of poverty and its status as compared to the traditional ways of poverty measurements. In a nut shell, the poverty measurement in a broader concept has been receiving much attention (Ruggeri Laderchi et al. 2003; Ravallion, 2011; Ferreira, 2011; and Meyer and Sullivan, 2012). \\
713
714Moreover, there have been extensive empirical studies on this issue such as Klasen (2009), Baulch and Masset (2003), Asselin and Vu(2008), Duclos et al.(2006) and Alkire and Santos(2014).
715They contend that there are multifaceted settings and matters that wrecked and lacerated human lives and those enormous issues beloved to be the root cause of poverty are intertwined that demand a comprehensive cross cutting solutions. \\
716In such setting markets usually fragmented, distorted, incomplete and imperfect (Tsui, 2002; Bourguignon and Chakravarty, 2003; Thorbecke, 2008) and monetary values are not good indicators and representatives of human well-being (Hulme and McKay, 2008; Thorbecke, 2008). Equally important, there were  convincing claims that state having sufficient income that can procure a basic baskets of goods doesn't diametrically imply it is going to be spent needlessly on those baskets of good (Thorbecke, 2008).\\
717
718What is more, poverty has been defined by one-dimensional and conventional monetary measures, such as income, expenditure, expenditure and other similar direct measures. However, there is a broad consensus that no one indicator alone can capture the multiple aspects that constitute poverty. As a result of this dynamism in thinking, replacing the previous Human Poverty Index (HPI), the Global Multidimensional Poverty Index (MPI) was developed in 2010 by the Oxford Poverty and Human Development Initiative (OPHI) and the United Nations Development Program and uses different factors to determine poverty beyond income-based lists. The global MPI is released annually by OPHI and the results published on its website.\\
719
720Multidimensional poverty is made up of several factors that constitute poor people's experience of deprivation – such as poor health, lack of education, inadequate living standard, lack of income (as one of several factors considered), dis-empowerment, poor quality of work and threat from violence. A multidimensional measure can incorporate a range of indicators to capture the complexity of poverty and better inform policies to relieve it. Different indicators can be chosen appropriate to the society and situation.\\
721 
722Multidimensional poverty analysis has also ventured beyond purely academic discussions into the broader policy debate, both within a number of national, regional and global frontiers. In the last five years, for example, Mexico's National Council for the Evaluation of Social Policy (CONEVAL,) adopted a multidimensional index as the country's official poverty measure and the Government of Colombia followed suit by adopting a poverty reduction strategy focused on five separate dimensions, and relying on a variant of the Alkire and Foster (2011a) approach for quantifying progress in 2011 (Francisco Ferreira and Maria Ana Lugo, 2012). Internationally, the MPI of Alkire and Santos (2010) which was reported by over 100 countries in the UN-DP's Development Report 2010, has also gained dominance.\\
723
724The MPI is an expansion of the one dimensional class of decomposable poverty measures proposed by Foster, Greer and Thorbeck (1984) developed and proposed by Alkire and Foster (2007). 
725The global MPI uses information from ten gauges which are structured into three dimensions: health, education and living standards, accustoming the same dimensions and weights as the Human Development Index (HDI, 2010). Every person is identified as poor or non-poor in every yardstick based on a deficiency cutoff (for detail notes refer to Alkire and Santos 2010). Health and education standards reflect attainment of all household members in their effort to salivate out of poverty. Then, every households deficit score is built based on a weighted mean of the denial they experience by means of a nested weight structure: The same weight across dimension and the same weight for every standard within dimensions. \\
726
727Three dimensions of the global MPI are equally weighted in the process of calculating the MPI, so that every indicator obtains a  weight. Thus, every measurement within the health and education dimension has a  weight and every indicator within the living standards dimension has a   weight.  Owing to missing/omission of data, when there are fewer than 10 indicators in the global MPI, the principle of equal value across dimensions and across indicators within a dimension still applies.\\
728 
729The mushrooming literature on the use of the MPI now includes studies such as Alkire and Foster (2011a), Chakravarty, Deutsch and Siber (2008), Deutsch and Silber (2005), Duclos, Sahn and Younger (2006) and Maasouri and Lugo (2008). There has been a vigorous debate on the conceptual and empirical virtues and demerits of the MPI since it was first included by the HDRO in the annual Human Development Report in 2010 (see, for example, Lustig, 2011; Silber, 2011; Alkire and Foster, 2011b; Rippin 2010; Ravallion, 2011, Bossert, Chakravarty, and D’Ambrosio, 2012).\\
730
731MPI has become an area of great debate for many prominent development economists and researchers although no one seems to dispute the point that, like development, poverty deprivations exist in multiple domains, and are often strongly correlated.  
732To date, save for some  reports, yet, to our knowledge no academic studies have estimated the MPI for Ethiopia. By merging data set 2004-2009 from the health, education and other sectors separately; the Oxford Poverty and Human Development Initiative reported the MPI to be 0.564, the percentage of MPI poor to be 87.3$\%$, and the average intensity across the poor to be 64.6$\%$ (OPHI, 2015). However, research on the dynamic MPI, inequality among the poor, and a country-wide map of MPI has never been conducted, and this thesis will endeavor to fill this gap using the recent ESS panel data sets. 
733\subsubsection{Brief Review Literature on Ethiopia's Socioeconomic Factors}
734
735In Ethiopia, agriculture is the main contributor and backbone of the Ethiopian economy contributing 84$\%$ as primary sources of food and income, 85$\%$ contribution to export sector, 73$\%$ contribution to employment, supplies 70$\%$  of the raw material requirements of local industries and 39$\%$ contribution to GDP and many other allied economic activities such as marketing, processing, expansion of aggro-processing industrial parks, blossoming of small and medium business enterprise, expansion and promotion  of export products and services depend on the agricultural sector (USAID-Ethiopia, 2015).\\
736
737Historically, in Ethiopia, rural poverty outweighs (see the next table below) and  only  20.5(21, 174, 205, 2017 est.) percent live in urban areas and historically poverty in rural of Ethiopia have been higher than urban poverty.
738\begin{quotation}
739...Although most Ethiopians are rural dwellers and subsistence farmers, the poorest 40 percent tend to be even more likely to live in rural areas and engage in agriculture. While educational attainment among average Ethiopians is low, it is even lower for the bottom 40 percent (1.5 years) compared to the top 60 percent (2.8 years)... (World Bank, 2016)
740\end{quotation} 
741
742Consequently, propositions that state urban poverty have offsetting effect to the gains in  the poverty reduction in the agricultural sector do not seem correct. On the other hand, according to World Bank Report (2014, 2015 and 2016)for the last decade, Ethiopian Agricultural growth ranged between 7.5-9.5 $\%$. Like wise, USAID-Ethiopia (2016) reported that Ethiopia has experienced strong and broad-based economic growth over the past decade, averaging 10.9 percent per year, over double the regional average.\\
743
744\begin{quotation}
745In addition, the high GDP growth has been relatively inclusive, as it has uplifted about 20 million people out of poverty and facilitated the achievement of most of the Millennium Development Goals (MDGs) targets, Ethiopia, Ministry of Finance and Economic Development (MoFED, 2016)
746\end{quotation}
747The following tables 5, 6 and 7 below reveal the contribution of each sectors and status of various socioeconomic indicators. \\
748
749The lack of infrastructural development, including roads and communication systems prevent farmers from being able to access markets. With the distorted markets and lack of these markets farmers struggle to obtain modern agricultural inputs such as seed, fertilizer, and tools which are critical to developing a modern farming system leaving the agriculture of Ethiopia at a barely sustenance level.  \\
750
751Poverty reduction has been the core objective of the Ethiopian Government. The most recent major pushes and  overarching policy response to agricultural productivity challenges, deep rooted poverty, and food insecurity in Ethiopia and broad strategies aiming at paving the ground for industrialization via stimulating agricultural output productivity, achieving robust economic  growth and creating a strong bond between the agricultural, manufacturing and industrial output and productivity, enhancing the service delivery in the public sector and other allied sectors greed towards  reducing poverty  under the umbrella of Agricultural Development-Led Industrialization (ADLI) strategy some how mobilized the Ethiopia's Economy. To this effect  some major Ethiopia's Economic Strategies have been crafted although their full scale implementation is debatable. \\
752
753To mentions some of the macro-level strategies, An Economic Development Strategy for Ethiopia (1994); Industrial Development Strategy (2002;   Sustainable Development and Poverty Reduction Program (SDPRP, 2002/03-2004/05); A Plan for Accelerated and Sustainable Development to End Poverty (PASDEP, 2005/06-2009/10); Growth and Transformation Plan I (GTP I, 2010/11-2015) and Growth Transformation Plan II (GTP II, 2015/16-2020) are among the major noes: Sources: MoFED Reports (2010, 2015/16).\\
754\!\begin{landscape}
755\begin{table}
756 \caption{ Review of Ethiopia's Poverty, Inequality, Well-being and sector-wise of Employment}
757\label{table5}
758\begin{tabular}{ |p{13cm}||p{2.5cm}|p{2.5cm}|p{2.5cm}| }
759\hline
760\multicolumn{4}{|c|}{Ethiopia: Poverty, inequality, well-being and sector of employment, 200 to 2011 }\\
761\hline
762 \hfill                                 & 2000 & 2005  &   2011\\
763\hline
764National absolute poverty headcount (National Poverty line)  &      44.2 $\%$ &   38.7$\%$ & 29.6$\%$    \\
765\hline
766Urban  & 36.9$\%$   &  35.1$\%$ & 25.7$\%$ \\
767\hline
768Rural  & 45.4$\%$  & 39.3$\%$  & 30.4$\%$\\
769\hline
770International extreme poverty headcount (US$\$$1.25 PPP Poverty Line) & 55.6$\%$ & 39.0$\%$ & 30.7$\%$ \\
771\hline
772Population (thousands) & 63,493 & 71,066 & 84,208 \\
773\hline
774Number of people living below the national poverty line (thousands)& 28,064 & 27, 066 & 25,102\\
775\hline
776Poverty depth (Measured at National Poverty Line)& 11.9$\%$ & 8.3$\%$ & 7.8$\%$\\
777\hline
778Urban & 10.1$\%$ & 7.7$\%$ & 6.9$\%$\\
779\hline
780Rural & 12.2$\%$ & 8.5$\%$ &8.0$\%$\\
781\hline
782\end{tabular}
783Sources: World Bank (2015)
784\end{table}
785\end{landscape}
786
787The series of tables (table 5) above and (tables 6-8) below present Ethiopian selected socioeconomic indicators and sector wise contributions. 
788In spite of all these positive progresses and  growth stories, Ethiopia has been suffering from chronic food shortages that have led to malnutrition, disease, famine migration and loss of lives. These food shortages are induced and precipitated by droughts, floods and soil depletion caused by climate change and biodiversity degradation (USAID-Ethiopia, 2014/15).  \\
789
790\subsubsection{Are  Ethiopia's "Unique Economic Strategies" Delivering High Growth?}
791
792According to a World Bank report (2015), "Ethiopia's Great Run; the Growth Acceleration and How to Pace It", it is reported that the agriculture and services sectors, were the main contributors to this accelerated growth, which was driven by high government investment in the energy, transport, communications, agriculture and social sectors.
793\begin{quotation}
794Over the last decade, Ethiopia has made remarkable progress in its economic growth exceeding other low income and Sub-Saharan African countries, with real gross domestic product (GDP) growth averaging 10.9 percent in 2004-2014. Ethiopia has moved from the second poorest in the world in 2000 and, if it can keep the current pace, it's on its way towards becoming a middle income country by 2025. Ethiopia's growth strategy stands out for its uniqueness in focusing on promoting agriculture and industrial development with a strong public infrastructure drive (World Bank, 2015)
795\end{quotation}
796
797Table 6 below, reports status of Ethiopian well-being Indicators. Although it looks very low as compared with other low income countries, these uni-dimensional monetary poverty measures indicate a better situation contrary to the country's multidimensional poverty reports. According to the Oxford Poverty and Human Development Initiative (OPHI, 2011), Ethiopia's Incidence of Poverty/Multidimensional Headcount Ratio \textit{(H)} was reported as 88.6$\%$, Average Intensity Across the Poor \textit{A} was revealed as 63.5$\%$ and the Multidimensional Poverty Index $(MPI=H{\times}A$ was reported as 0.562. 
798
799\begin{landscape}
800\begin{table}
801\caption{Ethiopia's Well-being Indicators: Poverty, Inequality, Employment }
802\label{table6}
803\begin{tabular}{ |p{13cm}||p{2.5cm}|p{2.5cm}|p{2.5cm}| }
804\hline
805\multicolumn{4}{|c|}{Ethiopia: Poverty, inequality, well-being and sector of employment, 200 to 2011 }\\
806\hline
807Poverty severity (Measured at National Poverty Line) & 4.5$\%$ & 2.7$\%$ & 3.1$\%$\\
808\hline
809\hline
810Urban & 3.9$\%$ & 2.6$\%$ & 2.7$\%$\\
811\hline
812Rural & 4.6$\%$ & 2.7$\%$ &3.2$\%$\\
813\hline
814Gini coefficient/National & 0.28 &0.30 & 0.30\\
815\hline
816Gini coefficient, Urban & 0.38 & 0.44 & 0.37\\
817\hline
818Gini coefficient, Rural & 0.26 & 0.26 & 0.27\\
819\hline
820Nutritional outcomes among children under 5 years of age &&&\\
821\hline
822Stunting & 58$\%$ & 51$\%$ & 44$\%$\\
823\hline
824Wasting & 12$\%$ & 12$\%$ & 10$\%$\\
825\hline
826Underweight &41$\%$ & 33$\%$&29$\%$\\
827\hline
828\end{tabular}\\
829Sources: World Bank (2015)
830%\listoftables{table5}
831\end{table}
832\end{landscape}
833As a matter of facts, apart from the initial market orientation reform after the downfall of the Military Junta in 1991, structural reforms have been limited to paper works and diplomatic language without prudent structural reform in the Ethiopian economic growth and development model. Although it was reported  as if Ethiopia has registered remarkable economic growth and arguably it has been the success story of developmental state economic models owing to the initial economic success, multidimensional poverty, sustainable and quality public services, gender inequality, risk prone primary economic activities,urban slams and so on remain the main challenge that hampered social transformation.\\
834
835\begin{quote}
836...although Ethiopia gradually moved in the direction of market oriented system, it continued to intervene in most sectors of the economy thereby not adopting some of the key recommendations....structural reforms have been absent from Ethiopia's growth strategy...inspired by the East Asian development State model and share some common features, it is also different from these countries both in conception and and outcomes (World Bank Report, 2016).
837\end{quote}
838
839On the contrary,  Ethiopia is still on the bottom of the low-income countries in the world, has the most fragile economic systems and is the most vulnerable climate change and disasters. According to the UN criteria, countries with less than $\$$400 level of per capita income countries are designated as low income countries and countries with less than $\$$750 per capita income as called less developed economics.  According to World Bank (2015) Report, Ethiopia is one of the lowest income countries with the Gross Domestic Product per capita 486.5 and 505.00 nominal per capita income U.S. Dollars in 2015. Thus, by all global and regional standards Ethiopian remains to be the low income country.\\
840
841\begin{landscape}
842
843\begin{tabular}{ |p{13cm}||p{2.5cm}|p{2.5cm}|p{2.5cm}| }
844\hline
845\multicolumn{4}{|c|}{Ethiopia: Poverty, inequality, well-being and sector of employment, 200 to 2011 }\\
846\hline
847 \hfill                                 & 2000 & 2005  &   2011\\
848\hline
849Proportion of households reporting shocks & & & \\
850\hline
851Life expectancy (years) & 52 & &63\\
852\hline
853Net attendance rate: Primary education (7-12 years of age)& 30.2$\%$ & 42.3$\%$& 62.2$\%$\\
854\hline
855Urban & 73.6$\%$ & 78.8$\%$ & 84.9$\%$\\
856\hline
857Rural & 24.3$\%$ & 38.8$\%$& 58.5$\%$\\
858\hline
859Immunization Rates (BCG, DPT1-3, Polio, measles) & & & \\
860\hline
861At least one shot & 83.5$\%$ & 76.0$\%$&85.5$\%$\\
862\hline
863All vaccines & 14.3$\%$ & 20.4$\%$ &24.3$\%$\\
864\hline
865Food price & n.a.& 2.0 $\%$ &19.0$\%$\\
866\hline
867\end{tabular}
868Sources: World Bank (2015)
869%\listoftables{table6}
870\end{landscape}
871
872
873\begin{landscape}
874\begin{tabular}{ |p{15cm}||p{2.5cm}|p{2.5cm}|p{2.5cm}| }
875\hline
876\multicolumn{4}{|c|}{Ethiopia: Poverty, inequality, well-being and sector of employment, 200 to 2011 }\\
877\hline
878 \hfill                                 & 2000 & 2005  &   2011\\
879\hline
880Proportion of households reporting shocks & & & \\
881Drought & n.a. & 10.0$\%$ & 5.0$\%$\\
882\hline
883Jobless & n.a. &1.0$\%$& 0.0$\%$\\
884\hline
885$\%$ crop loss (from LEAP) & 22.4$\%$ & 23.5$\%$ & 13.8$\%$\\
886\hline
887Share of population living in urban areas & 13.3$\%$ & 14.2$\%$ & 16.8$\%$\\
888\hline
889Proportion of households with at least one member engaged in & & & \\
890\hline
891Agriculture & 78.8$\%$ & 79.7$\%$ & 78.4$\%$\\
892\hline
893Services & 23.0$\%$ & 20.8$\%$ & 23.1$\%$\\
894\hline
895Industry  &  3.4$\%$  & 8.7$\%$  & 8.0$\%$\\
896\hline
897\end{tabular}\\
898Sources: World Bank (2015)
899%\listoftables{table7}
900\end{landscape}
901\subsubsection{Ethiopia's Policy Related, Data Related and Development Challenges}
902
903Although it's true that for the past 10-15 years, Ethiopia has been able to partly let fall its association with abject poverty and famine; it still remains to be one of the world's least income country. Despite Ethiopia's successes in poverty reduction, improvements in public services and outcomes;  development challenges remain extensive. Besides, although it seems true that the Government of Ethiopia (GoE) promotes a free market economy as it apparently claims, on the contrary, a birds-eye-view observation may reveal that following China's footprint and inspired by its state-led development, the incumbent government of Ethiopia fosters a developmental state with a with heavy hand and strong intervention of the state in the socioeconomic and market interplay.\\
904
905\begin{quotation}
906...Remarkable progress registered in Ethiopia is not without challenges, poverty remains widespread and the very poorest have not seen improvements. To the contrary even worsened, of consumption since 2005, which poses a challenge of achieving shared prosperity in Ethiopia. For example, prior to 2005, the growth in consumption in the bottom 40 $\%$ was higher than the growth in consumption of the top 60 $\%$  in Ethiopia.\\
907 
908However, this trend was reversed in 2005 to 2011 with lower growth rates observed the bottom 40 percent. Consumption growth benefited ,any poor households from 2005 to 2011, with the highest growth rates experienced by the decile below the poverty line. Nevertheless, the poorest decile did not experience an increase in consumption.  As a result, reduction in poverty  rates were not matched by reductions in poverty depth and severity  from 2005 to 2011. The negative growth of rate of the consumption of the bottom decile is robust to the choice of deflator and is concerning trend...(World Bank, 2015).\\
909
910\end{quotation}
911
912More importantly, Ethiopia's main development challenge continues to be abject poverty, unemployment and low-level socioeconomic service delivery and infrastructural development on top of  generating quality economic growth that will substantially reduce high poverty levels, particularly in rural areas. Despite the marked reduction in the country's poverty levels, an estimated 23 million (25 percent) of the population (measured by the traditional monetary measurement of poverty using income or consumption) continue to live below the poverty line for various reasons.\\
913
914Key among these are low agricultural sector production and productivity, low domestic resource mobilization, high
915levels of unemployment, especially among the youth, poor quality and inequitable distribution of basic social services, low access to finance, and a poor private business climate. Notwithstanding these challenges, Ethiopia has favorable economic opportunities and prospects. The country has abundant natural resources, a low cost and trainable labor force, an emerging middle class, and a developmental state with an ambitious vision, commitment, and strong sense of policy ownership UK Department for International Development(DFID-Ethiopia, 2016 report).\\
916
917The development process that Ethiopia puts in place  must ensure that every citizen lives in any part of the county or s/he subscribes to any faith in, whether s/he speaks any language or has any race and color; every citizen is entitled to same respect, same dignity, same decent life and same development... which the country can afford to any one else. It's demanding and depends on the countries endowment and inertia of the past. \\
918 
919As a matter of fact,  Ethiopia has seen a declining poverty over the last two decades, which is even higher the Horn of Africa Region. On the contrary, previously, it was focusing on the consumption part of poverty which is very inconclusive and incomplete form of measuring poverty. This approach also hides the dirty spots of poverty under the average analysis.\\
920
921But the problems can be addressed by choosing multidimensional poverty index (MPI) methods. By modeling multidimensional poverty alleviation strategies, fostering breakthrough thoughts to avoid being trapped in the main stream thinking on the conventional monetary poverty measures and thinking beyond money is  indispensable for resolving the poverty puzzles (Haider and Lade, 2017). \\
922
923Furthermore, by focusing on factors beyond just food intake and by putting more responsibility on the executive branch of the government as it's expanding the base, domains and angles from which poverty is analyzed. By taking the broad based and measurable responsibility, the government can show its commitment to reduce, control and eliminate poverty. By considering a more holistic concept of  good quality of life because people are entitle to not just basic components of food/diet, but are entitle to other basic need of life  that are critical for their success and development process as well.\\
924
925Likewise, despite the strenuous efforts poverty is still rampant and Ethiopia remains at the bottom of low income countries category. This could be explain maybe as the  policy was very much centerfold on certain  macroeconomic indicators yet failed to touch the most fundamental reality of development which is not about numbers. Development is about people, the basis of development strategies should be the target people and it ought to be done via the direct engagement of the people. It is of people, for people and by people. \\
926...xxxx...\\
927Development strategies on the poor should always be the priorities and governments must keep focus on these strategies. Robust, parsimonious, transparent and accountable system should be promoted to protect resources misdirection, stifling and embezzlement as the development endeavors ought to  start with people and it ends with people. No matter how good numbers look, if there is no change in the life of people across the board, then, that development is ending up benefiting the elites or to the powerful. Development is only development when it can touch and transform the lives of the weak, vulnerable  and marginalized. \\
928
929%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
930Moreover, the World Report (2015) revealed that deprivations in the in the health and education  is still quite high since Ethiopia face the problems of low rates of educational enrollment, access to enhanced sanitation and attended births. In the 2011 Global MPI assessment, the percentage share of the poor in Ethiopia was reported 87 $\%$, which is second poorest country in the world (OPHI, 2014). In other words, almost nine out of ten people were reported as MPI poor at least in one third of the weighted MPI indicators employed in the Global MPI.  These figures are fundamentally in stark contrast to the official reports of the Government of Ethiopian (GoE) and socioeconomic reports in the World Bank, IMF, and other Multilateral Development Banks pertinent to Ethiopia. Hence, the sources of these discrepancies and the debates around these discrepancies need to be given due attention and addressed properly.  \\
931
932\begin{quotation}
933The higher rates of poverty and slow progress recorded in the MPI (in Ethiopia) arise largely because of the divergence of between monetary poverty and the measures of the living standards used in the MPI. This divergence is due, in part, because the assets considered in the MPI do not include assets important in Ethiopia and the cutoff used in some dimensions is too high to reflect the recent progress (in Ethiopia)(World Bank, 2015).
934\end{quotation}
935
936However, we argue that the huge discrepancy between the conventional uni-dimensional poverty measurement and the Global MPI (that was measured my Oxford Poverty and Human development Initiative, 2015 and 2016) and the national MPI that we developed and assessed Ethiopian status of multidimensional intertwined in webs of interrelated and overlapping deprivations and  poverty indicators in living standards, human development and public/social services (education, health, mortality and exposure to risk and so on) domains in Ethiopia is not limited to the divergence between monetary poverty measurement and measures of living standard employed in the Global MPI as claimed  in the World Bank Report (World Bank, 2015).\\
937
938%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
939Further more, the intuitive question that we can ask here considering the following scenarios and stylized facts are: agriculture contributes a big share to the Ethiopian economy; the Ethiopian government continues to advocate state ownership of land whereby only usufruct rights (usufruct rights exclude the right to sell or mortgage the land) are bestowed upon  small landholders (Crewett et al, 2008); the present Constitution of Ethiopia, which was put into effect since January 1995, vests land ownership exclusively "in the State and in the peoples of Ethiopia."; the traditional and rain-fed agricultural productivity is limited to small household fragmented land holders; modernizing land ownership by giving title either to the peasants who till the soil, or to large-scale farming programs and mechanized frames are not existent; in Ethiopia, as in other developing countries, land has been considered as an important economic and social asset where the status and prestige of people is determined;..., then the logical arguments of gains of  economic growth are enjoyed by the upper segment(the rich) while lower strata are left behind as it was documents from experience from other countries (South Africa as a point in case)does not seem true as the Ethiopian economic growth was praised for its  inclusiveness.\\
940
941Why poverty in Ethiopia is still too deep and there exist abject poverty despite of the remarkable growth for more than a decade need further investigation  as it might lead to misleading conclusions and misinformed policy directions. In such circumstances, then what is the sources of this huge discrepancy between the conventional monetary measurement of poverty and MPI?\\
942%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
943Thought this research is limited to the national multidimensional over lapping deprivations and poverty measurement and analysis in Ethiopia, possible reasons for the discrepancy/divergence between the conventional and uni-dimensional monetary poverty measurement demand a rigorous and objective research independently ; we will suggest some plausible causes from observations of specific instances.\\
944
945\subsubsection{Challenges of finding authentic and reliable data in Developing countries}
946
947Many Ethiopians are skeptical regarding the the reliability, authenticity and data accuracy of  Ethiopian government owned statistical data as there might be Economic data manipulations, faking statistics  and faked fiscal, preferential and rigged statics reporting from other lower Federal and Regional Government employees or deliberate economic data manipulating from the top government officials. \\
948
949Rigged data reporting from the lower level might emanate from the lower levels as there is no well calibrated, technologically monitored, harmonized and unified central statistical reporting system. Likewise, when data is collected at local levels and go up to the federal structure until it reaches the Federal Agencies or Ministries, the bureaucrats in charge of that system benefit professional success and personal advancement when their numbers are big and meet the expectations of higher officials and party members. In conclusion, advocacy and policy implication based on estimates and counts of the poor and poverty measurement using the one dimensional measure of poverty in developing countries in in general and in Ethiopia in particular must be implemented with great care although the numbers are arguably crucial. This hold true to statistics archives in multilateral Development Banks(MDBs), multilateral Development agencies and other institutions for statistics. 
950\begin{quotation}
951...The world Bank publishes estimates of the numbers of poor people in the world. While everyone knows hat these numbers should be take with a pinch of salt, the numbers are arguably important... (Deaton, 2000) 
952\end{quotation}   
953
954\subsubsection{Lack of transparency in Data collection, Administration, Dissemination and Repositories}
955
956Here we observe two type of difficulties of data transparency. Either these problems may emanate from within the organization or endogenous (within official statistics), lack of institutional set, organizational set, poor performance  and lack of internal and data quality control, monitoring, evaluation and auditing system or beyond the control of the organization (external) or exogenous such as non existence/limited participation of civic society, lack of establishments and enterprises that independently and objectively - without any vested interest and ulterior motive-collect, administer and disseminate authentic, reliable, accurate and consistent data. \\
957Poor recording/unreasoning some important data, lack of proper, net worked and adjustable repository, archive and capacity development on how to use them:  for example demographic  and income related data and  lack of/limited institutions, reliable internet access, free  press or other check and balance mechanism (such as civic society, labor union and so forth) to validate the accuracy of data collection, administration and dissemination.
958Researchers, economists, investors, policy makings, donors and other stake holders ought not take the face value of such data  at might lead to a naive conclusions and that will result in missing to hit the target and intended objectives.\\
959
960Moreover, illiteracy, deficiency in law and order regarding data rigging, manipulation, made up and so forth, cultural constraints, practice of traditional and informal economic activities, heterogeneity of the society and living conditions, lack of standardized units of measurements, high variance and volatility of economic variables such as prices, wages, consumption, limited investment on establishing  secondary meta data storage alternatives, poor infrastructural and internal communications, insufficient legal frameworks and comprehensive and up to date directories. For detail discussions on comparative analysis of the one-dimensional and multidimensional poverty measure, you are kindly requested to refer "Why the new emphasis on MPI measurement?" under the Thematic Literature Review section.  \\
961\subsubsection{Levels and trends of Income Poverty and MPI}
962
963Correspondingly, there is a mismatch/discrepancy between monetary indicators of and social indicators of derivations. The sources of the variation could possibly be difference in the magnitude of market development and penetration; level of infrastructural development, culture, values, norms, and awareness of the rural population, geographic location and resources base etc.\\
964 
965Under those circumstances, evidence indicate that income poverty does not predict social deprivations and material and social deprivations do not predict each other strongly.  Neither do poverty and MPI headcount ratios  trend together. The graph below clearly depicts that one dimensional monetary poverty measure and multidimensional deprivation and poverty measures and do not necessary go together although they are complementary in many scenarios. 
966\begin{figure}
967\caption{Income and MPI Time Series Analysis from Previous Works Review}
968\label{Figure7}
969\begin{center}
970\includegraphics [scale=0.50]{INCOME_MPI_2017.pdf}
971\end{center}
972\end{figure}
973Sources: (Alkie, 2016)\\
974To conclude, MPI  is  a  very  adaptable  methodology  that  is amenable to  incorporate alternative  indicators,  cutoffs  and  weights  that  might  be  appropriate  in  regional  national, or  sub-national  contexts (Santos E. and Alkire, 2011).\\
975
976\subsubsection{Extremely Low Resources Base of The economic Growth and Rapid weather variability}
977
978As a point in case, an anticipated 25 million people were in need of emergency food aid in East Africa(UN, 2015). Of which over 10 million Ethiopians at this time need emergency food aid (FAO, 2015). Wealth of literature have documented as the Ethiopia economy has grown rapidly over the last decade, but from an extremely low base, and remains highly dependent on and vulnerable to the weather: 80$\%$ of the population of 100+ million are rural, and the vast majority make their living through rain fed agriculture(as marginal smallholder farmers in the highlands or pastoralist or agro-pastoralist herders in the lowlands).\\
979
980This is partly because the Horn of Africa, where Ethiopia is located, is the most drought vulnerable area on the continent. An anticipated 25 million people in East Africa are presently in need of emergency food aid (UN, 2015). Of which over 10 million Ethiopians at this time need emergency food aid (FAO, 2015).  
981  
982Nonetheless, progress has not been consistent across the country, and households that are dependent on agriculture remain vulnerable. This is partly because the Horn of Africa, where Ethiopia is located, is the most drought vulnerable area on the continent.\\
983 
984%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
985
986\subsubsection{Biodiversity depletion and  and Environmental Sustainability} 
987
988Most commentators believe that the most challenging environmental problems for Ethiopia are land degradation and climate change. 
989Fluctuations in biodiversity and climate change affects all walks of life but it seriously affects, the developing world  since the rural poor count on natural food sources produced traditional and are highly vulnerable to climate changes owing to poverty, recurrent drought, population surge, and inequitable land distribution, exhaustive utilization of natural resources, subsistence rain-fed erratic agriculture, and so forth.\\
990
991Equally important, climate change will have tremendous impacts on biodiversity, from ecosystem to species level. The most noticeable repercussion of climate change on biodiversity is that flooding, sea level rise and changes in temperature and ecosystem boundaries.  Under those circumstances shifts in a boundary, some ecosystems will expand into new areas; at the same time, others will become smaller. Owing to those variations, rainfall and temperatures habitats will change, and some species will not be able to keep up, leading to a sharp increase in extinction rates.\\
992%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
993\subsubsection{Corruption in Ethiopia}
994
995Scoring 34 points out of 100, Ethiopia is the 108 least corrupt nation out of 176 countries, according to the 2016 Corruption Perceptions Index by (Transparency International, 2016)\footnote{An example footnote}.  The result highlight the strong bond between corruption and inequality, which feed off each other to create a vicious circle between corruption, unequal distribution of power in society, and unequal distribution of wealth. 
996 \begin{quote}
997 In too many countries, people are deprived of their most basic needs and go to bed hungry every night because of corruption, while the powerful and corrupt enjoy lavish lifestyles with impunity.\end{quote} José Ugaz, Chair of Transparency International. The corruption, rent-seeking, and asset-grabbing for which Ethiopia is known are unlikely to disappear unless cultural, political, socioeconomic, and societal transformations are put in place.  \\
998 
999Despite Ethiopia having one of the fastest-growing economies in the world, it also remains one of the world's poorest countries owing to multifaceted root causes. As an illustration, corruption has played a huge role in hindering the development of the Horn of Africa nation. There seems wide consensus about the areas vulnerable to corruption in Ethiopia; markedly, facilitation payments and bribes being necessary to win and keep land leased from the state, land administration and distribution, tax, revenue and custom administration, low level and high level tender and contract administration, housing and natural resources developers, public procurement, the justice system, telecommunications, land procurement, licensing areas, police, public service, under pricing of national and international tenders, illicit financial flow, and the finance sector. Source: The Federal Ethics and Anti-Corruption Commission of Ethiopia website, accessed on January 30, 2017\\
1000
1001In conclusion, widespread and deep rooted general poverty,  high dependence on agriculture, underutilized natural resources, lack of industries and enterprises, lack of capital and technology, lack of basic infrastructures, vicious circle of poverty, enormously economically inactive demographic characteristics, inequality and discrimination in cultural characteristics, and dualistic economy in least developed countries in general and in Ethiopia in particular may never be addressed effectively without a well-functioning  multidimensional poverty analysis.\\
1002
1003As a result, it become more difficult to target multidimensional development and sustainability as  development is multifaceted and complex. It can be hard to strive for and measure the growth of a country/countries, when there are a variety of contradictory measure as it relates to development and progress. Therefore, there is an urgent and concrete  global, regional and national demand to shift from the main stream uni-dimensional poverty analysis to multidimensional poverty analysis. \\
1004
1005%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1006\subsection{Theoretical and Conceptual Literature review}
1007 \subsection{The Thematic Literature Review}
1008Measuring poverty in the multidimensional perspective has eminent importance because the target group, trends and poverty tracking, conclusions, implications and policy measures may differ greatly. Owing to this and other reasons, emphasis on multidimensional poverty measure, in general, and the AF approach together with the capability approach, in particular  in particular are becoming predominant (Suppa, 2016, Alkire and Santos, 2014; and UNDP, 2010).  
1009   
1010\subsubsection{Why the new emphasis on MPI measurement?}
1011In addition to the  normative, empirical and policy motives that explain the latitudinal status and why MPI is surging as a much preferred approach studying poverty form multiplicity setting and matters and hence the development of   multidimensional poverty analysis as the best resolution to the drawbacks in the uni-dimensional monetary measurement of poverty; according to (Alkire et. al, 2015 ) are the following among the others:
1012 \noindent We can \hfill Technical
1013\begin{enumerate}
1014  \item Data availability
1015 \item Computational and Methodological Developments\\
1016 \noindent It adds information: \hfill Empirical
1017 \item Monetary and Non-Monetary Household Deprivation Levels
1018 \item Trends in monetary and non-monetary deprivations
1019\item Associations across non-monetary deprivations
1020\item Economic Growth and Non-income Deprivations\\
1021\noindent It improves action: \hfill Policy
1022 \item National and International policy 'demand'
1023 \item Political space for new metrics
1024\end{enumerate}
1025Oxford Poverty and Human Development Initiative (OPHI, 2016) 
1026\begin{enumerate}
1027  \item Data availability:
1028Since the 1980s, access to more than 700 development indicators for more that 208 countries; more than 800 panel household surveys covering more than 120 countries, massive meta data on multiple household socioeconomic survey and other big data are available openly in the public domain that and the existing technology to process and analyses these data contributed to the emergence of MPI.
1029   \item Computational and Methodological Developments:\\
1030The extensive acceptance and  application of the Alkire-Foster MPI and the computational simplicity using the exiting software packages has become simple and much simpler that it was in the past. 
1031 \item Empirical Motivation:\\
1032 
1033Income poverty study was considered as a good proxy for the last 200 years. However, if we analyze income poverty vis-à-vis socioeconomic indicators and  psychological situations of poor households; the former (income/consumption) indicators do not necessary predict neither the socioeconomic nor the psychological status of poor households and realistic aspects of poverty. Wealth of literature have indicated that there is a mismatch between income poverty and social deprivations. Thus, it is argued that income is not a good proxy for material and other social deprivations.    
1034 \item Computational and methodological developments:\\
1035Increases of data availability together with increased computational power have led to the generation of new indices. To mention some of the most common well-being, poverty and inequality measures from the literature:
1036  \begin{itemize}
1037   \item HDI, IHDI, GDI, Canada Index of Well-being, etc.
1038   \item Going Beyond GDP Initiative,
1039   \item Global Peace Index and related,
1040   \item SIGI and gender-related,
1041   \item Doing Business Index,
1042   \item Good Governance, Transparency International, Mo Ibrahim,
1043   \item Social Protection, Global Hunger, Happiness,
1044 \end{itemize}
1045 \item Monetary and Non-Monetary Household Deprivation Levels:\\
1046 
1047For many years, income poverty was considered as a good proxy to the over all state of poverty. However, when we compare the income poverty  and the social poverty analysis to study the state of deprivations and poverty; we understand that there is a mismatch between income poverty and other social deprivation indicators. consequently, we deduce that income is not a good proxy for material deprivations and other social socioeconomic and psychological  deprivations (for more detail information, See Nolan and Whelan, 2011). \\
1048
1049According to Nolan and Whelan(2011), they reported that while 20 percent of people were tenaciously income poor, and 20 percent were continuously materially deprived. They further reported that only 10 percent of people are both persistently income poor and materially deprived. Further more, we believe that poverty like development has to be to be investigated from various spectrum of multiple domains and indicators of deprivations and poverty.  The exiting similar empirical evidence and our justification motivated this research to choose multidimensional material, social and psychological indicators as income does not reveal the full story of these factors.\\
1050 
1051In a like manner, other inherent problems associated with monetary measurement include 
1052\begin{itemize}
1053 \item non-sampling measurement error (accuracy)
1054 \item time and cost of survey (data collection)
1055 \item problem of comparability by rural-urban, age groups, international, and other social and cultural groups. 
1056 \item does not show how people are poor
1057 \item price distortion
1058 \item problem comparability and decomposability 
1059 \item it does not conquer with the notion that  poverty must be dealt village by village 
1060\end{itemize}
1061\item Trends in monetary and non-monetary deprivations:\\
1062Various authors such as François Bourguignon et al.(2010)  that assessed  trends of the Millennium Development Goals (MDGs) have argued that the trends of $1$/ day poverty did not match trends in other MDGs (Alkire, 2016)
1063 \item Associations across non-monetary deprivations:\\
1064 
1065The logical question that we have to look in to under this category is whether we can find a single non-monetary measure that captures all the monetary and non-monetary material and social deprivation. Again, we contend that it is betters to address the non-monetary deprivation from multidimensional perspective likewise. In the light of this, we believe that there is no bellwether measure that proxies the rest and indicate the trend.  \\
1066
1067 \item Economic Growth and Non-income Deprivations:\\
1068Empirical studies indicate that countries that enjoyed sustained economic growth over the last 30 years dis not proportionately reduce social problem of children malnutrition (underweight, stunted growth,waste); earning inequality, women empowerment and so forth.Equally, important, empirical studies reveal that there are areas where monetary poverty is zero while multidimensional poverty is very high. On the others hand, other countries with lower growth had made greater progress in social indicators (Alkire, 2016). As a result, it could be summarized that would not either be a better proxy. 
1069\item Levels and trends of Income Poverty and MPI:\\
1070Correspondingly, there is a mismatch between monetary indicators of deprivations and social indicators of derivations. The sources of the variation could possibly be difference in the magnitude of market development and penetration; level of infrastructural development, culture, values, norms, and awareness of the rural population, geographic location and resources base etc.\\
1071 
1072Under those circumstances, evidence indicate that income poverty does not predict social deprivations and material and social deprivations do not predict each other strongly.  Neither do poverty and MPI headcount ratios  trend together.
1073By and large, with all the imperfection and draw back in the MPI, it is the better indicator as it helps for targeting the poor in policy design and implementation with the existing data at our disposal. In essence, MPI can be employed by developing countries, policy design, implementations for better resources allocation and coordination; subgroup individual and geographic targeting, targeting precision  and monitoring and evaluation of development packages, programs and projects.\\
1074
1075\item National and International Demand:\\
1076On 1 December, 2016, High Level Panel Endorses Multidimensional Approach to Poverty at the Global Partnership Meeting in Nairobi; Nairobi, Kenya. \\
1077The Second High-Level Meeting (HLM2)held from 28 November- 1 December 2016, in Nairobi, Kenya  aims to amplify the positive impact of development co-operation over the next 15 years. (Nairobi Outcome Document, 1 December 2016 Nairobi, Kenya)\\
1078
1079\item Political Space for New Metrics/Measurements :\\
1080To strengthen policies that fight poverty. In particular, MPI supports these SDG priorities.\\
1081 Multidimensional poverty measures are becoming increasingly popular. So far over 50 countries around the globe have adopted  and implemented national or local Multidimensional Poverty Indices (MPIs) to assess their development strategies and track poverty. Poverty measures created using approaches such as the Alkire-Foster method (developed by OPHI) are attractive for those in charge of developing public policies because different dimensions and indicators can be selected to create specific measures for particular contexts (MPPN, 2016).
1082 \begin{itemize}
1083 \item Integrated, coordinated policy (break Silos)
1084 \item Inclusiveness and dis-aggregation by groups
1085 \item Universality-acute and moderate poverty
1086 \item Data Revolution do-able, and adds value
1087\item Global Monitoring complement $\$1.25$
1088 \end{itemize}
1089 \end{enumerate}
1090%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%  
1091 \subsection{Methodological Literature Review}
1092There has been much debate about the conceptual and methodological drawbacks of the conventional/uni-dimensional  monetary poverty measures following the seminal work by Sen (1979, 1981) on the capability approach. There has been a concerted effort to develop multidimensional poverty measures so as to respond to the urgent need for new methods of broad perspective poverty measurement. Particularly, there has been profound studies on issues such as theoretical studies by Sen(2000), Tsui(2002), Atkinson(2003), Bourguignon and Chakravarty (2003), and Alkire and Foster (2011).\\
1093
1094Proponents of the capability approach asserted that progress in development indicators and human development are more crucial than the changes in inputs like income and consumption. They further claim that analysis of multidimensional deprivations and poverty  and poverty should gravitate around lining standard and socioeconomic indicators such as household assets, stocks, and economics outcomes in lieu of focusing on economics opportunities, and  on flows or inputs (Clark and Hulme, 2005; Hulme and Shepherd, 2003) and more broadly on employ indirect poverty measures   (Baulch and Masset, and 2003; McKay and Lawson, 2003)\\
1095
1096Winsome axiomatic properties of the A methodologies have generated unique opportunity for poverty measurement.
1097Applications of AF go beyond poverty and also include energy, corruption, resilience, time use, well-being, empowerment, health, and so on (Alkire, 2016).\\
1098
1099Three core domains that are regarded as components of MPI are constructed of various indicators under each domain are robust and informative with respect to each domains that reflect the state of  multiple deprivation of individuals, households, communities etc. The three dimensions in this study were selected on normative as we as statistical considerations and are equally weighted  since each dimension  has a relative importance in terms  of intrinsic values. Under the health domain, four indicators were selected and half out of the overall weight assign to this dimension was given to the first indicator whereas the remaining half is equally shared by the rest three indicators. Likewise, under the education domain, the three indicators are  equally weighted under normative considerations. Finally, under the living standard domain the seven indicators are equally weighted by the same token.
1100
1101\subsubsection{Uni-dimensional Poverty Measurement}
1102
1103For centuries, the vast majority of empirical poverty studies have been using uni-dimensional poverty measurement of well-being, predominantly either using household's consumption or expenditure, or its income.  The majority of empirical studies on poverty profiles in Ethiopian have employed a one dimensional measure of well-being. Consumption rather than income is viewed as the preferred welfare indicator because consumption better captures the long-run welfare level than current income, and may better reflect households ability to meet basic needs. Income is only one of the elements that allow consumption. Consumption reflects the ability of households to access credit and savings at times when their income is very low. Hence, consumption reflects the actual standard of living (welfare). Consumption is also usually better measured in surveys than income (MoFED, 2014). \\
1104
1105Furthermore, the one dimensional approach, which deals with only the monetary aspect, does not reveal the whole picture of households deprivations for the following reasons:  First, the pattern of consumption behavior may not be uniform, so that attaining the poverty line level of income does not guarantee that a person will meet his or her minimum needs. Second, different people may face different prices, reducing the accuracy of the poverty line. Third, the ability to convert a given amount of income into certain functioning varies across age, gender, health, location, climate and conditions such as disability i.e. people's conversion factors differ. Fourth, affordable quality services, such as water, health and education, are frequently not provided through the market, and failing to take into account government provision of such services may overstate poverty. Fifth, employing the indirect measure of poverty gives no way to verify the intra-household distribution of income of the households. Sixth, participatory studies indicate that people who experience poverty describe their state as comprising deprivations in addition to low income. Finally, from a conceptual point of view, income is a general purpose means to valuable ends. Measurement exercises should not ignore the space of valuable ends (Sen, 1979 cited in Alkire and Santos, 2013).\\
1106
1107
1108In order to mitigate the disadvantages of the one-dimensional poverty measure, a number of approaches have been proposed to measure or analyze deprivations in multidimensional perspectives. The assessment of multidimensional occurrence poverty like development goes back to the influential works of Amartya Sen (1979, 1985 and 1987). More recently, the multidimensional poverty index (MPI) measures are gaining ground as the canonical measures of poverty, as absolute and relative monetary indicators, such as the income, expenditure or consumption approaches, may not only give a poor measure of the actual experience of poverty but may also lead to wrong policy implications (Alkire and Foster, 2007 and 2011a and Alkire and Santos, 2010).  \\
1109 \subsection{Succinct Multidimensional Poverty Measurement Literature Review }  
1110 
1111MPI poverty measure do not rely on marginal measures, but on
1112the joint distribution of deprivation. Moreover, the censored headcounts do not reflect the raw headcounts (marginal measures). They rather reflect identification and censoring of the non-poor. In other words, all  consistent sub-indices in the MPI produce $g_0(k)$ but not $g_0$. Unlike mainstream uni-dimensional poverty measure MPI do not make aggregation base on attainment such weighted prices, but rather based on deprivation space. To explain further, fundamentally multidimensional
1113aggregation is not possible in attainment space owing to draw back of lack of standards of units, wide spectrum of  achievement domains such as education, health social exclusion, safety, security and  decent living standard are betters captured in terms of deprivations of opportunities. \\
1114 
1115In a more abstract view, the uni-dimensional measure poverty has been widely used and a trend line despite of all its draw backs. Whereas MPI is the headline that fundamentally better captures major part of the matters and settings of individual, group and societal deprivations and hence aggregate ought to be done according to a concept of poverty as multiple deprivations, with explicit deprivation values and trade-offs and joint distributions. Consequently, although prices may be used when they are available and meaningful; MPI is by far the better option  when prices may not be available across all dimensions of poverty and are not meaningful (S. Alkire,  2011). Employing weighted prices for poverty aggregation becomes quite difficult when information on prices are not adequately addressed (there is no standard way of doing so and it extremely difficult to do it) in the existing Mega data available in the public domain. \\
1116
1117\subsubsection{Dashboards}
1118Although the dashboard approach consists of  more than 50 indicators, its  allure is inversely proportional to the multiple indicators of poverty and/or well-being. This approach doesn't only extrude vital information in poverty analysis but also the indicators are insensitive to the joint distribution of deprivations and could be useless for measuring extreme forms of poverty and destitution.\\
1119
1120To explicate the drawbacks with the dashboard approach to multidimensional poverty when the extrude vital information since they are insensitive to the joint distribution of deprivations, let's consider two hypothetical countries "E" and "S" below
1121    $$Country "E"=
1122%\begin{center}
1123\begin{blockarray}{rccccc}
1124 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} \\
1125\begin{block}{>{\small}r[ccccc]}
1126{1} & 0 & 0 & 0 & 0 \\
1127{2} & 0 & 0  & 0  & 0 \\
1128{3} & 0 & 0  & 0  & 0 \\
1129{4} & 1 & 1  & 1  & 1  \\
1130\end{block}
1131\end{blockarray}
1132%\end{center}
1133$$
1134and 
1135$$Country "S"=
1136%\begin{center}
1137\begin{blockarray}{rccccc}
1138 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} \\
1139\begin{block}{>{\small}r[ccccc]}
1140{1} & 1 & 0 & 0 & 0 \\
1141{2} & 0 & 1  & 0  & 0 \\
1142{3} & 0 & 0  & 1  & 0 \\
1143{4} & 0 & 0  & 0  & 1  \\
1144\end{block}
1145\end{blockarray}
1146%\end{center}
1147$$
1148
1149As we can see in the two scenarios above, the dashboard analysis would deduce a 25 percent deprivation in each indicator. As a result,  a naive conclusion may be drawn 25 percent multidimensional poverty in both countries. Nevertheless, using a more sophisticated and sensitive analysis would result as in the case of multidimensional poverty measurement and analysis would arrive at a different conclusion. To shed light on this, employing the union approach the MPI in both countries could be profiled that 100 percent whereas using the intersection approach, it's 25 percent in country "E" but 0 percent in country "S". 
1150This is the critical draw back of the dashboard multidimensional approach of poverty measurement and analysis. \\
1151
1152"It suffers because of their heterogeneity, at least in the case of very large and eclectic ones, and most lack indications about… hierarchies among the indicators used. Further, as communications instruments, one frequent criticism is that they lack what has made GDP a success: the powerful attraction of a single headline figure allowing simple comparisons of socioeconomic performance." Stiglitz-Sen-Fitoussi as mentioned in (Alkire, 2011).
1153
1154It's further argued that not only the dashboard approach does not catalyze expert, political, or public scrutiny and debate on these trade-offs, nor
1155promotes transparency and accountability; but also, it leaves the difficult questions about trade-offs completely open nor does it explain how the mechanisms of trade off takes place. \\
1156
1157Of course has been instrumental in measuring poverty, percentage of  the poor and has practical applications under Basic Needs Approach and Millennium Development Goals (UN, 2000) since it has advantage in sharing information on many dimensions of poverty, can draw on various data sources and can show information on disjoint populations, but dashboard approach is not only blind to joint deprivations but also this method of marginal measures does not answer the fundamental questions of a poverty methodology such as who is poor overall?; how many poor people are there?; and how poor are they?\\
1158
1159In other words, this uni-dimensional measure to each dimension shows the average share of the poor and can show information on disjoint populations; however, this has numerous shortcoming such as lack of hierarchies among the indicators;lack a single headline figure (such as GDP);leave the questions about trade-offs completely open; does not identify who is poor; and ignores joint distribution even when could reflect it  and so on (Alkire, 2016)\\
1160
1161\subsubsection{Composite Indexes}
1162
1163Composite idices poverty measure is a class of  single
1164summary measure. The prose of this measure includes providing a summary measure, useful for comparisons, ordering; can combine different data sources (superior to the dashboard in many cases); can combine information on disjoint populations; can draw on normalized indexes; and offering a hierarchy and make trade offs explicit (see Ravallion 2011).
1165
1166On the other hand,  this approach like the dashboard approach is plugged to drawbacks like implicit or no Identification and ignores joint distribution even when possible to capture. \\
1167
1168\subsubsection{Joint Distribution}
1169
1170This approach had been widely employed to study and trace the global distribution of income (both within and between countries) with a wide coverage of per capita income and life expectancy emphasis on multidimensional indicators of welfare poverty and inequality and their attributes.  This measure takes in to account  multiple measurement of poverty at a time and it was the main tool employed to monitor and evaluate progress in the MDGs.
1171
1172Furthermore, It argues that that individual well-being and social welfare
1173depend on the joint distribution of various attributes, and analysis based on a single attribute does not capture adequately the different dimensions of welfare. 
1174Welfare analysis  of multiple attributes by examining each individual attribute separately fails to account for the dependence among various attributes (Ximing Wu et al., 2008).
1175The critical draw back is it measures the marginal Distribution
1176(without reference to other dimensions) and hence variation  in the distribution of achievement matrix along row matrix (persons in this case) will result in a different outcome. In other words, Can have same marginal distributions and very different joint distribution? It is observed that the same marginals, different joint distribution result in very different situation in terms of poverty analysis. \\
1177
1178\subsubsection{Venn Diagrams}
1179
1180This welfare  tool has been used to study overlaps of deprivations in
1181different dimensions of multiple indicators of poverty.
1182The  strength of this tool is that it is a visual tool to explore overlapping binary deprivations, Considers the joint distribution of deprivations, intuitive and easy to understand for 2-4 dimensions. Whereas the downside are may not identify who is multidimensionally poor, no summary measure (thus, no complete ordering),regardless of the scale, every dimension is converted into the binary options, losing information on depth, difficult to read for 5 or more dimensions.\\
1183\subsubsection{Dominance Approach}
1184
1185This approach is employed as a poverty measurement tool in both the 
1186uni-dimensional example Atkinson (1987), multidimensional for instance Duclos, Sahn $\&$ Younger (2006) demonstrated how to make poverty comparisons using multidimensional indicators of well-being and how to check robustness of these comparisons to aggregation procedures and to the choice of multidimensional poverty lines.\\
1187
1188It confirms that whether poverty is evidently lower or higher regardless of parameters and poverty measures. The leeway of this tool is that it voids the possibility of contradictory rankings, Offers tool for strong empirical assertions about poverty comparisons, considers the joint distribution of achievements/deprivations, avoids controversial decisions on parameter values. However, it has significant drawbacks as well such as No summary measure, No complete ordering allows pair-wise dominance, but not cardinally meaningful difference, dominance conditions depend on relationship between dimensions, for 2+ dimensions, limited applicability for smaller data sets, stringent less intuitive conditions for dominance beyond first order (Alkire, 2016).\\
1189
1190 \subsubsection{Statistical Approaches}
1191
1192This tool is used for poverty identification, poverty aggregation, or both and are employed  during measurement design for the purpose of exploring relationships across variables and setting weights. Employing this too as practical advantages in poverty measurement as  addresses multidimensionality, Considers joint distribution,Multiple Correspondence Analysis (MCA) that can be used for the construction of composite indicators from multiple primary poverty indicators, and the computation of poverty and inequality indices with the composite indicator; can be used with ordinal data, and helps clarify relations among indicators: strengthen indicator design.\\
1193
1194However, this approach has the following drawbacks as well. Poverty identification and measurement are often not transparent, not straightforward for communicating, not checked for robustness, identification is mostly relative (based on percentiles of the score), comparisons across space and time may be difficult, no automatic normative or theoretical justification (Alkire, 2016). \\
1195 
1196\subsubsection{Fuzzy Sets Approach}
1197
1198This welfare analysis too, Fuzzy set approach explore how to be vaguely right argued Zadeh (1965), Cerioli $\&$ Zani (1990), Cheli $\&$ Lemmi (1995), Chiapero-Martineti (1994, 1996, 2000. It's a similar approach and is a type an extend Venn-diagram focusing on visual tool to explore overlapping binary deprivations in lieu of  identifying the deprived/
1199non-deprived or poor/non-poor, allow varying degrees of
1200membership $m_j{(x_{ij})}$ to each set.
1201This tool has typical strengths like offers summary measure, hierarchy among dimensions, explicit trade-offs and a complete ranking, can consider joint distribution of deprivations, and compatible with many aggregation methodologies (Chakravarty 2006).\\
1202Nevertheless, this approach has weakness such as justification of membership function is not straightforward,robustness tests are not mostly provided, some membership functions may misuse ordinal data, and  Fuzzy sets results may conflict with Dominance results. 
1203
1204\subsubsection{Axiomatic Approach}
1205
1206Since the publication of Sen's (1976) pioneering article on axiomatic method to measurement of poverty.
1207This tool develops poverty measures that comply with a number of
1208desirable properties in both the uni-dimensional and multidimensional poverty measurement by specifying a threshold level for all the attributes  and shortfalls of quantities of different attributes from respective thresholds for different welfare settings of individuals are aggregated in to an over all indicator of poverty. \\
1209
1210The advantages of this  approach include allows looking at joint distribution of deprivations, offers summary measure of poverty, provides clearer understanding on how measures behave due to different transformations (biggest advantage)where as the downside are relies on normative judgments (require various robustness tests), no single measure can satisfy all desirable properties (properties themselves often need strong justifications),final poverty measures can be difficult to interpret intuitively when they are made to satisfy many properties simultaneously (Chakravarty, 2006).\\
1211
1212\subsubsection{Counting Approach}
1213
1214This tool of poverty has been widely used in practice and policy since mid 1970s (Alikre, 2016). The measurement of poverty involves identification: the fundamental step of deciding who
1215is to be considered poor. 'A counting approach is one way to identify the poor in multidimensional poverty measurement, which entails the intuitive procedure of counting the number of dimensions in which people suffer deprivation' (Alkire et. al., 2015)
1216
1217It refers to a particular method for identifying the poor, entails 'counting the number of dimensions in which people
1218suffer deprivation, (…) the number of dimensions in which
1219they fall below the threshold', Atkinson (2003). 
1220Steps for identifying a poor person in the Counting Approach
1221\begin{enumerate}
1222\item Define a list of relevant indicators
1223\item Assign a weight to each considered indicator
1224\item Define a threshold (deprivation cutoff) for each indicator
1225\item Create binary deprivation scores for each person in each
1226indicator: "1" = deprived, "0" = non-deprived
1227\item Produce a deprivation score by taking a weighted sum of
1228deprivations
1229\item Set a threshold (or poverty cutoff) such that if a person has a deprivation score at or above the threshold, the person is
1230considered poor
1231\end{enumerate}
1232It is important to elucidate pros and cons Counting Approach as follows:
1233This approach  has advantages such as advantages such as having clarity, simplicity, transparency, intuition for identifying the
1234multiply deprived, allows looking at joint deprivations, and allows for both cardinal and ordinal variables. That said, to illuminate the disadvantages of this tool, it relies on the particular selection of indicators (appropriateness for the particular purpose), relies on the weights assigned to the dimensions/indicators, relies on dichotomies (deprived/non-deprived) so not sensitive to the depth for identification, sometimes a counting approach is combined with aggregation methodologies that are not intuitive, in certain cases incorrectly assigning cardinal meaning to ordinal values (for instance, poverty scorecards)
1235
1236\subsection{AF Method}
1237We select the AF methodology over other unitary and multidimensional deprivation and poverty measurements as the multidimensional poverty Index (MPI or ($M_0$) satisfies the following properties and axioms of multidimensional poverty measures:
1238 \begin{itemize}
1239  \item Symmetry
1240  \item Scale invariance
1241  \item Normalization
1242  \item Replication invariance
1243  \item Ordinality
1244  \item Poverty Focus
1245  \item Deprivation Focus
1246  \item Weak Monotonicity
1247  \item Weak Deprivation Re-arrangement
1248  \item Dimensional Monotonicity
1249 \item Decomposability
1250 \item Dimensional breakdown.
1251 Furthermore, the AF methodologies uniquely fulfills the following properties and are largely responsible for its growing use:
1252 \item Subgroup Decomposability, by which an assessment of subgroup
1253contributions to overall poverty can be made, facilitating regional
1254analysis and targeting;
1255\item Dimensional Breakdown, by which an assessment of dimensional
1256contributions to overall poverty can be made after the poor have
1257been identified, facilitating coordination;
1258\item  Ordinality, which ensures that the method can be used in cases
1259where variables only have ordinal meaning; and
1260\item Dimensional Monotonicity, which ensures that the removal of any poor person's deprivation reduces poverty, even if the person remains poor.\\
1261
1262Notably, properties of Multidimensional Poverty fall into three major  categories:
1263First, Invariance properties isolate aspects of the data that should not be measured. Second, subgroup properties connect poverty levels overall to levels obtained from data broken down by population subgroup or by
1264dimension. Third, dominance Properties concern the aspects of the data that should be measured and ensure that the poverty level responds
1265appropriately to certain changes in achievements (Alkire, 2016)\\
1266
1267Invariance properties isolate aspects of the data that should not
1268be measured.
1269 \item Symmetry (invariance to permutations of achievement vectors
1270across people),
1271 \item Replication In-variance (in-variance to replications of achievement
1272vectors across people)
1273\item Deprivation Focus (invariance to an increment in a non-deprived
1274achievement),
1275\item Poverty Focus (invariance to an increment in an achievement of a non-poor person).
1276\item Ordinality (allows meaningful evaluations of poverty when
1277variables are ordinal)\\
1278Subgroup properties connect poverty levels overall to levels
1279obtained from data broken down by population subgroup or by
1280dimension.
1281 \item Subgroup Consistency: (if poverty rises in a population subgroup and stays constant in the remaining population, while subgroup population sizes are unchanged, then overall poverty must rise)
1282\item Subgroup Decomposability: (overall poverty is a population weighted sum of the poverty levels in population subgroups).
1283\item  Dimensional Breakdown: (after identification has taken place and the poverty status of each person has been fixed,
1284multidimensional poverty can be expressed as a weighted sum
1285of dimensional components).\\
1286Dominance Properties concern the aspects of the data that should
1287be measured and ensure that the poverty level responds appropriately to certain changes in achievements.
1288 \item Weak Monotonicity (an increment in a single achievement cannot increase poverty)
1289\item Weak Rearrangement: (a progressive transfer among the poor
1290arising from an association-decreasing rearrangement cannot
1291increase poverty).
1292\item Dimensional Monotonicity: (requires poverty to fall as a result of an increment that removes at least one deprivation from among the poor)
1293 \end{itemize}
1294
1295By all means, MPI enables  to have targeted,  integrated, coordinated and inclusive policy designs, implementation, assessment and monitoring in a flexible and amenable manner by breaking the existing silos  that are barricades  to information sharing that facilitate learn from best innovative development endeavors around the globe and adoption to foster the cultural values, health services, enthronement, safety and security, and other fundamental compensates of human dignity. 
1296
1297 \subsection{Historical/Chronological Literature Review}
1298MPI is gaining wider acceptance to measure  the 17 SDGs, 169 target and 231 indicators. To put it differently, MPI is a better option to fill key sustainable development gaps left by the MDGs, such as the multidimensional aspects of poverty, decent work for young people, social protection and labor rights for all since poverty measures should reflect the multidimensional nature of poverty (Alkire, 2016).
1299   \begin{itemize}
1300   \item Poverty
1301   \item Hunger $\&$ Nutrition
1302   \item Health $\&$ Well-being
1303   \item Education $\&$ Learning
1304   \item Gender $\&$ Empowerment
1305   \item Water $\&$ Sanitation
1306   \item Energy
1307   \item Growth $\&$ Decent Work
1308   \item Infrastructure $\&$ Innovation
1309   \item Inequality
1310   \item Urban areas
1311   \item Sustainable consumption $\&$ production
1312   \item Climate Change
1313   \item Oceans $\&$ Seas
1314   \item Ecosystems $\&$ Biodiversity
1315   \item Peace $\&$ Justice
1316   \item Global Partnership
1317   \end{itemize}
1318In the $69^{th}$ session of the UN General Assembly, a resolution of the UNGA (A/RES/69/238) on 19 December 2014 reasserted the need for multidimensional poverty measurement as a necessary conceptual framework for the global community to measure and tackle extreme poverty.\\
1319
1320Chiefly, UNGA underlines the need to better reflect the multidimensional nature of development and poverty, as well as the importance of developing a common understanding among Member States and other stakeholders of that multidimensionality and reflecting it in the context of the post-2015 development agenda. Notably,it invites Member States, supported by the international community, to consider developing complementary measurements, including methodologies and indicators for measuring human development, that better reflect that multidimensionality (Alkire, 2016).\\
1321Another key points that indicate the wide acceptance of MPI is the Addis Ababa Accord of the Third International Conference on Financing for Development,(Addis Ababa Accord, 2015).
1322\\
1323
1324Markedly, it states that: 
1325\begin{quotation}
1326We further call on the United Nations, in consultation with the IFIs to develop transparent measurements of progress on sustainable development that complement GDP, building on existing initiatives. These should recognize the multidimensional nature of poverty and the social, economic, and environmental dimensions of domestic output. We will also support statistical capacity building in developing countries. We agree to develop and implement tools to monitor sustainable development impacts for different economic activities, including for sustainable tourism (Alkire, 2016).
1327\end{quotation}
1328
1329Equally important is  the Agenda 2063 Vision and Priorities. One of the top priorities  is to provide a robust framework for providing harmonized and quality statistics for the design and implementation as well as monitoring and evaluation of integration and development policies as well as development program in Africa. Indeed, this again reinforces the importance and wider broader practical implementation of the multidimensional measure of deprivations and poverty.\\
1330
1331Not to mention, in reports issued in November 2014, February and May 2015, the Sustainable Development Solutions Network (SDSN) presents the MPI as Indicator three of Goal 1 (“End poverty in all its forms everywhere). They asserted that so as to ensure the conceptualization of multidimensional poverty is firmly rooted in the Open Working Group Outcome Document and proposed SDGs, and they support the creation of a revised MPI. They went on saying that at a minimum this "MPI2015" would track extreme deprivation in nutrition, health, education, water, sanitation, clean cooking fuel and reliable electricity, to show continuity with MDG priorities.
1332
1333\begin{quotation}
1334We therefore propose using the Alkire and Foster method of calculation, and setting a threshold of multiple deprivations, to determine who is or is not considered poor (Alkire, 2016).
1335\end{quotation}
1336 
1337In addition, the The Atkinson Commission convened by the World Bank Chief Economist to Advise the World Bank on how to Measure and monitor global poverty inline to its goals of ending poverty; namely,to bring the number of extremely poor people, defined as those living on less than $\$1.9$ dollars a day (in 2011 Purchasing Power Party), to less than $3\%$ of the  world population by 2030. Not only to end  global poverty to  $3\%$ but also to boost shared prosperity, defined by promoting the growth of per capita real income of the poorest 40$\%$ of the population in each country has sought input from its advisers including from AF methodology.      
1338
1339%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1340%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1341\newpage
1342\section{Research Methods and Approaches}
1343From time to time, MPI is gaining wider acceptance in as much as its versatility, simplicity, and applicability to National and regional intertwined multiple deprivation and poverty measures, Global MPI, Governance, Targeting, Monitoring and Evaluation, Disaggregation, Communications, Innovative Participatory Methods and its inclusiveness of Private Sector and Civil Society. At the present time, over 50  countries have been developing and implementing the new multidimensional poverty analysis method in their stertorous endeavor to end/reduce poverty. \\
1344
1345There are  practical/empirical, policy implementation, ethical/normative and other multiple motivations for the development and adoption of the multidimensional poverty measurement. While laying the ground and emphasizing the importance of normative considerations, Sen (2000) stated \begin{quote}
1346Human lives are battered and diminished in all kinds of different ways, and the first task... is to acknowledge that deprivations of very different kinds have to be accommodated within a
1347general overarching framework.\end{quote} 
1348%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1349\subsection{Research Design:Meta-Analysis}
1350\subsubsection{Construction of Ethiopian MPI}
1351  \subsubsection{Data Sources}
1352The data sources for this study are multifaceted ranging from the ESS noble data, to various statistical archive such as UNESCO Institute for Statistics (UIS), World Health Organization(WHO), World Bank Group(WB), Ethiopia's-Ministry of Finance and Economic Development (MoFED), Ethiopia's-Central Statistical Agency (CSA) and so forth. Correspondingly, the approaches used in the data analysis vary from Stata 14, StatPlanet Desktop, Excel Spreadsheet software were applied in this study. Uniquely, the StatPlanet that enabled us to visually explore all the data sources within single intuitive interface, through integrated interactive maps and graphs. StatPlanet which consists of a powerful set of tools for automatically importing and visualizing Ethiopia's data pertinent to our study from various sources was meticulously applied in section 1 (subsection 1.4, 1.5 ...)  and  section 3 (sub section 3.7). As well, applied a blend of software and methodologies in the above mentioned sections and section 4 in particular. Similarly, multiple alternatives of data analysis(Stata 14 and excel spreadsheet) and presentation approaches like the LaTex software were intensively employed.\\
1353
1354Therefore, the best fit research design for such study is the meta-data-analysis as we used a statistical analysis of a large collections of data from multiple sources in order to analyze the body of evidences in our endeavor to answer the  fundamental questions of the study such as: what is the status of multidimensional poverty index (MPI) profile as compared to official reports based on uni-dimensional poverty measurement in Ethiopia?; whether multidimensional poverty is similar across regions as claimed by the official government reports or is there a significant difference among the four big regions of Amhara, Oromia, SNNP and Tigray?; Does multidimensional poverty remain robust across these major regions?; can an MPI  comparison be made among those regions?; and, does multidimensional poverty measures better allow for targeting specific MPI indicators, households and individuals? and so on.\\
1355
1356Furthermore, to shed light on more benefits of using the meta-data-analysis; it enables one to  establish statistical significance with studies that have conflicting results; to develop a more correct estimate of effect magnitude; to provide a more complex analysis of harms, safety data, and benefits;  and to examine subgroups with individual numbers that are not statistically. In short, it has greater advantages in greater statistical power, confirmatory data analysis, greater ability to extrapolate to general population affected, and considered an evidence-based resource.\\
1357
1358The meta-data-analysis research design will be the perfect fit as we are driving estimates of multidimensional poverty indexes from  various data sources pooled o as to produce a weighted average for the unit of analysis and unit of identification in the study. The advantage of using this approach is that aggregation of information usually leads to a higher degree of statistical power. In other words, combining estimates increases statistical power, computation of simulation models, produces fully transparent and reproducible synthesis and more robust estimates. Moreover, since we are integrating large collection of  data from various sources and statistical archives, a statistical analysis with a synchronized narrative discussions of research studies pertinent to this type of studies is the meta-data-analysis. \\
1359
1360Even if this is the best fit for our study as explained above, it is susceptible to some pitfalls such as, a meta-data-analysis of several small studies do not results of single large study; heterogeneity of effect sizes, heterogeneity of precision, selection, publication and quality control bias and so on.\\    
1361Sources: Gene V Glass (1976) "Primary, Secondary and Meta-Analysis of Research", Educational Researcher. 
1362\subsubsection{Exploring the ESS Data}
1363
1364Once the list of common variables in the panel data set has been identified, various visions of panel data set were created: the first $(called merge_panel_hh)$ contains only household-level variables, while the second $(called merge_panel_indiv_hh)$ contains both household-level and individual-level variables. 
1365As can be seen from the panel dataset archive, the unique identifiers are the variables $household_id$ and $individual_id$ $(See the file ERSS_Basic_Information_Document_Wave_2.pdf section 6.2.2 for details on merging into a panel)$.\\
1366 
1367It must be noted that Wave 2 includes some households in large towns, while Wave 1 did not.  By implication, these households have been dropped from the dataset before matching the waves.
1368%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1369The next logical steps that were performed while creating and building the MPI deprivation matrix. \\
1370
1371It should be noted that each vector provides information about the condition of deprivation of each individual in a specific indicator. The mean of this vector shows the incidence of each deprivation on the total population or (Uncensored Headcount Ratios)
1372\begin{enumerate}
1373 \item  Describe main characteristics of the dataset and each variable
1374\item Describe main characteristics of variables
1375\item Basic Command: The points mentioned below are some examples of basic commands. 
1376\begin{itemize}
1377
1378\item Categorical variable
1379\item Categorical variables with missing values
1380\item Two Categorical variables with missing values
1381\item Two Categorical variables cell results with missing values
1382\item Two Categorical variables row results with missing values
1383\end{itemize}
1384\item Set conditions and filters to analyze problems of missing data in each observation.
1385\end{enumerate}
1386Another simplified, yet powerful technique while creating and generating variables is that the significance of using loops that allow execution of the same command for a sequence of variables, numbers or other lists.\\
1387
1388In the context of Ethiopian household MPI, a thorough analysis of MPI dimensions and indicators was conducted in order to determine variables that best capture deprivation in the health, education and living standard domains. These include:
1389\begin{enumerate}
1390\item	scrutinize and analyze missing values:
1391 Final check to see the total number of missing values we have for each variable. The rule of thumb is these variables should not have high proportion of missing values at this stage. So as to carry out the missing value analysis the command might need to be installed by writing "findit mdesc" in the command window, and install it. 
1392\item Breakdown of the MPI Indices by population subgroup, region, dimension and other variables:\\
1393
1394Here, we analyzed the behavior of the MPI indicator among groups of interest, and check whether results re consistent to previous knowledge/evidence/survey reports
1395\item CRAMER's V Analysis:\\
1396 Cramer's V describes the association among indicators. It ranges between 0 and 1 where 
1397 \begin{itemize}
1398 
1399 \item 0 for the lowest possible association between variables, and
1400\item 1 for the largest possible association.
1401\end{itemize}
1402\item   Redundancy Analysis: \\
1403 Describes redundancy among indicators. The coefficient P is defined as the ratio between:
1404 \begin{itemize}
1405 
1406 \item the proportion of people with simultaneous deprivation in any two indicators, and
1407\item the lowest proportion of deprivation of those indicators independently.\\
1408In the redundancy analysis, the coefficient P takes values:
1409\item 0$\%$ when no one is identified as deprived in both indicators being considered, and
1410\item 100$\%$ when every individual who is deprived in the indicator with the lowest incidence of deprivation, is also deprived on the other indicator.
1411 \end{itemize}
1412 \end{enumerate}
1413Finally, single dummy variables corresponding to the different level of MPI indicators were generated and their frequencies were computed. 
1414%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1415Moreover, when working to create maps, you need to switch between two file locations:
1416\begin{enumerate}
1417\item Where the MPI datasets are stored
1418\item Where the downloaded shape files are stored\\
1419\end{enumerate}
1420The easiest way to do this is to use globals, as I've shown below (just copy in your 
1421file locations instead of mine). On top of that, changing the directory location to where the map data was downloaded is extremely important exercise. 
1422Transform the shape data from files in the map data folder into:
1423\begin{itemize}
1424\item a data file called $eth_data.dta$
1425\item a coordinates file called $eth_coord.dta$
1426
1427\end{itemize}
1428Note that the variable name in the genid() bracket must be the same as the region variable in the MPI dataset, in order to merge them together.\\
1429Critical considerations guided by multifaceted factors were made while developing indexes of dimensions, indicators, their respective weights and their deprivation cutoffs.  Among other things, the following factors were the basis were examined and reviewed.
1430\begin{itemize}
1431\item the global MPI, the dimensions, indicators, poverty cutoffs and weights
1432\item the Ethiopian context (values, culture, way of living and so on) and multiple settings and matters influencing poverty
1433\item the availability of the data in the panel database; and
1434\item the appropriateness, adequacy, expediency, seemliness and robustness of these data items after data filtering, data exploration and consultation. 
1435\end{itemize}
1436As a consequences of these factors, the key domains, indicators, weights and poverty cutoffs were established (for details see page 165-170 of this document).
1437  \subsection{Research Questions and Hypothesis}
1438This subsection briefly discuss the core questions that we wanted to investigate and answer in this study. Besides, based on the what we have observed and reviewed during the last steps of this study; this part will help us to develop the research problems.  At this stage, we scrutinized the core points of all our research endeavors by explaining what expected findings will be and we provided tentative  answers to the research questions that guide this study.
1439 \subsubsection{Research Questions}    
1440The key questions that motivated our study are fundamentally
1441whether  Ethiopia's official Multidimensional Poverty Index  (MPI) profile can be determined and identified  at national/federal level, regional level and can be further  decomposed by subgroups?
1442Whether comparisons of overlapping deprivations and multidimensional poverty levels be made among the four  major regional  states (namely, Amhara, Oromia, SNNP and Tigrai) of Ethiopia? Do MPI indexes provide insightful information that cannot be observed in the one-dimensional monetary poverty measures when decomposed and broken down  by region, gender, area, age group? What is the contribution of each MPI indicators and domains? How do we map the spatial distribution of MPI that enables policy makers for effectively targeting the poor? Why are the MPI indexes higher/lower than the orthodox one-dimensional monetary measuremnt of poverty?  \\
1443 
1444What is the status of multidimensional poverty index profile as compared to official reports based on uni-dimensional poverty measurement in Ethiopia? Whether multidimensional poverty is almost equal/similar across regions as claimed by the official government reports or is there a significant difference among the four big regions? Does multidimensional poverty pass the robustness test when one compares MPI across regions? Can MPI comparison be made among  those regions? And, does multidimensional poverty measures allow us to see how many households are experiencing deprivations at the same time? 
1445
1446\subsubsection{Research Hypothesis}
1447By and large, with all the imperfection and draw back in the MPI, it is the better indicator as it helps for targeting the poor in policy design and implementation with the existing data at our disposal. In essence, MPI enables policy makers  in developing countries for better policy design, implementations for better resources allocation and coordination; subgroup individual and geographic targeting, targeting precision, and monitoring and evaluation of development packages, programs and projects.\\
1448The purpose of the research is to analyses and discuss on the following core points:
1449
1450\begin{enumerate}
1451\item In certain circumstances when we deal with heterodox/multidimensional deprivations and  poverty in the living standards, socioeconomic and public services, tracking dynamics of poverty, assessing and translating the success of  antipoverty interventions in to multidimensional way, and dealing with social sanctions, violence and exclusions...the use of Alkire-Foster multidimensional deprivations and poverty measurement provides new insights and spotlights that were not adequately covered in the orthodox one dimensional monetary poverty measurement approach. 
1452\item To Measure MPI using the LSMS-IAS panel data set that has almost similar living standard indicators/ variables as in the case of one dimensional poverty measurement 
1453\item To measure poverty  employing the Alkire-Foster multidimensional approached at a higher cut-offs and compare it to the one dimensional poverty measurements reported earlier (for instance: in the World Bank, 2015 and Official Federal Government)in order to verify the prior claims thereof.
1454\end{enumerate} 
1455
1456Considering the predictive and testable criteria, we have designed our general research hypothesis as follows:\\
1457In many circumstances, the use of the Alkire-Foster multidimensional measurement of intertwined deprivations and poverty  provides extremely important and insightful information to the analysis of the situation of the poor; is higher than the official conventional/one-dimensional monetary poverty measurement; MPI can be decomposed and broken down by subgroups and indicators and comparison can be made among regional states and other sub groups. \\
1458
1459Besides, the study is based on the hypothesis that multidimensional poverty in Ethiopia are higher than the official government reports; MPI can be broken down by dimensions and decomposed by subgroups; comparisons cab be made among different regions, areas, age group etc and it remains robust. 
1460%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1461
1462 \subsection{Methodology Introduction}
1463
1464The overall achievement of each person $x_i$ is obtained by meaningfully and logically combining \textit{d} dimension. In short,  this is also referred as resources variable or welfare indicator variable. For instance, when the space is a set of income  sources, total income of a person $\textit{i}$, \begin{equation}
1465 x_i=\sum_{j}x_{ij}=x_{i1} \cdots x_{id} 
1466 \end{equation}
1467%This is a very critical condition 
1468Equally important, when the space is a set of commodities consumed, there is a set of \textit{d}  commodity prices designated by $(p_1 \cdots p_d)$. and total consumption expenditure of person \textit{i} is given by 
1469\begin{equation}
1470x_i=\sum_{i}p_{j}x_{ij}=p_{1}x_{i1} + \cdots + p_d x_{id}
1471\end{equation}
1472The overall achievement vector of a person/household is designated by $x=(x_1 \cdots x_n)$. To demonstrate, imagine that there are five persons in a society with over all income $\$25$, $\$18$, $\$15$, $\$33$, $\$31$. In this case $x=(25, 18, 15, 33, 31)$ is a vector indicating the income of the society. 
1473As can be seen, ordering/ranking the overall achievement vector; in other words, ranking/ordering individuals or households by  their  achievements yields (15, 18, 25, 31, 33).\\
1474 What is the cumulative distribution function for X? The cumulative density function that we apply in the robustness test of dominance can be plotted as follows:\\
1475\begin{center}
1476\begin{tikzpicture}
1477\begin{axis}[
1478    clip=false,
1479    jump mark left,
1480    ymin=0,ymax=3.5,
1481    xmin=14,xmax=35,
1482    xlabel={income},
1483    ylabel={cumulative distribution},
1484    every axis plot/.style={very thick},
1485    discontinuous,
1486    table/create on use/cumulative distribution/.style={
1487        create col/expr={\pgfmathaccuma + \thisrow{f(x)}}   
1488    }
1489]
1490\addplot [red] table [y=cumulative distribution]{
1491P(x) f(x)
149214   0
149315  1/5
149418  2/5
149525 3/5
149631 4/5
149733 1
149835 0
1499};
1500\end{axis}
1501\end{tikzpicture}
1502\end{center}
1503
1504The Cumulative Distribution Function (CDF) portrays that the share of the population having income less than a particular income level. This in turn will be an important tool for the robustness test of MPI domains, indicators and measures. 
1505
1506When employing MPI, policy makers are generally interested in aspects of a distribution or a vector that focuses on the Base (Poverty) deprivations where  welfare of the population below a certain level of income and socioeconomic achievement matrix is analyzed and diagnosed.\\
1507
1508According to Sen  (1976), Uni-dimensional poverty measurement involves two steps Identification and  Aggregation in detail discussed  below. \\
1509Here the purpose is to properly account, identify and exhaustively be able to answer the question who is poor?  This step dichotomies the population into a group of 
1510\textit{poor} and a group of \textit{non-poor} persons\\
1511To begin with, let's shed light on the main tool for multidimensional poverty analysis which is the basis for the methodology and methods of entire study : \textit{The Poverty Line} \textit{(z)} employing this approach it can be expressed that Person \textit{i} is poor if $x_i< {z}$
1512and is non-poor if $x_i ≥ \textit{z}$\\
1513where $x_i$ is the $i^{th}$element of vector x.\\
1514
1515The Poverty Line \textit{(z)} is extremely significance as it enables policy makers to identify a group/segment of poor people. Moreover, the poverty line is employed as a benchmark and the underpinning objective of a policy maker is to bring/lift up the deprived and the poor to \textit{z}. The vivid shortcoming is though important for poverty analysis, achievements of the non-poor above the poverty line is ignored.
1516
1517Besides, having the poverty line $\textit{z}$ helps us to execute censoring at the poverty line and it allows us to  generate a censored distribution of $\textit{x}$, referred as $x^{*}$, where 
1518
1519$x^{*}_{i} = {x_i}$ if  ${x_i} < {z}$ and \\
1520$x^{*}_{i} = {z}$  if  ${x_i} ≥ {z}$\\
1521
1522Another critical step is aggregation, the key focus here is to present the various alternatives of poverty measures and answer the question "how poor is the society?"
1523This step constructs an index of poverty summarizing the information in the censored achievement vector $x^{*}$ that implies for each distribution \textit{x} and poverty line \textit{z}, \textit{P(x;z)} or ${P(x^{*})}$ measures the level and extent of poverty dispersion in the society. \\
1524
1525Furthermore, in the aggregation approach we study behaviors of poverty measure. In other words,  we attempt to explain "how should a a poverty measure change owing to various data transformations?"
1526By and large, we shed light on the different attributes of poverty measure that operate under the properties that vary symmetrical with our intuitive considerations and those that do not obey the properties and may fluctuate unpredictably. Thus, synergistic analysis of poverty measure must made after explicitly filtering and double checking the underpinning properties that each poverty  measure fulfills. \\
1527
1528Equally important, since the fundamental axioms affect incentives while policy makers envisage to decrease, for instance headcount poverty; it is extremely important to digest the basic properties pertinent to each and every poverty measure. 
1529\subsection{Properties for Multidimensional Poverty Measures}
1530According to Alkire et al. (2015), policy maker pay great attention to how a poverty measure ought to behave in different situations in selecting poverty resumes so as to produce prudent, reproducible and consistent poverty measures and support the policy goals. The natural questions that would follow obviously are: should the poverty measure increase or decrease if the achievement of a poor person rises whilst there is no change in the achievements of other people in that category? Should the measure of poverty in a more populous region/state with a larger number of poor people be higher than the poverty measure in a small region/state with a smaller number of, but proportionally more, poor people?\\
1531
1532A policymaker seeking to alleviate poverty and improve the living standards of the poor ought to have a good sense of the various normative principles that the chosen poverty measures manifests. In this study, the set of properties for multidimensional study was based on the AF (2011a), that satisfy  the properties discussed in the next paragraphs. 
1533\subsubsection{Invariance Property}
1534
1535The invariance property requires that a poverty measure should not change under certain transformation of the achievement matrix. This, in turn, includes symmetry, replication, invariance, and scale variance. Besides, the two focus properties that fall in this category are  property focus and deprivation focus properties. 
1536
1537\begin{itemize}
1538\item Symmetry:\\
1539According to Alkire et al. (2015), symmetry property (sometimes referred to as anonymity property) requires that each person in a society is treated anonymously so that only deprivation matters and not the identity of the person who is deprived.  The underline message here is that as long as the deprivation profile of the society remains unchanged, swapping achievement vectors across people should not change the over all status of poverty. This type of arrangement may occur by pre-multiplying the achievement matrix by a permutation of appropriate order (Alkire et. al, 2015: P-53).\\
1540If an achievement matrix \textit{X'}  is obtained from achievement matrix \textit{X}  by as   $X'={\Pi}X$ where ${\Pi}$ is a permutation matrix of appropriate order, then 
1541\begin{equation}
1542{\Pi}(X';z) = {\Pi}(X;z).
1543\end{equation}
1544
1545\item Replication invariance:\\
1546The second invriance principle, replication invariance (also known as principle of population) requires that the level of poverty in a society to be standardized by its population size so that societies with different population size are comparable to each other (Alkire et al, 2015:p-53).  
1547If an achievement matrix  \textit{X'} is obtained from an other achievement matrix  \textit{X} by a replicating \textit{X} a finite number of times, then
1548\begin{equation}
1549\textit{P(X';z)} = \textit{P(X ;z)}.
1550\end{equation}
1551  \item Scale invariance\\
1552 
1553The third invariance principle, the scale invariance, demands that assessment of poverty ought not be affected by only changing the the scale of the indicators. To illuminate this notion, if the indicator the duration of completing schooling; then deprivation education and poverty should remain the same regardless of whether duration is guaged in months or year give that the deprivation cutoff is correspondingly adjusted.  (Alkire te. al., 2015: p-54)\\
1554If an achievement matrix \textit{X'} is obtained by post multiplying the achievement matrix \textit{X} by a diagonal matrix $\Lambda$ such that
1555$\textit{X'}=\textit{X}{\Lambda}$ and the deprivation cutoff vector \textit{z'} is obtained from \textit{z} such that $\textit{z'}= {\textit{z}{\Lambda}}$, then 
1556\begin{equation}
1557P(X';z')=P(X;z).
1558\end{equation}
1559A corollary of the scale in variance is the unit consistency which is briefly explained below;\\
1560
1561Unit Consistency:\\
1562For two achievement matrices \textit{X} and \textit{X"} and two deprivation cutoff vectors \textit{z} and \textit{z"}, if $P(X";z") < P(X;z),$ then
1563\begin{equation}
1564P(X"{\Lambda};z"{\Lambda}<P(X{\Lambda};z{\Lambda}).
1565\end{equation}
1566
1567\item Focus:\\
1568According to Alkire et al.(2015), the fourth invariance principle is focus. The key difference between the measurement of welfare and inequality and the measurement of multidimensional poverty is that while the former (welfare and inequality) are concerned with entire distribution in the sample space of the entire population of interest; multidimensional poverty measurement assessment deal the bottom or base of the of the distribution. \\
1569The  focus principle is vital since it requires poverty merely to respond to the achievements of the poor. In other words, this principle requires that poverty should not  change is there is an improvement in any achievement of a non-poor person. 
1570\item Poverty focus:\\
1571If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} such that $x'_{ij}>x_{ij}$ for some pair 
1572$(i,j)=(i',j')$ where $i'{\in}Z$, and $x'_{ij}=x_{ij}$ for every other pair $(i,j){\neq}(i',j'$ then,
1573\begin{equation}
1574P(X';z)=P(X;z).
1575\end{equation}
1576Note that the terms 'deprived' and 'poor' cannot be interchangeably used in the multidimensional framework of poverty measurement and analysis. This is so as someone can be poor yet not deprived in every singe indicator.\\
1577The average analysis of the uni-dimensional monetary poverty measurement is usually preferred over multidimensional measurements in policymakers of development countries like Ethiopia as this could  encourages these policymakers who are enthusiastically motivated by reducing the poverty figures to assist the poor to become non-poor in their non-deprived dimensions in place of addressing the dimensions in which they are deprived. \\
1578
1579Therefore, a second focus principle which is peculiar to the MPI, the deprivation focus is developed.
1580\item Deprivation Focus Principle:\\
1581 The deprivation focus principle requires that over all poverty does not change if there is an increase in the achievements of the non-deprived dimensions, notwithstanding whether it belongs to a poor or a non-poor person .\\
1582It is formally stated as follows: if an achievement matrix \textit{X'}  is obtained from another achievement matrix \textit{X} such that $x'_{ij}=x_{ij}$ for every other pair $(i,j){\neq}(i'j'),$ then
1583\begin{equation}
1584P(X';z)=P(X;z).
1585\end{equation}
1586When a union criteria is used to identify the poor, the deprivation focus principle implies the poverty focus principle. Whereas when an intersection criterion is employed to identify the poor, the poverty focus principle implies the deprivation focus principle (Alkire et al., 2015:p-56)
1587\subsubsection{Dominance Properties}
1588As discussed in (Alkire et al., 2015:p-57), the dominance property has two-the weaker version and the stronger versions. The stronger version require that a poverty measure strictly moves in a particular direction, given certain transformations in the achievements of the poor. However, the weaker version does not require a poverty measure to move in a particular direction but ensures the poverty measure does not move in an opposite (wrong) direction under certain transformation of the achievements.  
1589
1590The first dominance principle, monotonicity, requires that is the achievement of the poor person in a deprived dimension increases while other dimensions remain unchanged; then overall poverty should decrease. Informatively speaking, this principle considers that improvement in deprived achievements of the poor are good and shroud be reflected by resulting in poverty reduction. The weaker version of monotonicity ensures that poverty should not increase if there is an increase in any person's  achievement in the society. 
1591\item Stronger version of Monotonicity:\\
1592 If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} such  such that $x_{ij}<min{x'_{ij}, z_i}$ for some pair $(i,j)=(j',j')$ where $i'{\in}Z,$ and $ x'_{ij}= x_{}ij$ for every other pair $(i,j){\neq}(i',j'),$ then
1593 \begin{equation}
1594 P(X';z)<P(x;z).
1595 \end{equation}
1596 \item Weaker Version of Monotonicity\\
1597  
1598 If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} such  such that $x_{ij}<min{x'_{ij}, z_i}$ for some pair $(i,j)=(j',j')$ where $i'{\in}Z,$ and $ x'_{ij}> x_{}ij$ for every other pair $(i,j){\neq}(i',j'),$ then
1599 \begin{equation}
1600 P(X';z){\leq} P(x;z).
1601 \end{equation}
1602It ought to be noted that the monotinicity and weak monotonicity concepts are utilized in a similar manner as in the case of the one dimensional poverty analysis approaches. 
1603Nevertheless, the dimensional monotonicity, like wise to the deprivation focus principle is peculiar to the multidimensional settings of poverty measurement and analysis. According to Alkire and Foster (2011a), dimensional menotonicity requires that if a poor person who is not deprived in all dimensions, become deprived in an additional dimension then poverty should increase.  
1604\begin{quotation}
1605The dimensional monotonicity dimension ensure that we are not only concerned with the number of the poor in the society but also with the extent to which the poor are deprived in multiple dimensions-what we call the intensity of their deprivation (Alkire et al., 2015: p-58).
1606\end{quotation} 
1607It is presumed that a measure that satisfies monotonicity also satisfies dimensional monotonicity. \\
1608\item Dimensional Monotonicity\\
1609
1610If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} such  such that $x'_{ij}<z_j{\leq}x_{ij}$ for some pair $(i,j)=(i',j')$ where $i'{\in}Z$ and $ x'_{ij} = x_{}ij$ for every other pair $(i,j){\neq} (i',j',$ then
1611 \begin{equation}
1612 P(X';z) > P(X;z)
1613 \end{equation}
1614 
1615  \item Weak Dimensional Monotonicity\\
1616 
1617If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} such  such that $x'_{ij}<z_j{\leq}x_{ij}$ for some pair $(i,j)=(i',j')$ where $i'{\in}Z$ and $ x'_{ij} = x_{}ij$ for every other pair $(i,j){\neq} (i',j',$ then
1618 \begin{equation}
1619 P(X';z) {\geq} P(X;z)
1620 \end{equation}
1621 
1622\item Transfer principle:\\
1623
1624The third principle in the category of dominance principles, transfer, is concerned with the inequality among the poor. According Alkire et al.(2015: p-59), this principle is borrowed from the multidimensional inequality measurement literature that governs how a consistent poverty measure should behave when the distribution of measurements among the poor becomes more or less equal; while average achievements remains the same.\\
1625To put it in shortcut mathematical expressions: 
1626  
1627If  an achievement matrix \textit{X'} is obtained from \textit{X} such that $X'={\beta}X$ where ${\beta}$ is not a permutation or an identity matrix and ${\beta}_{ii}=1$ for all $i{\in}Z,$ then 
1628\begin{equation}
1629P(X';z)<P(X;z
1630\end{equation}
1631
1632\item Weak Transfer:\\
1633If  an achievement matrix \textit{X'} is obtained from \textit{X} such that $X'={\beta}X$ where ${\beta}$ is not a permutation or an identity matrix and ${\beta}_{ii}=1$ for all $i{\in}Z,$ then 
1634\begin{equation}
1635P(X';z){\leq}P(X;z
1636\end{equation}
1637
1638A great deal of literature we reviewed contend that the transfer principle in the multidimensional context that deals the spread of the distribution, is similar to its uni-dimensional counterpart. The key question that must be explored here is that should poverty go up or reduce owing to an association decreasing rearrangements among the poor? \\
1639
1640According to Tsui (2002), poverty should go down or atleast not increase since the association decrease rearrangement is likely to reduce inequality among the poor. On the contrary, Bourguignon and Chakravarty (2003) asserted that the change in the over all poverty should be conditional to the correlation between the  dimensions, that is whether they are substitutes or complements. When dimensions are believed to be substitutes, poverty should  not increase under the association-decreasing rearrangements. The further argued that that if dimensions are substitutes, an association-decreasing rearrangement helps both people compensate for their skimpy achievements in some dimensions with higher achievements in some other, a potential that was llimited for one of them before the rearrangement. \\
1641
1642On the other hand, when indicators are presumed to be complements, poverty should not decrease under the above explicated scenario. 
1643The gist is that the association-decreasing rearrangements has reduced the capacity of one of the persons to combine achievements and reach a certain level of well-being.
1644\item Weak Rearrangement (Substitutes):\\
1645If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} by an association-decreasing rearrangement among the poor, then 
1646\begin{equation}
1647P(X';z){\leq} P(X;z)
1648\end{equation}
1649
1650\item Converse Weak Rearrangement (Complements):\\
1651If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} by an association-decreasing rearrangement among the poor, then 
1652\begin{equation}
1653P(X';z){\geq} P(X;z)
1654\end{equation}
1655
1656\item Strong Rearrangement (Substitutes):\\
1657If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} by an association-decreasing rearrangement among the poor, then 
1658\begin{equation}
1659P(X';z){<} P(X;z)
1660\end{equation}
1661\item Converse Strong Rearrangement (Complements):\\
1662If an achievement matrix \textit{X'} is obtained from another achievement matrix \textit{X} by an association-decreasing rearrangement among the poor, then 
1663\begin{equation}
1664P(X';z){>} P(X;z)
1665\end{equation}
1666
1667To conclude, let's shed light on how are deprivation rearrangement properties related or different from the rearrangement properties? Generally speaking, if a poverty measure satisfies the (converse) weak deprivation rearrangements property, then the poverty measure will satisfy the (converse) weak rearrangement property,  and the converse is true as well. Again, a poverty measure that satisfies the (converse) strong deprivation rearrangement property surely satisfies the (converse) strong rearrangement property.\\
1668 
1669However, a poverty measure that satisfies the (converse) strong rearrangement property does not necessarily satisfy the (converse) strong deprivation rearrangement property. This entails, the main difference between the two sets of the properties lies in their stronger versions (Alkire et al., 2015:p-65).
1670\subsubsection{Subgroup Properties}
1671Subgroup properties deal with the interconnection between the overall (national) poverty  status and poverty in different subgroups of the population, and the intertwined between over all poverty and dimensional deprivations. The first principle of subgroup properties is subgroup consistency that ensures the change in overall poverty (national in our case) is consistent with the subgroup property.\\
1672
1673Suppose an achievement matrix X is divided into $m{\geq}2$ subgroups, such that$X^l$ and $n^l$ stand for, the achievement matrix and the population size of subgroup \textit{l}, for all $l=1,...,m$ and the subgroups are mutually exclusive and collectively exclusive: ${\sum}^m_{l=1}n^l=n.$
1674\item Subgroup Consistency:\\
1675If an achievement matrix \textit{X'} is obtained from the achievement matrix \textit{X} such that $p(X'^{l'};z)<P(X'^l;z)$ but $P(X'^l;z)=P(X^l;z)$ for all $l{\neq}l',$ and total population, as well as subgroup population, remain constant, then
1676\begin{equation}
1677P(X';z){<}P(X;z)
1678\end{equation} 
1679 \item Population Subgroup Decomposability:\\
1680 \begin{equation}
1681 P(X;z)={\sum}^m_{l=1}\left(\dfrac{n^l}{n}\right)P(X^l;z)
1682 \end{equation}
1683 The population subgroup decomposability principle is the pillar and extremely vital principle for targeting the deprived and the poor and for monitoring, evaluating and analyzing the well-being effect of antipoverty social programs. 
1684 \begin{quotation}
1685The population subgroup decomposability  property has been one of the most attractive properties for policy analysis as it can be particularly useful for targeting and monitoring progress in differential subgroups. It is worth noting that a poverty measure that satisfies population subgroup decomposability necessarily satisfies population subgroup consistency. But then, the converse is not true; in other words, population subgroup consistency does not necessarily imply population subgroup decomposability (Alkire et al., 2015:p-68).
1686 \end{quotation}
1687
1688Another crucial aspect of decomposition that of formidable relevance in the policy analysis of multidimensional poverty attribute is the breaking down  privation by deprivation across dimensions among the poor.\\
1689
1690The dimensional breakdown of poverty requires the overall poverty to be equal to a weighted sum of the dimensional deprivations after identification.
1691\item Dimensional Breakdown:\\
1692For an $n{\times}d$ dimensional achievement matrix \textit{X}, 
1693\begin{equation}
1694P(X;z)={\sum}^d_{j=1}w_jP_j(x._j;z)
1695\end{equation}, where $w_j$ the weight attached to dimension \textit{j} and $P_j(x._j;z)$ is the dimensional deprivation index after identification in dimension \textit{j}.
1696\subsubsection{Technical Properties}
1697Technical Properties warrants that measures behave within certain usual, convenient parameters. Moreover, to ensure that the poverty measures are meaningful, the non-triviality, normalization and continuity principles are expounded conversed below:\\
1698
1699The non-triviality principle requires that a poverty measure takes at least two different values. Unless a poverty measure takes two different values, it is stiff to distinguish a society with poverty (being poor) and a society without (non-poor) poverty. The normalization procedure/principle/ simply assets that the values of a poverty measure lie within the $0-1$ range. The minimum values zero (0), meaning there is no poverty in the society or an individual/household is non-poor; while the maximum value one (1), meaning poverty is at its maximum or the whole society /individual/household is poor. The continuity property prevents  a poverty measure from changing precipitously, given marginal changes in achievements. \\ 
1700To briefly revisit the key features that are vital to the MPI approach:
1701\item Normalization:\\
1702A poverty measure ought to be bounded between 0 and 1 for normalization purpose. 0 designates for non-poor/no poverty in the matrix of achievements and indicators while and 1 stands for highest possible poverty that a person, a group, a society or a population can possibly experience in their life time.
1703\item  Continuity\\
1704So as to prevent any sudden fluctuations in a poverty measure, it should be continuous on the achievements.
1705\item Invariance Properties\\
1706It asserts that poverty measures should not change under certain transformations of the achievement matrix.
1707\item Dominance Properties\\
1708It ensure that poverty measures should vary (increase or decrease) as a result of 
1709certain transformations in the achievement matrix
1710\item Subgroup Properties\\
1711It relates overall poverty to either groups of people or groups of domains. 
1712\end{itemize} 
1713In summary, when justifying why we employed  MPI as a better reflection of the multifaceted  interrelated and overlapping deprivations, poverty setting and maters, is advantage as   compared with the uni-dimensional conventional monetary measurement of poverty and its status among other multidimensional approaches, the unique features this the AF approach are succinctly presented and discussed under the principles of the invariance properties, particularly the principle of  deprivation focus; under the dominance properties-the principle of dimensional  monotonicity and weak dimensional monotonocity, weak deprivation rearrangement (substitutes), converse weak derivation rearrangement (complements), strong deprivation rearrangement(substitutes) and converse strong deprivation rearrangement (complements); under subgroup properties-dimensional breakdowns are some of the major peculiar features  specific to  MPI that make it superior and more attractive approach among other kinds of multidimensional  poverty measurement approaches.\\
1714
1715In the traditional monetary measures of poverty as well subjective perceptions have been employed  to measure poverty. Such measures of poverty stem from fundamental questions and perceived situations such as, "Do you have enough?"; "Do you consider your income to be very low, rather low, sufficient, rather high or high?". Besides, it may emanate from a judgment about minimum standards and needs like " what is the minimum necessary for a family of two adults and three children to get by?"; "what is the minimum necessary for your family?". Moreover, it may emanate from the poverty ranking from the community, like "which group are most  vulnerable in the village?" and so on. Based on the responses given for those questions, poverty measurement can be subjectively derived based on logical reasoning and judgment. Therefor, the normative motivation has been vital in the conventional uni-dimensional approaches of poverty measurement as well. The bottom line is subjective poverty measures not only may be employed to evaluate the situation of  particular household but also be set and inform the choices of poverty lines. 
1716%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1717
1718\subsection{Conceptual Framework}
1719
1720\subsubsection{Normative Motivation}
1721Well-being ought to be contextualized assessed and measured, not simply on the basis of resources but then, by the functioning and capabilities people relish. According to Sen (1992), Capability is thus a set of vectors of functioning, reflecting the person's freedom to lead one type of life or another 
1722\begin{quote}
1723A person's functioning  and her capability are closely related but distinct. Functionings, beings and doings that people values and achieve whereas capabilities, the various combinations oof functionings...that the person can achieve. 
1724\end{quote}
1725Looking at this assertion, a functioning could be better understood  as an achievement; whilst  s capability is the ability to achieve. 
1726The basis and focus of the MPI studies are Amartya Sen's capability approach that has been key in precipitating a fundamental reconsideration of the concepts of deprivations and key argument here is  that  well-being should be defined and studied, not merely by resources, but by the functionings and capabilities people enjoy. 
1727\begin{quotation}
1728Human lives are battered and diminished in all kinds of different ways, and the first task...is to acknowledge that deprivations of very different kinds have to be accommodated with a general overarching framework-Sen(2000)  as cited in Alkire (2016).
1729\end{quotation}
1730\subsubsection{Empirical Motivation}
1731
1732Empirical evidences show that monetary poverty measurement is not indicator of multidimensional deprivations as the levels and changes in monetary poverty and changes in settings of deprivations do not match.\\
1733
1734While trying to answer the empirical question whether conventional monetary approach is adequate to address all other settings of deprivations, wealth of literature revealed that there is a mismatch between monetary poverty level and other settings of deprivations and there is  lack of association between reduction in monetary poverty and reduction in deprivations in other indicators  (Ruggeri Laderchi (1997); Klasen (2008); Whelan et al. (2004); Bradshaw and Finch (2003); Wolff and De-Shalit (2007); Nolan and Whelan (2011); Global Monitoring Report 2013 (World Bank 2013)) \\
1735
1736Furthermore, empirical motivation argue that part of a lager problem is its lack of inclusiveness of economic growth in reducing non-monetary deprivations. It is not only that the new income generated by economic growth has been very unequally shared, but also that the resources newly created have not been utilized adequately to relieve the gigantic deprivations of the underdogs of society, India’s Uncertain Glory, Drèze and Sen (2013) as mention in (Foster, 2016). Finally, lack of association between the reduction in the conventional monetary poverty and reduction in deprivation in other indicators. Thus, there exists a solid ground  to recognize and expand deliberations to non-monetary deprivations.   \\
1737\subsubsection{Policy Motivation}
1738
1739How poverty is measured can influence policy differently. Monetary poverty measures recommend the employment of policies to impact income/consumption, growth policies,  that are budget lagged and redistribution of welfare and income transfers.  The key question that must be answered while addressing the synergy of poverty measurements and their implication to policy design in poverty targeting is that how could this consideration be expressed? Dimension by dimension, as in the dashboard approach? Or as part of a coherent notion of poverty? Is it enough to look at deprivations separately? Or is there a value added of looking at them together or jointly? Can we have same marginal distributions and very different joint distribution? While discussing such challenges, the under pinning understanding is that not only just the marginals but also the joint distribution of deprivations matters. It is argued that having the same marginals, while different joint distribution imply very different situation in terms of poverty measurements and assessments (Foster, 2016). \\
1740
1741From the perspective of policy motivation, how poverty is measured can influence policy and developing antipoverty programs. Forerunners of  the uni-dimensional believe  that monetary poverty measures indicated the use of policies to impact income/consumption, growth stimulant  policies and income transfers. However, the growth policies depend on lagged values and income transfer could have momentary effect, literature revealed that the growth policies did not achieve the envisaged targets in poor countries and income transfer are not sustainable  and hamper productivity as huge resources is  earmarked that could have been invested in other portfolio to boost productivity and reduce other deprivations (Foster, 2016).\\
1742
1743Multidimensional measures suggest holistic, comprehensive and inclusive policies on many domains such as education, health, living conditions and so on. Hence, it provides a way of coordinating policies/programs and can be used as a basis of communication among stakeholders. The underline message is that while a good poverty measure alone cannot manufacture potent policy, it can be designed with that goal in mind (Foster, 2016).\\
1744
1745\subsubsection{Miscellaneous  Motivations}
1746
1747According to Alkire and Foster (2016), a good poverty measurement should be able to:
1748 \begin{itemize}
1749 \item Ethical: Enhancing the fit between the measure and the phenomenon.  
1750\item produce the official statistics of multidimensional poverty
1751\item identify overall patterns of deprivation
1752\item to be decomposed and compare sub-national groups, e.g. regions, urban/rural, or ethnic groups
1753 \item compare the composition of poverty in different regions or social groups
1754 report poverty trends over time
1755 \item monitor the changes in particular indicators
1756 \item evaluate the impact of programs on multiple outcomes
1757 \item target geographical regions or households for particular purposes
1758 \end{itemize}
1759 We believe that the MPI fulfills the above requirements and is a good poverty measurement so far.  
1760 
1761\subsection{Methodological Framework}     
1762
1763To begin with, let's out line the methodological framework by shedding light on the following key points. For comparability of poverty estimates across various populations, we conjecture \textit{d} to denote a fixed set of dimensions in the multidimensional set of poverty measurements. The achievements of all the individual and/or households in a given society can be represented by an $n{\times}d$ dimensional achievement matrix \textbf{X}. Furthermore, let's shed light on the following essential points:
1764  \begin{itemize}
1765   \item Achievement: performance of a person in a dimension\\
1766 - $x_{ij}$ Achievement of person  $i (=1,…,n)$ in dimension $j (=1,…, d)$
1767  \item Achievement matrix: Summarizes achievements  such as  health domain indicators(malnutrition, child mortality, etc), education dimension indicators (years of schooling, school age child attendance etc) , living standards dimension indicators ( access to improved electricity, access to improved sanitation, access to improved housing, access to improved communications and transportation,  assets ownership etc), and so on of all \textit{n} persons in \textit{d} dimensions
1768 \item Achievement vector of a Person: May contain incomes from \textit{d} different sources or \textit{d} different commodities consumed
1769\end{itemize}
1770\[
1771X=
1772\begin{blockarray}{c c c c}
1773\BAmulticolumn{4}{c}{\text{Dimensions=column}}\\
1774\begin{block}{[c c c c ]}
1775x_{11} & x_{12} & \cdots & x_{1d} \\
1776x_{21} & x_{22} & \cdots & x_{2d} \\
1777\vdots & \vdots & \ddots & \vdots \\
1778x_{n1} & x_{n2} & \cdots & x_{nd}\\
1779\end{block}
1780\end{blockarray}\quad\text{achievemnts of\\ persons=raw}
1781\]
1782
1783As can be seen from the above matrix notation,matrix \textit{X} summarizes the joint distribution of \textit{d}
1784dimensions across \textit{n} individuals. Where as, row vector $ x_{i.} = {(x_{i1},..., x_{id})}$ summarizes the achievements of  person \textit{i} in all \textit{d} dimensions.
1785Besides, column vector $ x_{.j} = {(x_{1j},..., x_{nj})} $ summarizes the achievements in dimension \textit{j}  of all \textit{n}  persons.
1786The overall achievements of each person  is give by $x_i$ that is obtained by meaningfully combining \textit{d} dimensions. It's also refereed as  renounces variable or welfare indicator. To consider some practical cases, when the space is  set of income sources; total income of person \textit{i} is given by 
1787\begin{equation}
1788x_i={\sum}_jx_{ij}=x_{i1}+...+x_{id}
1789\end{equation}
1790Besides, when the space is a set odd commodities consumed, there is a set of \textit{d} commodity prices $p_1,...,p_d$ and total consumption expenditure of person \textit{i} which is written as follows:
1791\begin{equation}
1792x_i={\sum}_jp_jx_{ij}=p_1x_{i1}+...+p_dx_{id}
1793\end{equation}
1794
1795The notion can be extended to the notation of multidimensional space by invoking two practical assumptions for convenience. One vital assumption is that assuming the achievements of person \textit{i} in dimension \textit{j} can be represented in by non-negative real numbers, such that $x_{ij}{\in}{\Re}_+$ for all $i=1,...,n$ and $j=1,...n$. Another essential presumption based on the more is better utility principle; we assume that higher achievements are proffered to lower ones.\\
1796
1797In general, in the multidimensional context, this can be designated that the set of all possible matrices of size $n{\times}d$ by ${\chi}_n{\in}{\Re}^{n{\times}d}_{\dagger}$ and the set of all possible achievement matrices by ${\chi},$ such that ${\chi}=U_n{\chi}_n,$ then matrix \textit{X} contains achievements for \textit{n} persons in \textit{d} dimensions. Assuming that $X{\in}{\chi}$, the achievements of any person \textit{i} in all \textit{d} dimensions are represented by the \textit{d}-dimensional vector \textit{$x_i$} for all $i=1,...,n$. The achievements of any dimension \textit{j} for all \textit{n} persons, which is column \textit{j} of matrix \textit{X}, are represented by \textit{n} dimensional vector \textit{$x._j$} for all $j=1,...d$ (Alkire et al, 2015: p-25).\\
1798
1799%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1800To demonstrate some practical examples, a representative yet typical hypothetical data set with five dimensions and five persons resembles the  achievement matrix below. 
1801%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1802$$X_{ij}=
1803%\begin{center}
1804\begin{blockarray}{rccccc}
1805 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} & \thead {Asset\\ Ownership}\\
1806\begin{block}{>{\small}r[ccccc]}
1807Person\textsubscript{1} & 900 & 10 & Yes & Yes & 7\\
1808Person\textsubscript{2} & 300 & 7  & No  & Yes & 3 \\
1809Person\textsubscript{3} & 400 & 9  & No  & No  &  2\\
1810Person\textsubscript{4} & 800 & 12  & Yes & Yes &  8 \\
1811Person\textsubscript{5} & 1000 & 13 & No  & No  &   4\\
1812\end{block}
1813\end{blockarray}
1814%\end{center}
1815$$
1816$$Z_j =
1817 \left[{\begin{array}{ccccc} \hfill 500  &  \hfill 11  &  \hfill Yes  &  \hfill Yes  &  \hfill 5 \end{array}} \right] cutoffs$$ 
1818
1819Where $Z_j>0$ be the deprivation cut-off in indicator \textit{j}.\\
1820\subsubsection{Construction of Deprivation Matrix}
1821Deprivation matrix can be constructed in order to be able to identify poor and non-poor persons. \\
1822
1823In multidimensional poverty analysis, each dimensions can be a weighted or deprivation value based on is relative importance or priority. Let's denote the relative weight attached to dimensions by \textit{j} by \textit{w}, such that$w_j>0$ for all $j=1,...,d$. The weight attached to all \textit{d} dimensions are collected in a vector $w=(w_1,...,w_d).$ For simplicity we can restrict the weight so that they sum to the total number of considered dimensions which is give by,${\sum}_jw_j=d$.
1824
1825In more general cases, the weights could be normalized and they sum to one ${\sum}_jw_j=1$.
1826
1827Replacing entries: 1 if deprived, 0 if not deprived; we produce the following deprivation matrix: 
1828$$g^0=
1829%\begin{center}
1830\begin{blockarray}{rccccc}
1831 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} & \thead {Asset\\ Ownership}\\
1832\begin{block}{>{\small}r[ccccc]}
1833Person\textsubscript{1} & 0 & 1  & 0  & 0  &  0\\
1834Person\textsubscript{2} & 1 & 1  & 1  & 0  &  1 \\
1835Person\textsubscript{3} & 1 & 1  & 1  & 1  &  1\\
1836Person\textsubscript{4} & 0 & 0  & 0  & 0  &  0 \\
1837Person\textsubscript{5} & 0 & 0  & 1  & 1  &  1\\
1838\end{block}
1839\end{blockarray}
1840%\end{center}
1841$$
1842\textit{Identification-Weights:}\\
1843By introducing equal weights symbolized by $w_1, w_2, w_3, w_4 and w_5$ for each dimension, we produce the following matrix and weights:
1844$$g^0=
1845%\begin{center}
1846\begin{blockarray}{rccccc}
1847 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} & \thead {Asset\\ Ownership}\\
1848\begin{block}{>{\small}r[ccccc]}
1849Person\textsubscript{1} & 0 & 1  & 0  & 0  &  0\\
1850Person\textsubscript{2} & 1 & 1  & 1  & 0  &  1 \\
1851Person\textsubscript{3} & 1 & 1  & 1  & 1  &  1\\
1852Person\textsubscript{4} & 0 & 0  & 0  & 0  &  0 \\
1853Person\textsubscript{5} & 0 & 0  & 1  & 1  &  1\\
1854\end{block}
1855\end{blockarray}
1856%\end{center}
1857$$
1858$$w =
1859 \left[{\begin{array}{ccccc} \hfill w_1  &  \hfill W_2  &  \hfill w_3 &  \hfill w_4  &  \hfill w_5 \end{array}} \right] weights$$
1860 
1861Where ${w_i}$ is the weights. The weighted deprivation matrix can be summarized as follows:\\
1862By multiplying each dimensions by their respective weights, the weighted deprivation matrix is produced below:
1863$$g^-0=
1864%\begin{center}
1865\begin{blockarray}{rccccc}
1866 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} & \thead {Asset\\ Ownership}\\
1867\begin{block}{>{\small}r[ccccc]}
1868Person\textsubscript{1} & 0 & w_2  & 0  & 0  &  0\\
1869Person\textsubscript{2} & w_1 & w_2  & w_3  & 0  &  w_5 \\
1870Person\textsubscript{3} & w_1 & w_2  & w_3  & w_4  &  w_5\\
1871Person\textsubscript{4} & 0 & 0  & 0  & 0  &  0 \\
1872Person\textsubscript{5} & 0 & 0  & w_3  & w_4  &  w_5\\
1873\end{block}
1874\end{blockarray}
1875%\end{center}
1876$$
1877Moving on, the next logical procedure is identifying the poor and non-poor; in other words counting deprivations. In concise, it can be presented as follows:
1878\textit{Identification: Counting Deprivations} \\
1879With the presumption that there exist equal weights and are give by
1880\begin{equation}
1881{\sum}^d_{j=1}w_j=d
1882\end{equation}
1883For simplicity, let's further assumes ($w_i=1$) in the example give foe illustration purpose. \\
1884
1885In some particular cases where weights are equal and the sum to the number of dimensions (as the case below), the score is simply the the number of deprivation counts that the person experiences. Whenever weights are unequal yet aggregated to the number of dimensions, person \textit{i's} deprivation score is defined as the sum of the weighted deprivation counts (Alkire et. al., 2015: p-31) \\
1886The Counting Deprivation Vector can be constructed from weighted deprivations matrix just by counting the number of poor and non-poor in a give society as it can be seen in the matrix below:
1887$$g^-0=
1888%\begin{center}
1889\begin{blockarray}{rccccc}
1890 & \thead{Income} & \thead{Years of \\Education} & \thead{Improved\\ Sanitation} & \thead{Access to\\ improved \\ Electricity} & \thead {Asset\\ Ownership} \\ 
1891\begin{block}{>{\small}r[ccccc]}
1892Person\textsubscript{1} & 0 & 1  & 0  & 0  &  0\\
1893Person\textsubscript{2} & 1 & 1  & 1  & 0  &  1 \\
1894Person\textsubscript{3} & 1 & 1  & 1  & 1  &  1\\
1895Person\textsubscript{4} & 0 & 0  & 0  & 0  &  0 \\
1896Person\textsubscript{5} & 0 & 0  & 1  & 1 &  1\\
1897\end{block}
1898\end{blockarray}
1899%\end{center}
1900\quad
1901\begin{matrix}
1902\thead {Counting\\ Deprivations\\
1903 C}\\
1904 
19051\\
19064\\
19075\\
19080\\
19093
1910\end{matrix}
1911$$\\
1912The fundamental question "who is poor?" and hence the key notion of identification in any poverty analysis can then be answered either using the union approach or intersection approach. 
1913\textit{Identification-Union Approach}:\\
1914In any setting of multidimensional poverty analysis, the identification of poor and non-poor in the society is a critical stage. There basically two approaches in the multidimensional literature, the union approach and the intersection approach. Employing the union approach, a society, an individual or a household is identified as poor if deprived in any dimensions $c_i{\geq}1$.\\
1915
1916The first critical step of poverty measurement exercise is that it requires an identification function, that determines whether a society, person or a household is to be determined poor or non-poor. \\
1917It was reported from previous literature that the union approach often predicts very high numbers (Bourguignon and Chakravarty, 2003). 
1918\textit{Identification-Intersection Approach}:\\
1919A society, an individual or a household is poor if deprived in all dimensions $c_i=d$. It is a demanding requirement and may not represent the condition of poverty on the ground specially for developing countries like Ethiopia since it often identifies a very narrow slice of the population.  
1920\textit{Identification-Dual Cutoff Approach}:\\
1921In a space of multidimensional poverty analysis, employing the technique of dual cutoff approach for identification   purpose is specific to the MPI and hence its advantage over all other alternative multidimensional poverty measurement approaches.\\
1922While using the  dual cutoff approach, applying our normative judgment, we fix \textit{k}, and identify as poor is $c_i{\geq}k.$
1923%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1924\subsubsection{Identifying Deprivations in the society/population}
1925We considered this step and identification related for a specific/limited case in the above discussions. By the same token, we shall consider the similar approached when applied to the society/population. To fix matters short, let's start by defining the required threshold in each dimension for a society to be considered as deprived or non-derived. Such threshold can equally be perceived as setting the deprivation cutoff for an indicator, namely, the minimum level of achievement considered to be normatively sufficient in order to be non-deprived in that indicator-which is referred to as the dimensional  deprivation cutoff.\\
1926
1927When someone's achievement is strictly  below that cutoff, s/he is considered to be deprive. A typical multidimensional poverty measurement is given by an $n{\times}d$ dimensional achievement matrix \textit{$X_{ij}$}, where $x_ij$ is the achievement of person \textit{i} in dimension \textit{j}. It's further assumed that achievements can be represented by non-negative real numbers (i.e. $x_{ij}{\in}{\Re}_+$  and that higher achievements are preferred to lower ones.  Let's say for each dimension \textit{j}, a threshold (deprivation cutoff) $z_j$  is defined as a minimum achievement required in order to be non-deprived. Deprivation cutoffs are garnered in the \textit{d}-dimensional vector $z=(z_1,...,z_d).$ \\
1928 
1929Given each person's achievement in each dimension $x_{ij},$ if the $i^{th}$ person's achievement level in a given dimension \textit{j} falls short of the respective deprivation cutoff $z_j,$ the person is said to be deprived in that dimension (that is, if $x_{ij}<z_j.$
1930If the person's level is at least as great as the deprivation cutoff, the person is not deprived in that dimension.\\
1931 
1932Given the achievement matrix \textit{X} and the vector of deprivation cutoffs \textit{z}, one can obtain a deprivation matrix $g^0$ such that $g^0_{ij}=1$ whenever $x_{ij} < z_j$ and $g^0_{ij}=0,$ otherwise, for all $j=1,...,d$ and for all $i=1,...,n.$\\
1933In other words, if person \textit{i} is deprived in dimension \textit{j}, then the person is assigned a deprivation status of \textit{1}, and \textit{0} otherwise. The matrix $g^0$ summarizes the deprivation status  value all people in all dimensions of matrix \textit{X}. The vector $g^0_i.$  summarizes the deprivation status values of person \textit{i} in all dimensions, and the vector $g^0_.j$ summarizes the deprivation status values of all persons in dimension \textit{j}.\\
1934
1935Notice that the deprivations in each of the \textit{d} dimensions do not necessary have similar relative importance and hence, the vector $w=(w_1,...,w_d)$ of weights or deprivation values is employed to indicate the relative importance of a deprivation in each dimension. As a matter of fact, the deprivation value attached to dimension \textit{j} is denoted by $w_j$ such that $w_j>0$ for all $j=1,...,j$. Correspondingly, the weights attached to all \textit{d} dimensions are collected in a vector $w=(w_1,...,w_d).$
1936
1937From the matrix $g^0$ discussed above, a vector of deprivation score can be constructed for each person \textit{i} so that 
1938\begin{equation}
1939c_i={\sum}^d_{i=1}w_jg^0_{ij}
1940\end{equation}
1941
1942In other words, $c_i$ stands for the sum of weighted sum of deprivations suffered by person \textit{i}. \\
1943According to Alkire and Foster 2011a, the deprivation score is compared to the poverty cutoff designated by \textit{k}, which is the minimum score a person must have to be considered poor. In this context, a person is considered poor is $c_i{\geq}k.$. The poverty cutoff can \textit{k} can range from the union to the intersection criterion. Where, the union criterion requires $k{\in}(0, min_j(w)]$ and identifies a person as poor if the person is deprived in any dimension. \\
1944
1945However, the intersection criterion requires $k={\sum}^d_{j=1}w_j$ and identifies a person as poor only if s/he is deprived in all dimensions under consideration. In-between these two extreme criteria there is a room for intermediate criteria.\\
1946In this study, we consider a person is poor if she/he is deprived in 40 or more $\%$ of the total deprivations possible in the society. It can be noticed that as k increases the proportion of the poor is going to decrease \footnote{ When we are focusing on the poor, we designated "A" is the average of deprivations among the poor. It deals with how poor they (the poor) are.\\
1947
1948Censored Headcount: proportion of people who are deprived and poor.
1949Percent Contribution: the percentage of contribution of a dimension adds up to the overall.}\\
1950
1951It should be noted that unless someone may experience some deprivations, nonetheless, not identified as poor unless the union criterion is being applied. In so doing, the deprivations of those who have been identified as poor are then aggregated  in order to obtain a poverty measure. 
1952
1953As it is discussed in (Alkire et. al., 2015:p-110), the aggregation achievement approach consists of applying aggregation functions $f_s,$ to the achievements across dimensions for each person so as to obtain an over all achievement values $f_s(x_i;w).$\\
1954
1955Correspondingly, similar functions can be applied to the dimensional deprivations cutoffs in order to obtain an g 
1956  
1957\subsection{Identification of the Multidimensionally Poor and Non-Poor}
1958In the multidimensional measuring setting,  where multiplicity of variables are involved, identification becomes a challenging and complex exercise. Many approaches can be applied for the identification process in the multidimensional poverty measurement   and analysis study. To fix matters short, following the approached  discussed by Alkire et al.(2015), we followed the censored achievement approach. In so doing, we first determined who is multidimensionally poor in each dimension by comparing the persons achievement against the corresponding deprivation cutoff and thus accounting only the deprived achievements  and disregarding achievements above deprivation cutoff through the process of censoring for the identification purpose of the poor. \\
1959
1960Accordingly, counting approach is the most prominent method in the AF methodology. The concise process here is that a counting approach first identifies whether a person is deprived or not in each dimensions and then designates a person as poor according to the count of deprivations s/he experiences. The benefit of the AF methodology is threefold. Firstly, it has formalized the counting procedure of identification into a dual-cutoff approach, expounding the requirements of two distinct sets of threshold to define poverty in the multidimensional  context. One is the set of deprivation cutoffs, that identify whether a person iis deprived in with respect to each dimension. Then, a single poverty cutoff delimitation how deeply and broadly a person must be deprived in order to be considered as multidimensionlly deprived and poor.\\
1961    
1962Secondly, as a consequences of employing a dual-cutoff approach, the AF methodology takes in to account the joint distribution of deprivations at the identification step and not just the at the aggregation step. This is peculiar to the AF methodology as all other non-counting approaches used the union criterion. \\
1963
1964Thirdly, the AF methodology has interspersed the counting approach to identification with an aggregation methodology that extends the one-dimensional the Foster-Greer-Thobecke (FGT) class of poverty measure overcoming the limitation of pitfalls of the headcount ratio that is used by most counting methods but allowing intuitive interpretations. Therefore, the AF methodology draws together the counting traditions that are well known for their      practicability and policy appeal and the widely applied FGT class of axiomatic measure so as to examine multidimensional poverty that stands on the shoulder of both traditions (Alkire et al, 2015: p-149)\\
1965
1966Multidimensional poverty analysis consists of several factors that constitute poor people's multiple deprivation like poor/lack of affordable health services, lack of education, insufficient living standards, lack of income, dis-empowerment, poor quality of work, threats from violence and so on. To comprehensively capture and analysis these  settings and matters of multiple deprivations, we will employ the Alkire-Foster (AF) method, developed by Sabina Alkire and James Foster(2011) at Oxford Poverty  and Human Development Initiative (OPHI) is a flexible measure of well-being or poverty.  The AF method offers numerous advantages as itis amenable to incorporate different dimensions and indicators to develop measures relevant to a particular context. \\
1967 
1968In the light of this, the AF method can be applied to develop national, regional, global measures of poverty or well-being by integrating domains and indicators that are modified to the existing context. Also, they can be used to monitor and evaluate the efficacy of anti-poverty programs over time. By the same token, they can also be employed to target individuals for the public service programmes or conditional cash transfers (CCTs) against set criteria. \\
1969 
1970Equally important, the AF methods can be used by NGOs, governments, agencies, donors, and privates sectors to develop multipurpose measures such as effective allocation of resources that crucial to investing resources that are deemed to be most effective at reducing poverty. Policy makers can single out which deprivations comprise poverty, and pinpointing the most common among and within the groups, and so that policies will be crafted to to address those particular needs. Further more, the AF method is useful for identifying interdependence among deprivations and assist to identify poverty traps. \\
1971
1972Moreover, the AF method offers further benefits in reflecting the effects of changes in policies that are aiming at improving social program, say enhancing access and quality of education can be assessed and monitored quickly. Like wise, different dimensions, indicators and cutoffs can be employed to develop measures pertinent to specific uses, scenarios and communities and hence they are very flexible. Again, can be used for complementing other metrics and measurements of poverty such as income. Optionally, they can incorporate  income as one dimension of the several domains in a multidimensional measures\\
1973
1974Key aspects of MPI, that is a recently developed measurement of multidimensional deprivations of well-being is generally a holistic approach that passes the basic axioms of poverty measures and involves the following three major steps
1975\begin{itemize}
1976\item  Choice of space: How should we define poverty?
1977\item Identification: Who is poor?
1978\item Aggregation: How should the information of all the poor be aggregated to obtain an index/indices?
1979\end{itemize}
1980
1981 \subsubsection{Twelve Common Steps to Develop a Multidimensional Poverty measure}
1982
1983According to Alkire et. al. (2015), the AF methodology can be presented in 12 intuitive steps as follows:
1984\begin{enumerate}
1985
1986\item  Choose Unit of Analysis: The unit of analysis is most commonly an individual or household but could also be a community, school, clinic, firm, district, or other unit.
1987In this study, the unit of identification is the household while the unit of analysis is the individual. 
1988
1989\item  Choose Dimensions: The choice of dimensions is important but less haphazard than people assume. In practice, most researchers implicitly draw on five means of selection, either alone or in combination:\\
1990
1991Ongoing deliberative participatory exercises that elicit the values and perspectives of stakeholders. A variation of this method is to use survey data on peoples perceived necessities.
1992A list that has achieved a degree of legitimacy through public consensus, such as the universal declaration of human rights, the SDGs, or similar lists at national and local levels.\\
1993
1994Implicit or explicit assumptions about what people do value or should value. At times these assumptions are the informed guesses of the researcher; in other situations they are drawn from convention, social or psychological theory, or philosophy.
1995Convenience or a convention that is taken to be authoritative or used because these are the only data available that have the required characteristics.\\
1996
1997Empirical evidence regarding people's values, data on consumer preferences and behaviors, or studies of what values are most conducive to people's mental health or social benefit.
1998Clearly these processes overlap and are often used in tandem empirically; for example, nearly all exercises need to consider data availability or data issues, and often participation, or at least consensus, is required to give the dimensions public legitimacy.
1999
2000\item  Choose Indicators: Indicators are chosen for each dimension on the principles of accuracy (using as many indicators as necessary so that analysis can properly guide policy) and parsimony (using as few indicators as possible to ensure ease of analysis for policy purposes and transparency). Statistical properties are often relevant example, when possible and reasonable, it is best to choose indicators that are not highly correlated.
2001
2002\item  Set Deprivation Cutoff: A deprivation cutoff is set for each indicator. This step establishes the first cutoff in the methodology. Every person can then be identified as deprived or non-deprived with respect to each indicator. 
2003\item  Apply Poverty Lines: This step replaces the person's achievement with his or her status with respect to each cut-offs.
2004
2005\item  Count the Number of Deprivations for Each Person: General weights can be applied, however, in which case the weighted sum is calculated depending on normative value judgments of the importance of the indicators.
2006
2007\item  Set the Second Cutoff: Assuming equal weights for simplicity, set a second identification cutoff, k, which gives the number of indicators in which a person must be deprived in order to be considered multidimensionally poor. In practice, it may be useful to calculate the measure for several values of k. Robustness checks can be performed across all values of k. \\
2008
2009To shed light more on the the poverty cutoff, it can be comprehended as the share of dimensions in which a person must  be deprived so as to be  reckoned as multidimensionally poor. In this study, it was set in between k[33.3-40$\%$]of he weighted dimensions. This normative decision was substantiated by  statistical criteria and analytical validations. Among the calibration and validation test, it includes calculating the poverty for all possible poverty cutoffs and systematically cross checking the robustness of the results to changes in these values. 
2010
2011\item  Apply Cutoff k to Obtain the Set of Poor Persons and Censor All Non-poor Data: The focus is now on the profile of the poor and the dimensions in which they are deprived. All information on the non-poor is replaced with zeros (0). 
2012\item  Calculate the Headcount, H: Divide the number of poor people by the total number of people. 
2013
2014\item  Calculate the Average Poverty Gap, A: A is the average number of deprivations a poor person suffers. It is calculated by adding up the proportion of total deprivations each person suffers. 
2015
2016\item  Calculate the Adjusted Headcount, M0: If the data are binary or ordinal, multidimensional poverty is measured by the adjusted headcount, M0, which is calculated as H times A. Headcount poverty is multiplied by the average number of dimensions in which all poor people are deprived to reflect the breadth of deprivations. 
2017
2018\item  Set Weights: Read OPHI’s Working Papers on weighting dimensions of well-being and materials from OPHI’s workshop on setting weights in multidimensional measures (Alkire, 2016).
2019\end{enumerate}
2020 
2021The multidimensional poverty index, which is based on the AF method, is a matrix x containing available data which is size ${N*D}$ where \textit{D}-stands for number/s of indicators across a population of \textit{N}households/individuals, and and describes
2022for each individual the achievement in each dimension deemed relevant. Respectively, $y_{id}\geq0$ represents the achievement of individual $ {i}=1,..., {N}$ in dimensions ${d}=1,...,{D}$. Let the  row vector \textit{z} with $z_i>0$,  describes the deprivation cutoffs in indicator \textit{j}, i.e., the achievements necessary for not being considered as deprived in the respective dimension, and $w_j$ be the weight of indicator \textit{j} such that the weight sum to one. That is, normalized weight $ (\sum^d_{i=1}w_j=1)$. \\
2023  
2024Next, we construct a matrix of deprivations $g^0=[g^0_{ij}]$, whose typical element $g^0_{ij}$ is defined by $g^0_{ij}=w_j$ when $y_{ij}<z$ and $g^0_{ij}=0$ when $y_{ij}\geq z $.\\
2025
2026Moving on, using this information presented above, we then construct  the deprivation vector \textit{c}  by counting individual deprivations, i.e., the column vector's elements are $c_i=\sum^D_1\perp(y_{id}<z_d)$. In other words, we construct a vector \textit{c} of deprivation counts, whose $j^{it}$ entry $c_i=\sum^d_{j=1}g^0_{ij}$ represents the sum of weighted deprivations suffered by a person \textit{i}. Second a person is considered poor if his or her weighted derivation count is greater than or equal to \textit{k}.\\
2027 
2028Let $\rho_k$ takes values of 1 when $c_i\geq k,$ and 0 when $c_i<k$. To deal with poor people alone, we censor the deprivations of persons who are deprived and non-poor by constructing a matrix $g^0(k),$ obtained from $g^0$ by replacing its $i^{th}$ row $g^0_i$ with a vector of zeros whenever $\rho_k=0$. This matrix contains the weighted deprivation of all persons who are identified as poor and exclude deprivations of non-poor.\\
2029
2030\subsubsection{Estimation Framework: Aggregation and Measurement of MPI Poor and Non-Poor of Ethiopian Households}
2031
2032If a person is deprived in at least one third of the weighted measurements of all the dimensions, then s/he is recognized as multidimensional poor (or simply ‘MPI poor'). The pillar estimation frameworks and aggregation employed in this study are: the percentage of the population that is multidimensionally poor is simply the incidence of poverty, or headcount ratio (H). In other words, \textit{H}: can be understood as the 'Multidimensional Deprivation Headcount'(the share of poor individuals in the population). \\
2033
2034The mean percentage of indicators in which poor people are deprived is portrayed as the severity of their poverty (A) or simply \textit{A}: The 'Average Multidimensional Poverty Intensity' (the average percentage of simultaneous deprivations suffered by the poor individuals). The \textit{MPI} is computed by proliferating the amount of poor by the mean severity of poverty crosswise the poor (Thus, $MPI = H {\times} A$); as a result, it shows both the share of people in poverty and the extent to which they are disadvantaged. By the same token,  $M_0$: The Adjusted Headcount Ratio, $M_0 = H*A$, which accounts for both the incidence of poor individuals and the intensity of their multiple deprivations.\\
2035
2036Correspondingly, the AF methodology of multidimensional poverty measurement generates a class of measures that extends the FGT class of measures in natural ways. In fact, one has to be familiar with a step-wise analytic and intuitive presentation of how to obtain the Adjusted Headcount Ratio ($M_0$ ), Adjusted Poverty Gap ($M_1$ ) and the Adjusted Squared Poverty Gap (or FGT) Measure ($M_2$) before proceeding with a more formal implementation of the AF methodology (Alkire, 2014).\\
2037 
2038Note that in all three key measures ($M_0, M_1 and M_2$ ) are the deprivations experienced by people who have not been branded as poor (i.e. those whose deprivation score is below the poverty cutoff) are censored, hence not included; this censoring of the deprivations of the non-poor is consistent with the property of 'poverty focus' which is analogous to the one-dimensional case - requires a poverty measure to be independent of the achievements of the non-poor.\\
2039
2040In this study, in view of the fact that we want to amend for deviation in household size to make sure that measurements take into account that the poorest households typically have more members; we applied a weight $w_i=s_ih_i$  where $h_i$ is the household size and $s_i$  is the sample weight. Note that $w_i$  could be regularized so that  
2041\begin{equation}
2042{\sum}^n_{i=1}w_i=n
2043\end{equation}
2044Moving on, the sum of the destitute is given by  
2045\begin{equation}
2046H=\frac{1}{n}\sum^d_{i=1}\rho^k_i
2047\end{equation}
2048 $\rightarrow$ mean of $rho\_'k'$\\
2049\begin{itemize}
2050 \item $rho\_‘k’$: One identifier of the poor for each poverty cutoff \textit{k}
2051\item $c0k\_‘k’$: One censored deprivation score for each poverty cutoff  \textit{k}
2052\end{itemize}
2053
2054To relate better to statistic, the identification function can be equivalently expressed as ${\mid}c_i{\geq}k$ where ${\|}[.]$ is an \textit{identification function} that takes a value of \textit{1}if the indicated condition $c_i{\geq}k$ is true for $i^{th}$ person, and \textit{0} otherwise.\\
2055
2056Built of the basis of the FGT unidirectional poverty measure, the aggregation step generates a spectrum of parametric class of intertwined and multifaceted  deprivations and poverty measures. Like wise, the FGT measures that are viewed as a mean of n appropriate vector built from the original data and censored using the poverty line; the Adjusted Headcount Ratio in the AF methodology, signified as $M_0(X;z)$, is the mean of the censored deprivation score sector:
2057
2058$M_0=\mu(c(k))=\dfrac{1}{n}\times\sum^{n}_{i=1}$
2059
2060Based on this matrix, we construct a censored vector of deprivation counts $c(k)$ which differs from vector c in that it it counts zero deprivations for those not identified as multidimensionally poor. As a result, multidimensional poverty index (MPI) is the mean of the matrix $g^0(k)$ multiplied by the number of columns it contains ${d}$. \\
2061
2062In other words, $MPI=d\mu(g^0(k))$, in which $\mu$ denotes the arithmetic mean operator. The multidimensional poverty index can be broken down in to two measures: the multidimensional headcount ratio (\textit{H}) and the average deprivation share among the poor (\textit{A}). \textit{H} is the proportion of people who are poor and is gauged by $H=\frac{q}{n}$ where \textit{q} is the number of poor people. The fraction of weighted indicators where person is deprived is $c_i(k).$ The intensity or average of that fraction among the poor is then expressed as $A=\sum^n_{i=1}\frac{c_i(k)}{q}$.\\
2063
2064In other words,  MPI creates various temporary variables in the process of calculating the induces ad the most important are;
2065\begin{itemize}
2066 \item $rho\_‘k’$: One identifier of the poor for each poverty cutoff \textit{k}
2067\item $c0k\_‘k’$: One censored deprivation score for each poverty cutoff  \textit{k}
2068\end{itemize}
2069
2070The incidence or headcount ratio \textit{(H)}, the percentage share (proportion) of people who are poor or incidence of multidimensional poverty can also be expressed using the alternative notation as
2071\begin{equation}
2072H(X;w)=\dfrac{{\sum}^n_{i=1}{\|}[c_i{\geq}k]}{n}
2073\end{equation}
2074
2075The draw back of the headcount ratio \textit{H} is that it violates the dimensional monotonicity as it remains the same while the number of deprivation rises for a person. 
2076
2077In turn, poverty intensity (average deprivation share among the poor) \textit{(A)} is the average deprivation score across the poor or the average deprivation share among the poor. Recall that the censored deprivation score $c_i(k)$ stand for the share of possible deprivations experienced by  a poor person \textit{i}. Thus, the average deprivation score across the poor is given by 
2078\begin{equation}
2079A=\frac{1}{q}\sum^n_{i=1}\rho^k_i\rightarrow
2080\end{equation}
2081 mean of $rho\_'k'==1$\\
2082Alternatively, the average intensity/severity of multidimensional poverty or the average share of dimensions (proportion of weighted deprivations)people suffer at the same time or deprivation share among the poor can be expressed commonly as:
2083\begin{equation}
2084A={\sum}^q_{i=1}\dfrac{c_i(k)}{q}
2085\end{equation}
2086
2087The other core measure of the AF class is $M-0$, the 'Adjusted Headcount Ratio', $M_0 = H{\times}A,$ which accounts for both the incidence of poor individuals and the intensity of their multiple deprivations   or the weighted sum of deprivation  poor people experience divided by the total number of people/population is given by,
2088\begin{equation}
2089M_0(X;z)={\mu}(c(k))=X{\times}A=\dfrac{q}{n}{\times}\dfrac{1}{q}{\sum}^q_{i=1}c_i(k)=\dfrac{1}{n}{\sum}^q_{i=1}c_i(k)
2090\end{equation}
2091or 
2092 
2093\begin{equation}
2094M_0={\mu}(c(k))=\dfrac{1}{n}{\sum}^n_{i=1}{\sum}^d_{j=1}w_j{g^0_{ij}(k)}
2095\end{equation}
2096
2097Equivalently, in the notation below where the identification function for the $i^th$ is multiplied by the weighted deprivation score $c_i$ of the $i^{th}$ person. This censors (replaces by 0) the deprivation of the non-poor. The sum of the deprivation scores thus censored by the identification function, divided by $n{\times}d,$  provides the alternative equation below:
2098\begin{equation}
2099M_0(X;z)=\dfrac{1}{nd}{\sum}[{\|}(_i{\geq}k){\sum}^d_{j=1}w_jg^0_{ij}(x_{ij})]
2100\end{equation}
2101We emphasized on $M_0$ here as it a has a distinctive properties that indicate a plus benefit of using this approach as it ($M_0$) satisfies the ordinality property which states that whenever variables and their corresponding cutoffs are modified in such a way that their scale is maintained (hence the admissible transformation) the poverty value should not change. This is so because the identification method is performed based on the counting approach procedures that dichotomizes achievements int to deprived and on-deprived; equivalent transformation of the scale of the variables  will not impact the set of people who are identified as poor. It can further be noted that the weights are attached to deprivations are principally independent of the scale of indicators and are implemented after the deprivation status has been determined. \\
2102
2103This  asserts that the MPI is more relevant for consistency in targeting within policies or programs aiming at improving the well-being of the deprived and the poor employing the ordinal indicators. \\
2104Likewise, aggregation procedure to obtain the $M_0$ measurement is conducted employing the censored deprivation matrix, that represents the deprivation status of each poor person in every dimension and also use the \textit{1} otherwise \textit{0} dichotomy. In this (aggregation) procedure, the deprivations oof the poor are weighted, yet again, the weights are independent of the indicators' scale and  are implemented after the deprivation status of the poor has been determined. Therefore, equivalent transformation of the scale of variables will not affect the aggregation of the poor and so will not affect the over all poverty values. \\
2105
2106The fact that$M_0$ satisfies the ordinality property is extremely important when poverty is analyzed from the capability perspective, since  a great deal of core functions are commonly gauged using ordinal or (ordered categorical) variables. To sum up, the Adjusted Headcount Ratio changes when the number of deprivation rises for a person and hence it satisfies the dimensional monotonicity. Moreover, of the numerous gains gains of employing the Adjusted Headcount Ratio, $M_0$ conveys information on deprivations, it is valid can be applied for ordinal data, it is simple for computation and by and lager have extremely important properties such as subgroup decomposition and dimensional breakdown. 
2107\subsubsection{Multidimensional Poverty Measurement Subgroup Decomposition}
2108The subgroup decomposition property (which also holds true in he uni-dimensional poverty measurement approach)has been so vital in analyzing poverty by regions, by ethnic group, by gender, by religion, and by other subgroups as defined the context of the ground that fits. The $M_0$ satisfies population subgroup decomposability principle similar to the FGT class of indices (Foster, Greer, and Thorbecke 1994). Population subgroup decomposability enables us to comprehend and monitor the subgroup $M_0$ levels and compare them with the over all $M_0.$ According to Alkire et. al.(2015: p-163), the population share and the achievement matrix of subgroup \textit{l} are designated by $v^l=\dfrac{n^l}{n}$ and $X^l,$ respectively. Thus, the over all $M_0$ can be expressed as:
2109\begin{equation}
2110M_0(X)={\sum}^m_{l=1}v^lM_0(X^l)
2111\end{equation}
2112Given the additive form of equation (34), it is possible to compute the contribution of each sub group to over all poverty. Let's represent the contribution of subgroup \textit{l} to aggregate poverty by $D^0_l,$ which can be formulated as
2113
2114\begin{equation}
2115D^0_l=v^l\dfrac{M_0(X^l)}{M_0(X)}
2116\end{equation}
2117
2118It can be observed that the contribution of subgroup \textit{l} to aggregate poverty depends both on the level of poverty in sub group \textit{l} and on the population share of the subgroup. Whenever the contribution to poverty of a region or some other group greatly exceeds its population share, this suggests that there exists unequal distribution of deprivations and multidimensional poverty in the country; with some regions or groups bearing  a disproportionate share of poverty. Succinctly, the sum of the contribution of all groups is equal to one.  \\
2119
2120Alternatively, the contribution of subgroups to the overall adjusted headcount ratio could be written as:
2121
2122\begin{large}
2123Population Subgroup:
2124\end{large}
2125A poverty measure respects the property of subgroup decomposability of a poverty measure for evaluation of poverty reduction programs,  an extension of monotonicity requirement, and other core practical reasons in poverty measurement the subgroup consistency property has tremendous significance. \\
2126 
2127By the same token, monotonicity, the key property of subgroup decomposability requires poverty to fall when one person's poverty level is reduced. Where as subgroup consistency requires aggregate poverty to fall when one group's poverty level is reduced. However, a reduction in one group's poverty may be accompanied by both increase and fall in individual incomes depending on the magnitude of an increment and decrements. In this case, when the deduction is bigger than the surge and hence it is fully offset so that there will be overall poverty reduction in the group. \\
2128
2129Given that a mutually exclusive and exhaustive population subgroup with the following elements:\\
2130The population size of matrix \textit{X} is \textit{n}, matrix \textit{X} is divided into two population subgroups as:
2131\begin{itemize}
2132\item Group 1: $X_1$ with population size $n_1$
2133\item Group 2: $X_2$ with population size $n_2$
2134\item And notice that $n=n_1+n_2$
2135\end{itemize}
2136Then, population subgroup decomposability is is possible and under the following conditions. A poverty measure is decomposable if 
2137\begin{equation}
2138P(x)= \frac{n_1}{n}P(x_1)+\frac{n_2}{n}P(x_2)
2139\end{equation}
2140
2141Again, the contribution of each group to overall poverty  may be computed as 
2142\begin{equation}
2143C(X_1)=\dfrac{n_1}{n}\dfrac{P(X_1)}{P(x)}
2144\end{equation}
2145It must be noted that additive and/or subgroup decomposability implies subgroup  consistency, yet the converse does not necessary hold true. \\
2146
2147\subsubsection{Multidimensional Poverty Measurement Dimensional Breakdown}
2148A second for of breakdown for $M_0$ that helps to identify deprivations experienced by the poor measurement progress in alleviating deprivations is breakdown by dimension in lieu of population subgroup.\\
2149
2150A multidimensional poverty measurement that satisfies the dimensional breakdown property can be expressed as a weighted sum of dimensional deprivations, wt which the particular case of $M_0$ referred to as the censored headcount ratios. This property is very essential as it enables us to analyze the compositions of multidimensional poverty. In this study for example, after decomposing the over all poverty in Ethiopia by region, gender, area(residence), age-group and so forth; we examine the the MPI dimensional break down by indicators, domains and across these subgroups to study how different groups have different dimensional deprivations, that is different poverty compositions. \\
2151
2152While dealing with MPI breakdown by dimensions, we have to shed light on censored headcount ratios and uncensored headcount ratios that are essential concepts. \\
2153\begin{Large}
2154Censored Headcount Ratio:\\
2155\end{Large}
2156Censored headcount ratio of a dimension is defined as the percentage of the population who are multidimensionally poor and simultaneously deprived in that dimension.  For illustration, let's denote the $j^th$  column of the censored deprivation matrix $g^0(k)$ as $g^0_{.j}(k)$ and mean of the column for the chosen dimension as:
2157\begin{equation}
2158h_j(k)=\dfrac{1}{n}{\sum}^n_{i=1}g^0_{ij}(k)
2159\end{equation}
2160Then, $h_j(k)$ is simply the censored headcount ratio of dimension \textit{j}. What is the interpenetration of $h_j(k)?$ The censored headcount ratio $h_j(k)$ is the proportion of the population that rea identified as poor $c_i{\geq}k$ and are deprived in dimension \textit{j}.\\
2161The additive nature of the $M_0$ measure allows it is to be expressed as a weighted sum of the censored headcount ratios, where the weight on dimension \textit{j} is $w_j,$ the relative weight assigned to that dimension.  We have already seen in expressions (31) and (32) that $M_0=\dfrac{1}{n}{\sum}^n_{i=1}{\sum}^d_{j=1}w_j{g^0_{ij}(k).}$  This expression can be reformulated as:
2162\begin{equation}
2163M_0=\dfrac{1}{n}{\sum}^n_{i=1}{\sum}^d_{j=1}w_j{g^0_{ij}(k).}={\sum}^j_{j=1}w_j[\dfrac{1}{n}{\sum}^n_{i=1}g^0_{ij}(k)]={\sum}^d_{j=1}w_jh_j(k)
2164\end{equation}
2165
2166According to Alkire et al.(2015), discussions depending on the censored head count ratios could be supported by the percentage contribution of each domains to aggregate poverty. The censored headcount ratio displays the extent of deprivations among the destitute however, not the relative values of dimensions. Various dimensions may have equal censored headcount ratios yet very different contributions to over all poverty. This is so because the contributions relay on the censored headcount ratio and the weight (value) assigned to each dimensions. For illustration, let's say the contribution of dimension \textit{j}to $M_0$ by ${\omega}^0_j,$ where
2167\begin{equation}
2168{\omega}^0_j(k)=w_j\dfrac{h_j(k)}{M-0}
2169\end{equation}
2170for each $j=1,...,d.$ Whenever the contribution of poverty of a certain indicator greatly exceeds its weight, there i a relatively high censored headcount ratio for this indicator. \\
2171
2172\begin{Large}
2173Uncensored (Raw) headcount Ratio
2174\end{Large}
2175The uncensored (raw) headcount ratio of an indicator and dimension is defined is defined as the proportion of the population that are deprived in that indicator/dimension. It summarizes deprivations of the poor (censored headcount) with deprivation of the non-poor. The uncensored headcount ratio of dimension \textit{j} is computed from the (uncensored) deprivation matrix $g^0$  as the mean of the $j^{th}$ column vector $g^0_j.$ Therefore,
2176
2177\begin{equation}
2178h_j=\dfrac{1}{n}{\sum}^n_{i=1}g^0_{ij}
2179\end{equation}
2180
2181is the uncensored (raw) headcount ratio of indicator/dimension \textit{j.}\\
2182The censored headcount ratio generally differs from the uncensored headcount ratio except when the identification criterion employed is the union. Where as in such cases, a person is identified as poor if the person  is deprived in any dimensions, thus, no deprivations are censored. To this effect, the censored and uncensored headcount ratios are equal under such scenarios. \\
2183
2184In this study, we emphasized on the critical principles that multidimensional poverty measurement should fulfill; for example equation (1)to (21) and  Adjusted Headcount Ration analysis equations (22)to (41) as many of our poverty indicators are practically of ordinal scale. Nevertheless, if all indicators are cardinal, the expressions (equations) for the Adjusted Headcount Ratio can be expanded to reflect additional measures such as depth of deprivations poor people experience below the deprivation cutoff in each dimension. Briefly, the identification procedure is exactly similar to what we have demonstrated above with respect to $M-0.$
2185The only difference lies in the aggregation procedures. The detail discussions and mathematical expressions can be found in (Alkire et.al., 2015: P-173-176).\\
2186\subsection{Creating Ethiopian Multidimensional Poverty Maps}
2187Under this section, we presented the approach adopted in creating the Ethiopian MPI multidimensional poverty maps. 
2188\subsubsection{Ethiopian MPI Spatial Data Analysis in Stata}
2189According to Maurizio Pisati (2012), Stata users can perform spatial data analysis using a variety of user-written commands published in the Stata Technical Bulletin, the Stata Journal, or the SSC Archive. Let’s succinctly elucidate the use of six such commands: spmap, spgrid, spkde, spatwmat, spatgsa, and spatlsa. Besides, two most commonly applied pair of Stata commands/suites for fitting spatial regression models are spatreg and sppack.  A space represents as a plane, i.e., as a flat two-dimensional surface.\\
2190Furthermore, according to (Bailey and Gatrell 1995), in spatial data analysis, we can extricate two conceptions of space. These are entity view, space as an area filled with a set of discrete objects and field view, space as an area covered with essentially continuous surfaces.\\
2191
2192Moving on, information about spatial objects can be classified into two categories such as spatial attributes and non-spatial attributes. The spatial attributes of a spatial object consist of one or more pairs of coordinates that represent its shape and/or its location within the study area. N other words, the spatial attributes of a spatial object consist of one or more pairs of coordinates that represent its shape and/or its location within the study area. The non-spatial attributes of a spatial object consist of its additional features that are relevant to the analysis at hand.\\
2193
2194Moreover, according to their spatial attributes, spatial objects can be classified into two basic types.  Points (point data) and polygons (area data). A point si is a zero-dimensional spatial object located within study area A at coordinates (si1, si2). Points can represent several kinds of real entities, e.g., dwellings, buildings, places where specific events took place, pollution sources, trees and so forth.\\
2195  
2196In addition, in order to draw a geographical map, you need one or more files defining the boundaries of the geographical area of interest and, possibly, of its administrative subdivisions. The Esri shapefile (so as to obtain the shapefiles, go to http: www.diva-gis.org/gdata and download the data for your country) is one of the most common vector formats for storing geospatial data of this kind. A shapefile is actually a set of three mandatory files, plus one or more optional files.  Mandatory files such as .shp, stores the coordinates of the spatial objects; .dbf, stores the attributes of the spatial objects; .shx, indexes the spatial objects. Other relevant optional file:  .prj: defines the spatial reference system.\\
2197(To obtain the shapefiles go to http:www.diva-gis.org/gdata and download the data for your country).\\
2198
2199%Then, use the following command to transform the shapefile in a Stata file: shp2dta using ETH$_$adm1, database(region) coordinates(map) genid$(id)$ gencentroids(center). In the new data set called "region" check the id for each region (br id NAME$_$1) VARNAME$_$1. In the collapsed results data set generate a new variable "id" following the structure of the data set "regions.dta"\\
2200
2201
2202With spmap, you can graph data onto maps and produce results. spmap is a user-written command by Maurizio Pisati (2014). The following process explains how to use spmap. 
2203 spmap is aimed at visualizing several kinds of spatial data, and is particularly suited for drawing thematic maps and displaying the results of spatial data analyses.
2204
2205A very important  application in mapping poverty is  $spmap$ in which its functioning rests on three basic principles: in other words the ado procedures which shows the impression how the information of interest and geographical data/polygon coordinates function in the practical applications.
2206
2207First, a base map representing a given study region S made up of M polygons is drawn.  Second, at the user's choice, one or more types of additional spatial objects may be superimposed onto the base map. In the current version of spmap, six different types of spatial objects can be superimposed onto the base map: polygons $({via\; option\; polygon()})$, polylines $({via\; option\; line()})$, points $({via\; option\; point()})$, diagrams $({via\; option\; diagram()})$, arrows $({via\; option\; arrow()})$, and labels 
2208$({via\; option\; label()})$. Third, at the user's choice, one or more additional map elements may be added, such as a scale bar $({via\; option\; scalebar()})$, a title, a subtitle, a note, and a caption$(via title_options)$.\\
2209 
2210Proper specification of spmap options and sub-options, combined with the availability of properly formatted spatial data, allows the user to draw several kinds of maps, including choropleth maps, proportional symbol maps, pin maps, pie chart maps, and noncontiguous area cartograms. Whereas, providing sensible defaults for most options and supoptions, spmap gives the user full control over the formatting of almost every map element, thus allowing the    production of highly customized maps.
2211
2212%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2213\begin{enumerate}
2214\item	Obtain and install the spmap, shp2dta, and mif2dta commands. 
2215\item	Search the web for the files that describe the map onto which you want to graph your data. You can use ESRI shapefiles or MapInfo Interchange Format.\\
2216 
2217ESRI shapefiles are the more common. In this format, there are three files associated with a map: an .shp shape file, a .dbf dBASE file, and an .shx index file. You need only the .shp and .dbf files. You translate those files into a format usable by Stata with the shp2dta command. Doing so creates two .dta datasets, one corresponding to each file.\\
2218 
2219The MapInfo Interchange Format consists of two files with suffixes .mif and .mid. You translate these files with the mif2dta command. Just as with shp2dta, doing so creates two .dta datasets. 
2220\item	Look at the translated .dbf (.mid) file. It is a .dta dataset and you just use it. Examine the dataset to determine the coding used by the map's authors to designate areas. For instance, 1 might mean Addis Ababa and 2 Amharaa in one dataset, and 1 might mean Oromia and 2 Tigray in another. 
2221\item	You have data you want to plot onto a map. Let’s assume that the data are stored in a Stata .dta dataset. You need to modify your .dta dataset to use the same location coding as that used in the map. Call that variable id. 
2222\item	Merge on id the translated .dbf (.mid) dataset with your dataset containing the statistics to be graphed. 
2223\item	With the merged dataset in memory, make the graph by using spmap. You will tell spmap about the other translated dataset (the coordinate dataset) 
2224\end{enumerate}
2225
2226We neither have the database on Ethiopian Map coordinates nor do we have the same codification in the shape file (map coordenates.dta file) and in the original data. That is why we have to generate a variable that matches in the two data sets.\\
2227
2228Order and codes for regions from map data (this is the spelling from the map database, but please adjust the label define command as appropriate): When working to create maps, you need to switch between two file locations:
2229\begin{enumerate}
2230
2231\item Where the MPI data sets are stored
2232\item. Where the downloaded shape files are stored
2233The easiest way to do this is to use globals, as I've shown below (just copy in your file locations instead of mine).
2234\end{enumerate}
2235Change the directory location to where the map data was downloaded.
2236
2237Transform the shape data from files in the map data folder into:
2238\begin{itemize}
2239\item a data file called $eth_data.dta$\\
2240\item a coordinates file called $eth_coord.dta$\\
2241 \end{itemize}
2242Note that the variable name in the genid() bracket must be the same as the region variable in the MPI data set, in order to merge them together.
2243
2244\subsubsection{MPI Robustness Analyses and Statistical Analysis/Inference}
2245This section sheds light on not only tools that can be used to test robustness of pairwise comparisons but also the robustness of over all rankings with respect to the initial choice of parameters. The core questions that this section tries to answer are:
2246How accurate are the estimates? If they are used for policy drafting while targeting the poor, what is the chance that they
2247are mistaken? How sensitive policy prescriptions are to choices of
2248parameters used for designing the measure ('Robustness
2249Analyses')?  And how accurate policy prescriptions are subject to the sample from which the they are computed ('Statistical Inferences')? \\
2250
2251In the Ethiopian context The Ethiopian Federal Government of Ethiopia (GoE), regions and Woredas develop two types of budgets: recurrent and capital budgets.
2252\textit{Recurrent budget}: is spending on items that are consumed and are repeated every budget year, such as salaries for teachers, health workers and development assistants, offices' operational costs, medicines, books and electricity. Salaries for employees of public bodies dominate the recurrent
2253budget.\\
2254
2255\textit{Capital budget}: is spending on items that will last for several years, like buildings, roads and water points. Money is needed to build schools, health post, water points and roads. Once constructions under the capital budget, for schools, health posts, etc., are completed, the maintenance and service provision in most
2256cases continues to be assigned in the recurrent budget.  \\
2257If we take a look at the budget breakdown for the 2018 budget year or 2010 Ethiopian Calendar budget yer (all figure in Ethiopian currency Birr: considering the current exchange rate, July 12, 2017: $1\, US \,Dollar = 23.25 \, Ethiopian \, Birr$.
2258
2259\begin{quotation}
2260...the proposed spending would see 117.3 billion Birr out of next year's budget redirected to the nine regional states and two city administrations, allocated based on a special formula devised by the House of Federation (HoF). The second largest chunk, in tune of 114.7 billion, is allotted to federal government's capital spending complementing the 81.84 billion Birr recurrent expenditure of the federal bodies. The last component which is a seven billion allocation is the support for regional states in their effort to meet the Sustainable Development Goals (SDGs).\\
2261
2262Based on HoF’s formulation, which takes into consideration the revenue potential and expenditure needs of the regional states, Oromia is entitled to 39.85 billion Birr out of 117.3 billion subsidy fund followed by the Amhara and Southern regional states each looking to cash a check for 24.98 and 23.25 billion Birr, respectively, in the coming fiscal year. Somali Regional State is the next biggest earner on the subsidy budget taking home 11.54 billion Birr followed by Tigray securing 6.97 billion Birr, Afar 3.49 billion, Benishangul-Gumuz 2.11 billion, Addis Ababa 1.64 billion, Gambella 1.54 billion, Dire Dawa 1.02 billion and Harari 878.74 million Birr.\\
2263
2264Both Dire Dawa and Addis Ababa seem to be the latest inclusions to the subsidy illegibility list in connection to their status of being federal cites and not regional states. According to the principles guiding the subsidy grants, it is only the nine regional states which are original founders of the federation which are entitled to subsidy ...\\
2265 Source: http://thereporterethiopia.com/content/budget-breakdown: website  accessed on July 10, 2017.
2266\end{quotation}
2267Therefore, the robustness analysis/test was applied since the researcher sought to test if the regional comparisons are robust and statistically significant.  Besides, as we planned to scrutinize is there is a steepest decrease in poverty in the regions and dimensions, we conducted a test if the inter-temporal comparisons are robust and statistically significant. It's vital to conduct a robustness test analysis as comparisons may alter when parameters change. It is also essential to carryout statistical test since difference in estimates may be of the same magnitude, yet statistical inference may not be the same.
2268\subsubsection{Robustness Analyses}
2269According to Alkire et al.(2015: p-234), the parameters that derive the multidimensional poverty estimates and poverty comparisons based on the Adjusted Headcount Ratio $M-0$ measure and its partial indices are based on the following parameter values:\\
2270We employed a tool that tests the extreme form of pairwise robustness as applied in the stochastic dominance analysis that is simply an extreme form of robustness. Further more, of the two key types of  robustness analysis; namely, dominance analysis for changes in the poverty cutoff-with respect to poverty cutoff similar to the uni-dimensional dominance/sensitivity test procedure  and multidimensional dominance technique; and rank robustness analysis-with respect to weights and with respect to deprivation cutoffs. \\
2271\begin{itemize}
2272\item The Set of Indicators:
2273The set of indicators designated by $j=1,...,d$;
2274\item The Set of Deprivation Cutoffs: The set of deprivation cutoffs (represented by vector \textit{z});
2275\item The Set of Weights or Deprivation Values denoted by vector \textit{w} and
2276\item the poverty cutoff denoted by \textit{k}
2277\end{itemize}
2278It should be noted that a change in any of these parameters could affect the aggregate poverty estimate or comparisons across regions and countries. 
2279To build the dominance test procedure, we focused on first-order stochastic dominance technique to explain and pinpoint consistent comparisons with respect to poverty cutoff of the two most prominently used poverty measures; namely the Adjusted Headcount Ratio ($M-0$) and the Multidimensional Headcount Ration \textit{(H).}\\
2280Consider the notations of two univariate distributions  of achievements \textit{x} and \textit{y} with the cumulative distribution function of (CDF) $F_x$ and $F_y,$ where $F_x(\theta)$ and $F_y(\theta)$ are the share of population in the distribution \textit{x} and \textit{y} with achievement levels less than ${\theta}{\in}\Re_+.$\\
2281Distribution \textit{x} first-order stochastically dominates distribution \textit{y} (or $x\, FSD\, y$ if and only if
2282\begin{equation}
2283F_x(\theta)  {\geq}  F_y(\theta)
2284\end{equation}
2285for all ${\theta}$ and $ F_x(\theta) < F_y(\theta)$ for some ${\theta}.$ Strict FSD requires that $F_x(\theta)<F_y(\theta)$ for all ${\theta}.$
2286
2287To shed light on how this concept can be applied to unanimous pairwise comparisons using $M_0$ and \textit{H} between any two  distributions of deprivation scores across the population. Give that deprivation cutoff vector \textit{z} and a given weighting vector \textit{w}, the FSD instrument can be employed to assess the responsiveness of any pairwise comparisons to different poverty cutoff \textit{k}.\\
2288
2289To illustrate, let's designate the (uncensored) deprivation score vector by \textit{c}. Recall that an element of \textit{c} stands for the deprivation score and a large deprivation score implies a lower level of well-being. More importantly, the FSD was applied to convert deprivations into attainments by transforming the deprivation score vector \textit{c} in to an attainment score vector $1-c,$ and to employ the tool directly on the deprivation score vector \textit{c} as explained in the Alkire and Foster(2011a). Without applying any transformation in the more direct technique, a person is identified as poor if the deprivation score is larger than or equal to the poverty cutoff \textit{k}; unlike in the attainment space where a person is identified as poor if the person's attainment falls below a certain poverty cutoff. 
2290
2291\subsubsection{Rank Robustness Analyses}
2292The Dominance Analysis for changes in the poverty cutoff with respect to poverty  cutoff and multidimensional dominance across all comparisons discussed above is an extreme for of robustness. Although Stochastic Dominance (SD) conditions are very important foe pair by pair analysis and serve as the strongest possible comparisons; these conditions, however, could be too stringent and may not hold for subgroups or countries. As a resolution to such pitfall, the rank robustness analysis that evaluates the extent to which a ranking, that is, an ordering of more than two entries obtained under a specific set of parameters' values, is preserved when the value of some parameter is modified is applied. An intuitive  and useful technique that assist us to evaluate a robustness of a ranking is to compute the percentage of pairwise comparison that are robust to variations in parameters. In other words, working out the proportion of pairwise comparisons that have the same ordering. \\
2293
2294Furthermore, vital optional ways to assess the robustness of ranking is by calculating a rank correlation coefficient between the original ranking of entities and those obtained with alternative parameters. Two prominently applied  approaches to compute rank correlation coefficient are the Spearman rank correlation coefficient $(R^{\rho})$ and the Kendall rank correlation coefficient $(R^{\tau})$.\\
2295The natural producer here is that first various ranking of countries or subgroups are generated for various specifications of parameters such as different weighting vectors or different deprivation/poverty cutoffs. Then, the pair-wise ranks and rank correlation coefficients are computed.\\
2296 
2297To shed more light on this, consider the set of ranks of correlation across m population subgroups is denoted by $r=(r_1, r_2,...r_m$, where $r_l$ is the rank attributed to subgroup \textit{l}. The subgroups may be ranked by their level of multidimensional headcount ratio, i.e., the Adjusted Headcount Ratio, or any other partial and consistent sub-indices.\\
2298 
2299Correspondingly, designate the rank for an optional specification of parameters by $r'$, where $r'_l$ is the rank attributed to subgroup \textit{l}. If the initial and alternative specifications yield the same set of ranking across subgroups, then $r_l=r'_l$ for all $l=1,...,m$. On the other hand, if the two specifications yield different/opposite sets of rankings, then, $r_l=r'_{m-l+1}$. Here, we state the two sets of ranking as perfectly negative whereas the former case are stated as perfectly positive. \\
2300
2301To cut issues short, the Spearman rank correlation coefficient can be expressed as 
2302\begin{equation}
2303R^{\rho}=1-\dfrac{6{\sum}^m_{l=1}(r_l-r'_l)^2}{m(m^2-1)}
2304\end{equation}
2305Logically, for the Spearman rank correlation coefficient, the square of difference in the two ranks of each subgroup is commuted  and an average is taken across all subgroups. The $R^{\rho}$ is normally bound between \textit{-1} and \textit{+1}. The lowest value of \textit{-1} is obtained when two ranking are perfectly negatively correlated with each other, whereas the largest value of \textit{+1} is obtained when the two ranking are perfectly positively correlated with another.\\
2306
2307The Kendall's Tau rank correlation coefficient is based on the number of concordant pairs and discordant pairs. Thus, $R^{\tau}$ is the difference in the number of concordant and discordant pairs divided by the total number of pairwise comparisons. Thus, the Kendall's Tau rank correlation coefficient can be expressed as 
2308\begin{equation}
2309R^{\tau}=\dfrac{ Number of Concordant Pairs-Number of Discordant Pairs}{\dfrac{m(m-1)}{2}}
2310\end{equation}
2311Notice that like $R^{\rho}$, $R^{\tau}$ also takes the values in between \textit{-1} and \textit{+1.} If there exist ties, this measure ought to be adjusted for these ties, and the adjusted Tau ($\tau$) is said to be tau-$\beta$.
2312
2313\subsubsection{MPI Confidence Interval and Statistical Inference:Significance Test}
2314Sampling distributions allow us to make statistical inference about the unobserved true population parameter in relation to the
2315observed sample statistic using two approaches; namely, different estimation methods (confidence interval)and conducting hypothesis tests (computing p-values). \\
2316Different sample surveys, even when conducted at the same time and despite following the same research design could most probably provide a different set of estimates for the same population parameter. Thus, it is paramount to compute the measures of confidence or reliability for each estimates from a sample survey.\\
2317The procedure of determining the standard deviations of an estimate leads us to estimating the standard errors. Thus, the lower the magnitude of the standard error, the larger the reliability of the corresponding estimates. Determining standards errors in turn are key for hypothesis testing (computing the p-value), for constructing the confidence intervals, both of which are vital for robustness analysis, statistical inference, and by and large for drawing policy conclusions.\\
2318
2319The commonly posed question here are is the aggregate poverty larger or smaller in one region that another region? In rural area or in urban residents? In male headed or female headed households?  and other  similar question. Moreover, has the overall poverty increased or decreased over time? Often time, such conclusions related to the population are made from or inferred from a sample since garnering data  form the population could be too expensive or impossible. \\
2320
2321Uniquely, inferential statistics such standard errors (SE) and confidence Interval (CI) deal with inference about population based on the behavior of the samples. Both SEs and CIs do assist us to determine how likely are the results obtained from a certain sample (or samples) are the same results that would have been gained from the entire population. There are various methods to estimating standard errors, the most common ones are: 
2322\begin{enumerate}
2323\item Simple Random sampling with Analytical Approach: Formula that either give the exact to the asymptotic approximation of the standard error (Yalonetzky, 2010). For the detail discussion and mathematical derivation of the simple random sampling, stratified sampling with analytical approach see (Alkire et. al.,2015:p-249-252).
2324\item Re-sampling  Approach: Standard errors ad confidence intervals may be computed via bootstrapping (Alkire and Santos, 2014)
2325\end{enumerate}
2326The analytical approach is based on the premises that the sample survey used for estimating the population parameter are significantly smaller in size as compared to the population size under considerations.\\
2327The practical motivation behind the  second presumption that states    each sample should be treated as drawn from the population with replacement; is the size of the sample survey as compared to the population size. To conduct the bootstrapping method (re-sampling approach)the following conditions are required: random artificial sampling are drawn from the data-set; an estimate is produced from the artificial sample and stored; and assuming the artificial samples are independent and identically distributed (iid), the standard error is calculated using the artificial sample estimates. \\
2328
2329For  the multidimensional and censored headcount ratios, SEs can be computed employing the following methods:
2330\begin{equation}
2331M_0(X;z,w,k)=\dfrac{{\sum}^n_{i=1}c_i(k)}{n}
2332\end{equation}
2333\begin{equation}
2334A(X;z,w,k)=\dfrac{{\sum}^n_{i=1}c_i(k)}{q}
2335\end{equation}
2336
2337\begin{equation}
2338H(X;z,w,k)=\dfrac{{{\sum}^n_{i=1}\|[c_i{\geq}k}]}{n}
2339\end{equation}
2340\begin{equation}
2341h_j(X;z,w,k)=\dfrac{{{\sum}^n_{i=1}\|[(c_i{\geq}k){\bigwedge} (g^0_{ij})}]}{n}
2342\end{equation}
2343Note that ${\bigwedge}$ is the logical 'and' operator. The standard error of the subgroups' $M_0s$ and partial and consistent sub-indices may be computed in the same way and so we only outline the standard errors of equations (45)-(48).\\
2344
2345To explain confidence intervals (CI) succinctly, the probability that confidence interval contains the parameter is referred as confidence interval (CI). By recalling the central limit theorem and using the notions under the normal distribution, we are enable to compute the the critical value associated with the significance level which is give by the inverse of the standard normal distribution. \\
2346Further aspects are found, e.g., in Alkire and Foster (2011a,b). Alkire et. al., (2015) provide a more comprehensive discussion).
2347
2348\subsubsection{Multidimensional Poverty Analysis and Hypothesis Testing}
2349The most vital step of any scientific method study involves testing the hypothesis to determine if it is true. This is a critical stage within the scientific method. The observations must be tested to make sure they are unbiased and reproducible. In economics, extensive testing and observation is required because the outcome must be obtained more than once in order for it to be valid. In other words, hypothesis testing can be defined as an act in statistics whereby an analyst tests an assumption regarding a population parameter. The methodology employed by the analyst depends on the nature of the data used and the reason for the analysis. Hypothesis testing is used to infer the result of a hypothesis performed on sample data from a larger population.\\
2350
2351Let's briefly describe the following concepts pertinent to hypothesis testing procedure and the likely root causes. \\
2352Sampling Error (estimation error): The plausible root causes of  sampling error are:
2353\begin{itemize}
2354
2355 \item i) the difference between a sample statistic and the true population parameter that the sample statistic is being used to estimate. 
2356 \item ii) the error caused by observing a sample instead of the population, caused by sample to sample variation.
2357 \item  iii) in practice the exact sampling error for a statistic is typically unknown but statistical theory provides procedures for estimating the of the sampling error, called the  standard error.          
2358 \item iv) the likely size of the sampling error can generally be reduced by increasing the sample size, although generally there are also costs involved with larger sample sizes.
2359\end{itemize}
2360
2361Conversely, the root causes of the con-sampling error could possibly be 
2362\begin{itemize}
2363\item i) a catch-all term for the deviations from the true population parameter that are not a function of the sampling error (e.g. poorly worded questions or untruthful respondents). 
2364\item ii) much more difficult to quantify than sampling error.
2365\end{itemize}
2366
2367More generally, testing of hypothesis involves:
2368\begin{itemize}
2369\item i) the standard approach to assessing whether an observed value of a variable or an observed relationship between two or more variables derived from sample data is real, that is holds true in the population or is a result of mere chance. 
2370\item ii) is an inferential statistics approach, that allows the researcher to use characteristics derived from sample data to make inferences about population characteristics. 
2371\item iii) involves comparing empirically observed findings with theoretical expected findings. 
2372\item iv) estimates the statistical significance of findings. 
2373\item v) involves posing opposing hypotheses about a population characteristic or the relationship between two or more population characteristics and then testing those hypotheses.
2374\item  vi) asks how often the observed results could be expected to occur by chance, if the answer is relatively frequently , then chance would remain a viable explanation of the effect but if relatively rarely, then chance would not be a viable explanation.
2375\end{itemize}
2376
2377\begin{large}
2378Hypothesis Testing Technical Issues:\\
2379\end{large}
2380Hypothesis testing is one of a fundamental problems in
2381statistics. A hypothesis is (usually) an assertion about the unknown
2382population parameters such as $\beta_1$ in
2383$ Y_i = \beta_0+ {\beta_1}X_i + {\epsilon_i}.$ 
2384Using the data set, the researcher has to determine whether
2385an assertion is true or false.\\
2386%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2387\begin{large}
2388Hypothesis Testing Breaking Down:\\
2389\end{large}
2390In hypothesis testing, an analyst tests a statistical sample, with the goal of accepting or rejecting a null hypothesis. The test verifies the analyst whether or not his primary hypothesis is true. If it isn't true, the analyst formulates a new hypothesis to be tested, repeating the process until data reveals a true hypothesis.
2391%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2392\subsubsection{Testing a Statistical Hypothesis}:
2393Statistical analysts test a hypothesis by measuring and examining a random sample of the population being analyzed. All analysts use a random population sample to test two different hypotheses: the null hypothesis and the alternative hypothesis. The null hypothesis is the hypothesis the analyst believes to be true. Analysts believe the alternative hypothesis to be untrue, making it effectively the opposite of a null hypothesis. This makes it so they are mutually exclusive, and only one can be true. However, one of the two hypotheses will always be true.\\
2394
2395Corresponding, let's  further discuss the null hypothesis and alternative hypothesis to concisely  comprehend theses closely intertwined concepts.\\
2396\begin{large}
2397Null Hypothesis:\\
2398\end{large}
2399Null hypothesis and alternative  involves the following essential points:
2400\begin{enumerate}
2401
2402\item in hypothesis testing it is the claim or statement about a population parameter that is assumed to be valid or true unless the observed data contradicts this assumption. 
2403\item the hypothesis that two variables are not related or that two statistics (e.g. means or proportions) are the same. 
2404\item symbolized as $H_0$. 
2405\item hypothesis test can either reject the null hypothesis, in which case the alternative hypothesis may be true, or fail to reject the null hypothesis.
2406\end{enumerate}
2407\begin{large}
2408Alternative Hypothesis (research hypothesis):\\
2409\end{large}
2410\begin{enumerate}
2411
2412 \item in hypothesis testing it is the opposite claim or statement about a population parameter from the null hypothesis. 
2413 \item the hypothesis that two variables are related or that two statistics (e.g. means or proportions) are different. 
2414 \item the hypothesis that the researcher expects to be supported, although this perspective is controversial.  
2415 \item  symbolized as $H_1$ or $H_a$.
2416\end{enumerate}
2417%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2418Commonly, a test procedure is specified by a test statistic $T = T(X)$, with the rejection region $S_l = S_l(\alpha)$, where $S_l$ is significance level, taking the form $T(X) > c$. For a pre-specified significance level $\alpha$, the critical value \textit{c}  is chosen (possibly in a data-dependent way and also dependent on $\alpha$) to control the size of the test, although one often resorts to asymptotic approximations. \\
2419%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2420Moreover, hypothesis testing is an inferential procedure in
2421which we test to see if we have sufficient evidence to
2422reject a null hypothesis $(H_0)$ in favor of an alternative
2423hypothesis $(H_a)$. These two hypotheses are meant to reflect the
2424research hypothesis being tested.
2425We choose between $H_0$ and $H_a$ by computing a test
2426statistic from a set of data, which quantifies the
2427strength of our evidence against $H_0$.\\
2428
2429Since we know the sampling distribution (z-distribution) that our test statistic follows, we can calculate the probability that
2430it would be a certain size if $H_0$ were true. If we get a
2431statistic that would be unlikely to occur if $H_0$ were
2432true, then we will reject $H_0$ in favor of $H_a$.\\
2433
2434The probability of obtaining a test statistic at least
2435as large as that which we obtained, if the null
2436hypothesis were true, is called the p-value. Small
2437p-values are considered to be evidence again $H_0$. If our p-value is small enough, we reject $H_0$, otherwise we fail to reject (FTR) $H_0$. In so doing, we are able to compute p-values because we know the sampling distributions of our test statistics. The field of statistics we are considering here is the parametric as opposed to(non-parametric which not based on known distributions) and all distributions thereof statistics which is   based on known normal distributions. \\
2438
2439Confidence intervals allowed us to find ranges of reasonable values for parameters we were interested in. Hypothesis testing will let us make decisions about specific values of parameters or
2440relationships between the following five parameters.\\
2441\begin{enumerate}
2442\item ${\pi}$ - population proportion (for instances adjusted headcount ratio among various age-groups  of Ethiopian Households 
2443\item $\mu$ - population mean (for example comparing poverty headcount ratio in Ethiopian Households using the multidimensional vs uni-dimensional poverty analysis
2444\item $\mu_1$−$\mu_2$ - difference in population means (example: compare average Multidimensional Headcount Ratio of of male-headed and female-headed Ethiopian Households)
2445\item $\mu_d$ - population mean difference (for paired data) (for instances: adjusted headcount ratio between urban and rural residents of Ethiopian Households)
2446\item $\pi_1$ − $\pi_2$ - difference in population proportions (example: compare proportion of multidimensional headcount ratio among Oromia vs. Tigray, Amhara vs. Oromia, SNNP vs. Tigray, Tigray vs Amhara and so forth)\\
2447
2448\end{enumerate}
2449
2450The hypothesis testing procedure  (on the five parameters stated above and other similar parameters) can be performed in
24514 steps. Note that how these steps are defined is
2452subjective; other researchers and textbooks may define
2453them differently, but the outcome maybe the same.
2454\begin{enumerate}
2455\item Set up the null $\&$ alternative hypotheses
2456\item State the significance level and the corresponding
2457critical value
2458\item Compute the test statistic $\&$ p-value
2459\item Make the statistical decision and interpret your
2460results. 
2461\end{enumerate}
2462
2463%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2464
2465In addition, hypothesis is a claim about a population characteristic (parameter)
2466\textit{Null Hypothesis}: the status-quo - initially assumed true\\
2467\textit{Alternative Hypothesis}: the researcher's proposal - what you hope to show\\
2468The gist of the hypothesis testing is to help us make a cogent and prudent decisions under the following scenarios.\\
2469Reject the null hypothesis in favor of the alternative only with convincing/significant evidence is provided. We may not say that we accept the alternative, since  only that we have significant evidence to reject the null. \\
2470
2471This is so because we could have made a mistake while formulating a hypothesis as follows:
2472Further illustrations on forms of Hypotheses:\\
2473\textit{Null Hypothesis}: $H_0$ : population parameter = some hypothesized value.\\
2474\textit{ Alternative Hypothesis}:$H_a$ : population parameter ${\neq}$ that same hypothesized value (two-sided) OR\\
2475 $H_a$ : population parameter > that same hypothesized value (one-sided to the right) OR\\
2476 $H_a$ : population parameter < that same hypothesized value (one-sided to the left)\\
2477 
2478%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2479In conclusion, we should be careful on deciding which theory of testing a statistical hypothesis to support. In this study, we adopted the decision making that is  based on the 'rare event' concept. Since the null hypothesis is the status-quo, we presume that it is true unless the observed result is extremely unlikely (rare) under the null hypothesis.\\
2480
2481The implication and under pinning definition is that if the data were indeed unlikely to be observed under the assumption that $H_0$ is true, and therefore we reject $H_0$ in favor of $H_a$, then we say that the data are statistically significant.\\
2482
2483%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2484In general, hypothesis tests can be used to draw inference on a wide variety of parameters. In this study, we are interested just in drawing inference upon the population mean, which is denoted $\mu$. For illustration, let's suppose from different previous experience that documented Ethiopia's poverty status (World Bank 2015 and 2016); using the Ethiopian national poverty line, we know that national absolute poverty headcount ranges  from 38.7$\%$ - 29.6$\%$. Besides, we can observes from the same report that absolute poverty headcount for urban and rural residents  ranges from 35.1$\%$ -25.7 $\%$ and from 39.3$\%$-30.4$\%$ respectively.\\
2485Sources: http://www.investopedia.com/terms/hypothesistesting.asp/ixzz4moT0HfdT; and website accessed on July 12, 2017.\\
2486
2487We believe that maybe the uni-dimensional poverty measures do not reflect the accurate reality on the ground since the average poverty measures usually used in this (uni-dimensional) approach understates the real nature, dimension, shape and forms of poverty as the monetary measures of poverty may not necessarily capture the various setting and matters of poverty as it may be much deeper and wider that just beyond lack of money. \\
2488\subsubsection{Formulating Main Hypothesis of this Study}
2489Under such circumstances, the main hypothesis that we are most interested in this study should be the research hypothesis denoted by $H_a$  which means the multidimensional headcount ratio in Ethiopia does not lie between the ranges of 38.7$\%$ - 29.6$\%$.  In fact, it is much higher than these figures. The same is for urban and rural poverty figures.\\
2490The other hypothesis is the null hypothesis, denoted by $H_0$ which states that the mean poverty headcount ratio in Ethiopian household lie between 38.7$\%$ - 29.6$\%$ nationally. Besides, this null hypothesis asserts that the  absolute poverty headcount for urban and rural residents  ranges from 35.1$\%$ -25.7 $\%$ and from 39.3$\%$-30.4$\%$ respectively.\\
2491
2492${H_0}$: the population mean ${\mu}$ poverty headcount ratio in Ethiopian household lie between 38.7$\%$ - 29.6$\%$.\\
2493${H_a}$: ${\neq}{\mu}$.\\
2494Considering the law of lager numbers and the central limit theorem, we assumed that as $n{\rightarrow}{\infty},$
2495$({\widehat{M_0}}-M_0)$  $\overrightarrow{d}$ 
2496Normal $(0, \, \dfrac{\delta^2_0}{n}),$ where $\delta^2_0={\in[\widehat{c_i}}(k)-M_0]^2$ is the population variance of $M_0$ (Alkire et. al.,  2015: p-244).\\
2497
2498Likewise, suppose  give  two regions, say Oromia and Tigray. The population achievement matrix are denoted by $X_1$ and $X_2$ respectively. We seek to test the null hypothesis $H_0$: $M_{0.1}-M_{0.2}=0,$ which implies that the poverty in region Oromia is not significantly different from the poverty in region Tigray. With regard to any of the three alternatives: i) $H_a : M_{0.1}-M_{0.2}{\neq}0,$ which means that one of the two regions is significantly poorer than the other; or ii) $H_a : M_{0.1}-M_{0.2}>0, $  which means that region Oromia is significantly poorer than region Tigray; or iii) $H_a : M_{0.1}-M_{0.2}<0,$ which mean that region Tigray is significantly poorer than region Oromia and so on. \\
2499For the first alternative, we need to conduct a two-tailed test, while for the other two alternatives, we should perform a one-tailed test. \\
2500
2501Finally, given that a sample of $\widehat{X_1}$ of size $n_1$ is collected from $X_1$ and a sample of $\widehat{X_2}$ of size $n_2$ is collected from $X_2$, where samples $\widehat{X_1}$ and $\widehat{X_2}$ are presumed to have been drawn independently of each other. We denote the estimated Adjusted Headcount Ratio from the samples by $\widehat{M_{0.1}}$ and $\widehat{M_{0.2}}$ respectively.\\
2502By the law of large numbers and the central limit theorem, 
2503$(\widehat{M_{0.1}}-\widehat{M_{0.2}} \overrightarrow{d})$ Normal $(0, \, \dfrac{\delta^2_{0.1}}{n_1})$ and $(\widehat{M_{0.1}}-\widehat{M_{0.2}} \overrightarrow{d})$ Normal $(0, \, \dfrac{\delta^2_{0.2}}{n_2}).$ The difference of two normal distributions is a normal distribution as well. Thus
2504
2505\begin{equation}
2506((\widehat{M_{0.1}}-\widehat{M_{0.2}}-(M_{0.1}-M_{0.2}))\overrightarrow{d} Normal (0, {{\delta^2_{{0.1}-{0.2}}}}),
2507\end{equation}
2508where ${{\delta^2_{{0.1}-{0.2}}}}=\dfrac{\delta^2_{0.1}}{n_1}+\dfrac{\delta^2_{0.2}}{n_2}.$ Note that, as we presumed independent samples, the covariance between the two Adjusted Headcount Ratios is zero. Thus, the standard error of $\widehat{M_{0.1}}-\widehat{M_{0.2}},$ designated by $se_{\widehat{M_{{0.1}-{0.2}}}},$ can be estimated by the following equation:
2509\begin{equation}
2510se_{\widehat{M_{{0.1}-{0.2}}}}=\sqrt{se^2_{M_{0.1}}+se^2_{M_{0.2}}}
2511\end{equation}
2512where $se^2_{M_{0.1}}$ is the variance of $\widehat{M_{0.1}}$ and $se^2_{M_{0.2}}$ is the variance of $\widehat{M_{0.2}}$
2513%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2514
2515%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5
2516\subsection{Components of the Ethiopian National and Regional  MPI: Dimensions, Indicators, Deprivation Cutoffs and Weights}
2517In this study, the multidimensional poverty index is constructed based the Global Multidimensional Poverty Index (MPI) adopting the procedures of indicated in the Human Development Report 2010 (UNDP, 2010, Alkire and Santos, 2014) under the family of AF Method. However, necessary adjustment were in indicators and their respective weights to fit to our data sources, the variables thereof and to better reflect the Ethiopian context. 
2518By and large, the unit of unit of analysis is the individual, however, the unit of identification, is a blend of an individual and a household.\\
2519
2520As a matter of fact, the Ethiopian societal value encourages people to live together even when they are over 18 years old and do share common resources. Thus, it is hard in some cases to observe individual deprivations (for example in the living standard categories which are utilized jointly at the household level). In the event that it is difficult to disentangle individual and household deprivations, it is cogent to discern penury and destitution at the household level.  In such scenario, the underlined message is that a deprived household in an indicator implies that its members are presumed to be deprived in that indicator as well. By the same token, if a household is multidimensionally destitute, all other members are perceived to be multidimensionally poor. \\
2521
2522\subsubsection{Normative Decisions: Choice of Indicators}  
2523 
2524The three domains namely, health, education and living standards are weighted equally. 
2525The study of multidimensional poverty in Ethiopia are based on the AF method and three waves of  panel data for more than 5000 household in Ethiopia was collected in 2011-2012, 2013-2014, and 2015-2016. We further analyzed and identified the poor by sub-group decomposition by gender, age-group, location,region and so on . Plus, we scrutinize the break down of MPI by domains and indicators.\\
2526
2527Based on the practices that we learned in our review of  wealth of literature, it is observed that each domains are equally weighted. Hence, the nutrition indicator takes 50$\%$ of the health domain's weight which is (16.7$\%$ as it is believed to be the most critical indicator of this dimension.  The other three indicators in this dimension share the remaining weight equally. Thus, the distribution of the weight for each indicators of the health dimension is 5.6$\%$.\\
2528
2529Correspondingly, the three indicators in the education dimension are equally weighted as it is argued that all indicators here are equally important. Therefor, the education dimension's weight is equally shared among the three indicators and each indicators is weighted by   $w=11.1\%$\\
2530
2531Finally, the indicators in the living standard dimensions  are articulated based the national and global MPI reviewed previous research works. The eight indicators in this domain as well are equally weighted and the weight for each indicator is 4.1$\%$.\\
2532
2533
2534%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2535\subsection{ Access, Affordable and Quality of Health Services in Ethiopia}
2536Taking a birds eye-view at some health indicators/proxies, wealth of literature have documented that Ethiopian Households practicing open defecation, rural(percentage of rural population (2012) in Ethiopia was 44.3 percent.\\
2537To illuminate the above statement, households practicing open defecation refers to the percentage of the population defecating in the open, such as in the fields, forest, bushes, in other open bodies or disposed with solid waste which is summarized in figure 8 below (WHO/UNICEF  Joint Monitoring Program(JMP) for Water Supply and Sanitation, 2012). Figure 8 below reports the percentage of Ethiopian households practicing open defecation by comparing with  selected Sub Saharan African Countries with developing economies. Considering the above stated proxy of health indicator, close to one in two Ethiopian households are exposed to lack of hygiene related diseases although Ethiopia is registering tremendous achievements in the health sector recently.
2538\begin{landscape}
2539\begin{figure}
2540\caption{Ethiopian Rural Residents Practicing Open Defecation (Percentage of Rural Population by 2012}
2541\label{Figure8}
2542\includegraphics [scale=0.45]{EthDefecation1graph.png}%{plot}
2543\end{figure}
2544\end{landscape}
2545Considering other indicators of affordable and quality health in Ethiopia, we shed light on health expenditure per capita (current U.S. $\$$) (2014).\\
2546Total Health expenditure is the sum of public and private  expenditure  as a ratio of total population. This proxy indicator too covers the provision of health services (preventive and curative), family planning activities, nutrition activities, and emergency aid designated for health but does not include provision of water sanitation. Source: (World Health Organization Global Health Expenditure database,2014). \\
2547
2548As can be seen from  figure 9 below,  Ethiopian health expenditure per capita is very low as compared with selected Sub Saharan African Countries. \\
2549
2550\begin{figure}
2551\caption{Comparison of Ethiopia's Health Expenditure per capita(current U.S. $\$$, 2014) with Selected Sub Saharan African Countries}
2552\label{Figure9}
2553\includegraphics [scale=0.45]{EthHealthExpenditure1graph.png}
2554\end{figure}
2555Another critical indicator of poor services is the mortality rate (indicated in figure 10 below), infant (per 1,000 live birth) 2014. Infant mortality is the number of infants dying before reaching one year of age, per 1,000 live birth in a given year. This indicator is the lowest even as compared with other least developing countries in the Sub-Saharan Africa (UNICEF, WHO, World Bank, UN DESA Population Division, 2014).
2556"The infant mortality rate is 59 deaths per 1000 live births, while the overall under-five mortality rate is 88 per 1000 live births. A total of 67$\%$ of all deaths of children aged under 5 years in Ethiopia take place before a child's first birthday."
2557
2558The Ethiopian Demographic and Health Survey (EDHS)Report shows that there has been a steady reduction in the  trends in childhood mortality in Ethiopia since the 2000 EDHS survey. The following quotation summarizes the facts pertinent to infant and child under 5 year old death rate: 
2559\begin{quotation}
2560... child, and under-5 mortality over the last 16 years. For example, under-5 mortality rates for the 5 years preceding the survey declined from 166 deaths per 1,000 live births to 123 deaths per 1,000 live births in 2005, to reach 67 deaths per 1,000 live births in 2016. Similarly, infant mortality decreased from 97 deaths per 1,000 live births, to 77 deaths per 1,000 live births, and to 48 deaths per 1,000 live births in the same period EDHS, 2016).
2561\end{quotation} 
2562\!\begin{figure}
2563\caption{Ethiopian Households Infant Death rate, crude (per 1,000 people, 204 as Compared with SSA Countries}
2564\label{Figure10}
2565\includegraphics [scale=0.45]{EthInfantMortality1graph.png}%{plot}
2566%\label{Figure10}
2567\end{figure}
2568%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2569\newpage
2570
2571\subsubsection{The Health Dimension} 
2572
2573(Sources: WORLD HEALTH RANKINGS: LIVE LONGER LIVE BETTER 2017, accessed on April 21, 2017; for details refer to www.worldlifeexpectancy.com/ethiopia-life-expectancy).\\
2574
2575Health is core to human ecstasy and well-being as it makes a crucial contribution to a country's economic progress,a cohesive and self-reliant society, communities live longer, people are productive, save for internationalism welling, protect their environment,have positive impact on development, prosperity, security  and stability, poverty reduction and saw the hopeful seeds for the future. This is also in alignment to the Goal 3 (Ensure healthy lives) of the Sustainable  Development Goals.\\
2576
2577Conversely, deprivation of this fundamental component results in lack of all what we mentioned above, self esteem, decent life and so forth. Therefor, health is the pillars of the overlapping deprivation in the AF multidimensional measurement of poverty.\\
2578\subsubsection{Status of Malnutrition in Ethiopia}
2579
2580There are two basic types of malnutrition/underestimation. The first and most important is protein-energy malnutrition (PEM). It is basically a lack of calories and proteins. Protein-energy malnutrition is the more lethal form of malnutrition/hunger that leads to a principal type of growth failure. Besides, the two type of acute malnutrition are wasting, (also called marasmus)or nutritional edema, (also called kwashiorkor). Wasting(children who are too thin for their height)  is characterized by rapid weight loss and in its severe form can cad to death. Nutritional edema is cased by insufficient protein in the diet.\\
2581
2582Where as stunting(children who are too short for their age) is a slow, cumulative process and is caused by insufficient intake of some nutrients that is believed  to affect 161 million children world wide. It is also reported that nearly half (3 million) of the deaths in children under 5 year of old are attributed to underestimation. Also, it puts children at a greater risk of dying from common infections, increase the frequency and severity of such infections, and contribute to delayed recovery.
2583To sum up, more than 161 million kids are too short for their age; while around 51 million are too thin for their height. On the contrary, over 42 million kids are overweight in the lens of global perspective (Global Nutrition Report, 2015). \\
2584
2585The Causes of Chronic Malnutrition Crisis in Ethiopia has been under the world spotlight for many years. According to the latest WHO data published in may 2014, Malnutrition Deaths in Ethiopia reached 28,560 or 4.75$\%$ of total deaths. The age adjusted death rate was 48.19 per 100,000 of population ranks Ethiopia number 14 in the world where as percentage of children under 5 who are stunted are more than 40$\%$ (WHO, 2010-2016, UNICEF, 2016).
2586\begin{quotation}
2587Malnutrition in the Ethiopian context has been described as a long-term year round phenomenon due to chronic inadequacies in food intake combined with high levels of illness (National Nutrition Policy-Draft 2003)
2588\end{quotation}
2589The malnutrition cataclysm in Ethiopia is attributed to several reasons, immature agricultural systems, uneducated and overused caregivers, and insufficient health care systems and sanitation. The lack of agricultural infrastructure and poor farming techniques have led to critically low crop yields. Poor education for female caregivers and societal inequalities over caregiver roles has led to especially malnourished youth who play into a disturbing viscous cycle. Compounding these issues, the health care and sanitation systems of Ethiopia have thus far been unable to cope with the explosion of malnourished cases. \\
2590
2591In our data sources from wave I, II and III of the Ethiopia Socioeconomic Survey; panel done in 2011-2012, 2013-2014, and 2015-2016; save for children under five year old, the height and weight of household members are not measured in the waves and hence cannot be included in the malnutrition indicator in the health domain. To put it differently, the malnutrition indicator and follow up analysis in the health dimension includes children under 59 months or five years old. \\
2592
2593Furthermore, height and weight were garnered from all children aged 6-83 months.  These data were employed to compute the three regularly used child nutritional status indicators. The three frequently practiced anthropocentric indicators to measure child nutritional status include height-for-age, weight-for-age, and weight-for-height. Measured by the three indices, children with score of below minus two standard deviations (-2SD) from the reference population are considered as moderately malnourished (for details see the WHO 2006). The main indicator in the health domain, nutrition of children below 5 years old (an individual being considered as deprived  if any child under 5 years old with nutritional information is undernourished) \\
2594
2595Other indicators of daily health functioning are made use of as other indicators of the heath domain as above mentioned the survey has no information on child mortality. In such cases, an individual is deprived in the health functioning if s/he has difficulties with activities of daily living in seeing, hearing, walking, remembering, keeping self-care, communication and absentee from their daily activities. Similarly, respondents were inquired "for how many days were you absent from your usual activity due to the health problems?" Referring to this question a household is deprived if any member of the household has any disease or injury and was absent for 4 weeks as a result of the injury or disease related problem. \\
2596
2597Considering the difficulties households/or household members face challenges in executing/performing the activities of the daily life; an individual/a household member is deprived if s/he has difficulties with activities of daily living in seeing, hearing, walking, remembering, keeping self-care, communication and absentee from their daily activities. \\
2598
2599In the like manner, respondents were asked  "how many meals, including breakfast are taken on average per day for adults and children?" Pertinent to this question,  an individual is considered to be deprived if meals were served everyday including  breakfast are taken  on average less than three time per day for adults  15 years and above,  and less than four times for children less than 15 years.\\
2600In the last indicator of the health domain, respondents were asked "have you faced any health problems during the last 2  months?", related to this question, a household is deprived if any member of  is deprived in self reported health if any member of the household faces health problems in the last 2 months.\\
2601
2602Correspondingly, the 2016 Ethiopia Demographic and Health Survey (EDHS) released by Central Statistics Agency (CSA) showed that infant mortality has fallen from 97 deaths per 1,000 live births in 2000 to 48 in 2016. By the same token, children under 5 years old  mortality rate has markedly decline from 1666 to 67 deaths per 1,000 live births. Elaborating on the findings of the survey, it was reported that a significant reduction was also was witnessed in fertility rate from 5.5 children per women in 2000 to 4.6 children in 2016. This demonstrates a decline of 0.9 children. Moreover, 35 percent of married women were fond to be using a modern method of family planning; which is a five fold increase from 6 percent in 2000. 
2603Source: CSA website, "Infant, Under-5 Mortality Rate Registers Significant Decline" online website accessed on 15 August 2017.  
2604%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2605\subsubsection{Education Dimension}
2606\begin{quotation}
2607
2608Education can transform lives, but today millions of young people grow up without acquiring basic reading, writing or math skills. This is not only a wasted opportunity, but a great injustice to the children who most need a good education to succeed in life. Fixing this problem will be possible when learning for all becomes a priority for every country.  For example, Liberia, Papua New Guinea and Tonga improved early grade reading substantially in a very short time. Between 2009 and 2015, Peru achieved some of the fastest growth in learning outcomes due to concerted policy action.  We hope to see many similar success stories. The 2018 World Development Report on education lays out the concrete steps that can help solve our learning crisis.\\
2609Education has long been critical to human welfare, but it is even more so in a time of rapid economic and social change. The best way to equip children and youth for the future is to place their learning at the center. The 2018 World Development Report (WDR) explores four main themes: 1) education's promise; 2) the need to shine a light on learning; 3) how to make schools work for learners; and 4) how to make systems work for learning.
2610Jim Kim President at the World Bank(2017: website accessed September 27 2017. 
2611\end{quotation}
2612
2613Education is a basic human right and civil right. Specially  literacy is a human right; that being able to read and write is personally empowering and, in a world that relies more and more on technology, simply necessary.\\
2614It results in skilled labor and manpower with physical or mental 
2615endeavors or active engagement toward the production of goods and services. \\
2616
2617A country having proper manpower planning, optimal  utilization of its labor forces, huge investment on its people, far ranging, continuous and  dynamic capacity building on ts human development, ceaseless improvement in productivity, sustainable and inclusive development,... has the power to ensure societal transformation and reduce/eradicate poverty, inequality, unemployment and other related social ills. People in those countries  have the freedom where they want to be and what to do(Sen, 1981), freedom to choose and decide over their will, low disparity of opportunities, outcomes and autonomous, adequate and standard education and training facilities. \\
2618
2619Societies who are deprived of those key factors to development are unable  to unlock the poverty trap and live under the set of factors or spirals of sets of  events by which poverty, once occurred is likely to continue if there is no big push and multifaceted intervention programs to unlock this poverty trap.
2620We briefly analyzed the status of education in Ethiopia based on the available data continent for Interactive Stat Planet analysis as shown below:\\
2621
2622\begin{figure}
2623\caption{Enrollment in Tertiary Education Institutions per 100,000 Inhabitants (both sexes )in Ethiopia Until 2014}
2624\label{Figure11}
2625\includegraphics [scale=0.45]{EthEducTeriEnrol1graph.png}%{plot}
2626\end{figure}
2627Figure 11 below,  portrays enrollment in tertiary education per 100,000 inhabitants (both sexes). It is calculated by dividing the total number of students  enrolled in tertiary education in a given academic year by the country's population and multiplying the result by 100,000. This indicator shows the general level of participation in tertiary education by indicating the proportion of (or density) of students within a country's population.\\
2628
2629Despite Ethiopia's strenuous effort to expand access to education, enrollment  in tertiary education per 100, 000 inhabitants, both sexes is 781 which is very low. Even if our interest in this study is enrollment in primary and pre-secondary education, we considered enrollment in tertiary education due to the similarity of the database. However, in the other case the database (in the interactive Statplanet) is categorized as enrollment in private and public and it is just raw data that may lead to wrong conclusions. We expect this could be due to the national policy differences on education where education is free in the Ethiopian case, yet this may not be true in the other countries in comparison. \\
2630\newpage
2631Considering female enrollment in upper secondary education in Ethiopia, female (number) (2012), it has generally registered a resounding triumph in education. Yet, there are concerning observations regarding female participation; particularly the total numbers of female students enrolled in (public and private) uppers secondary education institutions (UNESCO Institute for Statistics, 2012).
2632\begin{figure}
2633\caption{Female Enrollment in Upper Secondary in Ethiopia, until 2012 in (number)}
2634\label{Figure12}
2635\begin{center}
2636\includegraphics [scale=0.45]{EthFemaleEnrol1graph.png}%{plot}
2637\end{center}
2638\end{figure}
2639With regard to expenditure on education as percentage of total government expenditure (percentage) (2012); total general (local, regional and central) government  expenditure on education (current, capital and transfers), expressed as a percentage of total general government expenditure on all sectors (including health, education, social services etc.). It includes expenditures funded by transfer from international sources to government. Public expenditure includes spending by local/municipal, regional and national governments(excluding households contributions)on educational institutions (both public and private), educational administration,  and subsidies for private entities (students/households and other private entities)(UNESCO, 2012)
2640\begin{figure}
2641\caption{Expenditure on Education as a Percentage of Total Government Expenditure Until 2012}
2642\label{Figure13}
2643\begin{center}
2644\includegraphics [scale=0.45]{EthExpendEduc1graph.png}%{plot}
2645\end{center}
2646\end{figure}
2647As can be seen from figure 13 above, percentage expenditure as a percentage of total government in Ethiopia  is very large as compared to SSA countries. In fact, education takes the third/or fourth largest share in Ethiopia's Federal budget distribution for the last decade. As a result, an other crucial overlapping deprivation domain of the MPI is education.  What is more, this domain is directly related  to Goal 4 (Ensure inclusive and equitable quality education) and implicitly related Goal 8 (Promote... economic growth, full and productive employment and decent work for all) of the Sustainable Development Goals (2015-2030).\\
2648For this reason, education is a fundamental component well-being analysis under the AF methodology as it is the main supply-side vicious circle slack. \\
2649
2650The education indicators and their respective cut-offs for this study; therefor, are year of schooling, child school attendance, and school enrollment. Under the first scenario a household member is non-deprived if any member of the household older than 15 has completed five years of schooling. While a household is deprived if no household member older than 15 has completed five years of schooling. In other words, if we observe that atleast one member of the household with five or more years of education then, regardless of other members with missing data, we classify the household as non-deprived. Besides, in the ESS panel data, a member of the household was asked "are you currently attending school?"; regarding school-age children attendance, a household is non-deprived if any school-age (7-13) child is attending school. A household is deprived if any school-aged(7-13) child is not attending school. \\
2651
2652Finally, a household/member was asked "which grade are you attending?"; referring to third and last indicator under the education, a household is non-deprived if any post primary school age adolescent (14-25) is attending post primary school. Where as a household is deprived if any post primary school age adolescent (14-25) is not attending post primary school. \\
2653%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2654\subsubsection{Living Standard Dimension}:
2655
2656Having sustainable and continuously improved access to adequate food, housing, water, sanitation services and energy are key to basic human existence not to mention a productive and economically active life. Besides, ownership and entitlements of fixed and productive assets play critical role by stimulating development efforts in the education, health and other social sectors aiming  at reducing/ending overlapping deprivations in particular, poverty in general, inequality and unemployment in the developing world. Consequently, these variables are considered as the main factors influencing  multiple destitution and indicators of MPI as these factors tremendously impact development issues in developing countries in general and in Ethiopia in particular.\\
2657
2658Unlike to the conventional measurements of living standard, direct(income and consumption)  and  (such as GDP, Real/average GDP per capita, Index Human Development (HDI), Satisfaction With Life Index, Happy Planet Index, GDP-Purchasing Power Parity (PPP), Index of Human Poverty (HPI), and other indirect measures of living standards and inequality); MPI employs indirect indicators measurements in the  living standards domain. By and large, these measure are similar or proxies to the measurements targets in the Sustainable Development Goals like Goal  1(End Poverty in all its forms), Goal 2( End hunger, achieve food security), Goal 6(Ensure availability... of improved water, sanitation... for all), Goal 7 (Ensure access to ... improved/modern energy for all,  Goal 11(make... human habitat inclusive, safe...) and o forth.\\
2659  
2660  \subsubsection{Sources of Energy Supply in Ethiopia}
2661    
2662Having a modern energy unlocks access to improved health care, improved education, improved economic opportunities, sustainable growth and, even, longer life. Not having it is a major constraint on their social, economic development, ending poverty, inequality, unemployment, social transformation and emancipation. 
2663
2664Ethiopia, like other Sub-Saharan countries is rich in energy but poor in energy supply. It's currently utilizing 6.5$\%$ of its potential(60,000 MW) (the International Energy Agency's (IEA) Africa Energy Outlook, 2014). Another big challenge in Ethiopia is the Energy poverty. Ethiopia's electricity consumption/population is 0.07 MWh/capita which is lower than its neighbors Sudan (electricity consumption/population is 0.25/capita) and Kenya (electricity consumption/population is .17 MWh/capita) (IEA, 2014). The irony is that both this countries are buying electric supply from Ethiopia.  \\
2665
2666Ethiopia has plentiful renewable energy resources and has a potential to develop over 60,000 megawatts (MW) of electric power from hydroelectric, wind, solar and geothermal sources. Currently the country has developed approximately 4,000 MW of installed generation capacity which is short of the demand to serve a population of over 95 million people. GTP II has a new target to increase generation capacity to over 17,000 MW by 2020, with an overall potential of 35,000 MW by 2037.\\
2667
2668Ethiopia energy supply is mainly based on biomass (90 $\%$ of households, 70 $\%$  of industries, and 94 $\%$ of service industries and enterprises consume biomass as energy sources. Besides, it's reported that households account for 88 $\%$, industry 4$\%$, transport 3 $\%$ and others 5 $\%$ of total energy consumption. 
2669
2670Dubbed "the water tower of Africa", Ethiopia has long sought to harness the power of the rivers that tumble from its highlands. Flagship dam projects were central to the modernization plans drawn up by the Italian administration of 1936-1941 and by the former emperor, Haile Selassie, in the 1960s. Gibe III is the latest in a series being built along the Omo River by the government, which is also constructing what will be the largest-ever dam in Africa when it opens, in theory, next year: the Grand Ethiopian Renaissance Dam on the Blue Nile.\\
2671 
2672Together these projects are intended to turn Ethiopia, which has scarce minerals but enormous hydro-power potential, into a renewable-energy exporter. Gibe III alone is expected to generate as much electricity as currently produced by the whole of neighboring Kenya, which has enthusiastically signed up to buy some of its power. The export earnings will help to plug Ethiopia's gaping current-account deficit, while the cheap power will provide a timely fillip to its nascent manufacturing sector.\\
2673\begin{quotation}
2674Energy is arguably one of the major challenges the world faces today,touching all aspects of our lives. For those
2675living in extreme poverty, a lack of access to modern energy services dramatically affects health, limits opportunities and widens the gap between the haves and the poor. The vulnerability of the poor is only worsened  with recent challenges from climate change,a global financial crisis,and volatile energy prices (UNDP, 2009).
2676\end{quotation}
2677\begin{figure}
2678\caption{Share of Total Primary Energy Supply in Ethiopia Until 2014}
2679\label{Figure14}
2680\begin{Large}
2681\includegraphics [scale=0.60]{ETHIOPIA4_Energy_Supply.pdf}%
2682\end{Large}
2683\end{figure}
2684
2685The dominant Ethiopian Energy supply based on n biomass continues to take the lion's share followed by oil since 1970 to present as it can be seen in the graph below.  
2686\begin{figure}
2687\caption{Components of Total Primary Energy Supply in Ethiopia}
2688\label{Figure15}
2689\begin{Large}
2690\includegraphics [scale=0.60]{ETHIOPIA5_Energy_Supply.pdf}%
2691\end{Large}
2692\end{figure}
2693\subsubsection{Access to Improved Electricity}:
2694
2695Here, we succinctly  discussed  Access to electricity, rural (percentage of  rural population) a trend analysis based on the data base (2012).\\
2696Abject poverty and lack of access to improved energy sources attributed to the heavy dependence of the population on traditional fuels(EIA, 2015?). According to the World Bank, Sustainable Energy for all (SE4All, 2012)database, Ethiopian rural households access to electricity, percentage of rural population with access to electricity; though it shows a rapid growth it is still extremely low (less than 10 $\%$)Source Organization: World Bank, Sustainable Energy for all (SE4ALL) database from World Bank, Global Electrification database.  \\
2697
2698
2699Reviewing Ethiopian rural households access to electricity, percentage of rural population with access to electricity; though it shows a speedy growth, it is still extremely low. As it can be seen and inferred from the World Bank, Sustainable Energy for all (SE4ALL) database,(2012)-the World Bank, Global Electrification database accessed on Dec. 4/2016, we produced (EthRElectric2trendgraph) time series trend analysis that reveals the Ethiopian rural household access to electricity. 
2700\begin{figure}
2701\caption{Rural Households in Ethiopia Access to Improved Electricity(percentage of Rural Population): Trend Analysis}
2702\label{Figure16}
2703\begin{center}
2704\includegraphics [scale=0.45]{EthRElectric2Trendgraph.png}
2705\end{center}
2706\end{figure}
2707Considering Ethiopian households trend of access to improved electricity, merely, 7.6 of Rural(percentage rural population) Ethiopian have access to electricity despite its huge sources of renewable energy potential. Ethiopians are among those with the lowest access to electricity in  the Sub-Saharan Africa region. Although with a growing trend,  access to electricity (percentage of population) is around 26.6. \\
2708 \begin{figure}
2709 \caption{Ethiopian Rural Households  Access to Improved Electricity(percentage of Rural Population): Static Analysis}
2710 \label{Figure17}
2711  \begin{center}
2712 \includegraphics [scale=0.45]{ETHElectric1Graph.png}%{plot}
2713\end{center}
2714\end{figure}
2715Looking at figure 16 and 17 above, it can be observed that both figures report  Ethiopian rural households access to improved electricity as the percentage of rural Population. In can further be noted that while figure 16 demonstrates time series analysis of Ethiopian rural households access to improved electricity; figure 17 presents access to electricity at a point in time. \\
2716Based on the Ethiopia Socioeconomic Survey (2011-2012, 2013-2014 and 2015-2016), the MPI living standard indicators in this study are:
2717
2718\begin{enumerate}
2719\item Deprived in Electricity(SDG indicator Goal 7): 
2720Other Sources of Energy: Without use of power sources economic development is unthinkable.\\
2721Household in the (ESS)surveys were asked "What is the main source of light for the household?" For the purpose of identification, a household is non-deprived in electricity if it has electricity meter-private, electricity meter-shared, and solar energy. But then, a household is deprived in electricity if in electricity if it gets electricity from  electricity from generator, electrical batter, Lantern, Light from dry cell with switch, Kerosene light lamp (imported), Kerosene lamp (local kuraz), Candle/wax, Fire wood, and Other specify\\
2722
2723\item COOKING FUEL (SDG indicator Goal 7):
2724In the ESS panel data sets, households were asked "What is the main source of cooking fuel?"; adopting the UNDP indicators and based on the existing literature,  a household is non-deprived in cooking fuels if uses Kerosene, Butane / gas, Electricity, Solar energy, and Bio gas.  Whereas a household is DEPRIVED IN COOKING FUEL if: it uses  Collecting fire wood, Purchase fire wood, Charcoal, Crop residue / leaves, Dung / manure, Sawdust and Others.\\
2725
2726\item Ethiopia: Water Sanitation and Hygiene Indicators (WASH) Services (SGD indicator 2 and 6):\\
2727There is persistent and critical water $\&$ sanitation crisis in Ethiopia. Country wide, 43 percent of Ethiopians lack access to safe water and 72 percent lack access to improved sanitation.  Recurring droughts result in famine, food shortages, and water-related diseases, as people are forced to rely heavily on contaminated or stagnant water sources(UNICEF, 2016?). \\
2728
2729In the ESS (2011-2012, 2013-2014, and 2015-2016) respondents were asked "What type of bathing facilities does the housing unit have?"
2730Based on the literature and practical situations on the ground,a household is non-deprived if it uses Bathtub/private, Bathtub/shared, Shower/private, and Shower/shared for bathing. However, a household/ household is deprived in bathing facilities if it uses a room reserved for bathing/private, a room reserved for bathing/shared, a fixed place for bathing, and others.\\
2731\item Access to improved Water:\\
2732The most critical and pervasive living standard indicator is access to improved water.\\
2733
2734According to the WHO/UNICEF Joint Monitoring Programme (JMP) (2010), access to an improved Water Sources refers to the percentage of the population using an improved drinking water sources that includes piped water om premises (piped household water connections located inside the users dwelling, plot or yard),
2735and other improved drinking water sources (public taps or standpipes, tube wells or boreholes, protected dug wells, protected springs, and rainwater collection.\\
2736
2737In the ESS, respondents were inquired "What is the main source of drinking water in the rainy and dry season?" Based on exiting literature, UNDP definition, global MPI, and facts pertinent to the society, a household is non deprived if it has tap inside the house, private tap in the compound, public tap, shared tap in the compound, communal tap outside compound, protected well / spring, private, protected well / spring, shared and others. In contrast, a household is deprived if: it gets water from Water from kiosk/retailer, unprotected well / spring, river / lake / pound. \\
2738
2739To elucidate the load, the time investment and cumbersome nature of collecting woods so as to supply their energy consumption from local sources and time spent on fetching water let's consider some stunning cases below:
2740\begin{quotation}
2741
2742"Hiwot used to spend up to six hours a day collecting one Jerrycan of water for her family. And Hiwot's story of balancing farm work, raising her child and collecting water for her family is not uncommon in her small Ethiopian community. 22-year-old Haymanot, 19-year-old Birey, 21-year-old Frewoyini ... each of these women shared the same story of putting aside education to focus on chores each dependent on the task of acquiring water.\\
2743
2744As 26-year-old Muzey's bio states, "It upsets her to give dirty water to her children, but she has no other choice."\\
2745
2746The outside world would likely know nothing of these human stories, if not for charity: water's new "Someone Like You," campaign, which was fully released today in conjunction with World Water Day. (Charity:water, 2017; accessed April 14, 2017)"\\
2747\end{quotation}
2748
2749%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2750\subsubsection{Improved Housing Services}
2751
2752According to the UN-Habitant Report (2013), one-third of the urban population in developing countries resides in slums with slum prevalence the highest in Sub-Saharan Africa (62$\%$).  It also indicates, the slum prevalence in Ethiopia is 80$\%$ being the highest in the world.  Majority of the  Ethiopia population is you median age being 18.9, coupled with a 2.45$\%$ annual population growth rate and more than 4.5$\%$ urbanization growth rate is putting considerable pressure on the demand for housing.\\
2753
2754Ethiopia, one of the least urbanized countries in Africa, is expecting large increase in its share of urban population from current around 20 $\%$ to 40 $\%$ by 2030. Considering the inability of many rural farms to provide a subsistence livelihood, high urbanization rates are inevitable. \\
2755
2756However, the sluggish nature of socioeconomic and infrastructural development and rapid population surge in most developing countries makes effective and efficient service delivery difficult for governments, to satisfy the need of their residents. Shelter/housing is the key basic necessities for human survival and key indicator of well-being. The current housing deficit in urban areas is estimated between 1 to 1.5 million. On the other hand, its supply falls in short far behind for long period and becomes pervasive and critical challenge in most cities of developing countries in general and Ethiopian cities in particular. Hence,in regional states and capital cities of Ethiopia, providing affordable housing is one of the key priorities now  \\
2757
2758Why do we consider improved walls, floor and roof of a house as best indicators of decent as safe home? As an illustration let’s consider what it means by households deprivation in improved and quality floor. People with lack of access to improved floor, in other words, who live in a 'dirty' floor (if the floor of the house is made from mud / dung, reed / bamboo and wood planks and the situation of  the house is too old and dilapidated and too narrow to accommodate families, where the health and dignity of families is compromised).\\
2759
2760In such scenario, children are particularly the most affected group as their growth cycle is attached to the floor and are easily affected by foreign harmful bacteria, virus, and germs and so on that live in the earth floor. Dust is kicked up, spills and puddles stick around, and the floors of peoples' homes become a breeding ground for mosquitoes, parasites, and all of the disease that comes with them. Nevertheless concrete floors are non-affordable and shortage in supply for poor Ethiopian families. \\
2761
2762Consequently, health members of the household living in such floor will deteriorate owing to influenza and pneumonia, diarrhea diseases, asthma and so forth. Treating this takes the scanty income of the household leading to borrowing (if there exist any sources of accessible and affordable credit and saving scheme) and deteriorating the existing multiple overlapping deprivations and poverty. This in turn worsens the poverty situation of the household and drags down to the bottom of poverty trap. Targeting this will assist Ethiopian households to possess a healthy and sustainable homes thereby improving their well-being. 
2763Each year diarrhoea kills around 760 000 children under five years old. Globally, there are nearly 1.7 billion cases of diarrhoeal disease every year.  It is the second leading cause of death in children and a leading cause of malnutrition in children under five years old.  \\
2764
2765Diarrhea is one of the leading (fourth) causes of childhood death in Ethiopia. According to the latest WHO data published in may 2014 diarrhoeal diseases Deaths in Ethiopia reached 41,287 or 6.87$\%$ of total deaths. The age adjusted Death Rate is 49.54 per 100,000 of population ranks Ethiopia number 35 in the world. In practical/experimental study in Rwanda, healthy floors have been shown to reduce the incidence of childhood diarrhea by 49$\%$ and parasitic infections by 78$\%$. Dirt floors have been shown to harbor parasites, bacteria, and pathogens dangerous to children and the adults living in these homes (EarthEnable, 2017; website accessed on April 21, 2017).  By the same token, similar explanation can be give to the contribution of improved walls and roof of the main dwelling to the decent living and its multiplicative effect to the poor households who are deprived to it.\\
2766
2767A crucial indicator for deprivation and status of the poor in the living standards is indicators of deprivations in the ROOF, WALL, and FLOOR of the main dwelling. In the ESS, respondents were asked the following interrelated questions: "The 'ROOF', 'WALL', 'FLOOR' of the main 'DWELLING' are predominantly made of what material?" In the ESS, respondents were asked "How many rooms (excluding the kitchen, toilet and bathroom) does the household own?"; considering on the standard practices and literature, a household is non-deprived in the 'ROOF' of the main dwelling if the roof of the house is made from  Corrugated iron sheet, Concrete / Cement, Asbestos, Bricks and others.\\
2768
2769On the other hand,  household is deprived in the  'ROOF' of the main dwelling if the roof of the main dwelling is made from Thatch, Wood and mud, Reed / bamboo, and Plastic canvas. Besides, a household is non-deprived in the 'FLOOR' of the main dwelling if the floor of the house is made from  Parquet of polished wood, Cement screed, Plastic tiles, Cement tiles, Brick Tiles, Ceramic / marble tiles, and others. Yet, a household is deprived in the  'FLOOR' of the main dwelling if the floor of the house is made from mud / dung, reed / bamboo and wood planks.\\
2770
2771A household is non-deprive in the walls of the main dwelling if the walls are made from stone and mud, stone and cement, blocks, plastered with cement, and others bricks, mud bricks (traditional) and others. Nonetheless, a household deprived in the walls of the main dwelling, if the walls of the house are made from Wood and mud, Wood and thatch, Wood only, Stone only, Blocks/unplastered and Reed / bamboo.
2772
2773\item Overcrowding: \\
2774Overcrowding and lack of tenure security are pressing issues
2775in Ethiopia's major towns, and are highly correlated with poverty
2776A household is non-deprived if less than or equal to 3  people live in a room. But, a household deprived in Overcrowding if more than 3 people live per room.
2777\item Households' Fixed, Productive and Other Assets:\\
2778Asset ownership is thought to be a key indicator of welfare and asset acquisition is often believed o be a manifestation of improving living standards of households. Depletion of assets, on the contrary, entails a shrinking households wealth and hence a decline in welfare. In the ESS, information on ownership of some selected key indicators was collected from households. In so-doing, household assets were itemized as modern and traditional farm inputs, furniture, Electronics and electronic related, personal items, and others assets.  
2779
2780\item   Households' Fixed, Productive and Other Assets Possession:\\
2781Control and ownership of assets within Ethiopian households is not only reckoned as the most essential indicators of household well-being but also predicts the bargaining power over a numbers of issues within the household. While possession of these assets could be a manifestation of enhancing living standards of households; disposition or lack of these assets could gauge the a complex conditions of disadvantages that individuals and a community may experience and  that would entail a shrinking household wealth/stock and hence a decline in well-being.\\
2782
2783It may result in an involuntary position and condition of households or a community at the margin of societal, political,economic, ecological and biophysical systems depriving them from access to resources, assets, services, restraining freedom of choice, hampering  the development of capabilities and ultimately causing multiple overlapping deprivations and poverty owing to the products of multiple market, institutions, policies, and so forth (Gatzweiler et al., 2011). In the ESS, households' fixed, productive and other assets are summarized as modern and traditional farm implements, home furniture, communications and entertainment equipment, households durable and a few items such as automobiles, bikes and jewelries, electronics, personal items and other assets.\\
2784
2785In light of these, a household is deprived if it does not own at least one of the assets used for access to information such as (phone (mobile or fixed), radio/tape recorder, Tv, and, Satellite Dish; access to transportation assets like private car, Bicycle, motor cycle, cart animal drawn for transporting good and people, camel, horse, donkey, Bajaj; access to income generating household equipment: Sewing machine, weaving equipment, gold, silver; access to relatively luxurious goods  such as refrigerator, electric stove, sofa set; and access to household productive assets such as oxen, cows, modern plough, pigs, sheep, goats, chicken, water pump.\\
2786
2787Finally, after summarizing all the assets into one category, A household is non-deprived if it has 3 or more small assets, Private car, and Privately owned house. However, a household is deprived if it has less than  3 small assets, no Private car and no private houses.
2788                                                
2789\item Households' Livestock Ownership:\\
2790
2791Livestock contribute to the well-being of many of the poor households in the world in general and developing countries in particular. According to FAO (2012)"Pro-Poor Policy Initiative  Livestock sector development for poverty reduction" document, roughly over 2.6 billion people in the developing world have to make a living  the growing integration of global markets provide both  new opportunities and threats to the livelihood of poor and small-scale livestock producers, traders and processors. 
2792farming. 
2793\begin{quotation}
2794Agricultural productivity gains and/or diversification into high value agricultural products that lead to increased income through increased value of output per area of land and, more importantly, per unit of labor input are essential means of raising rural income and improving food security. Because lager share of the rural poor keep livestock, because livestock can make important contributions to sustainable rural development, and because the demand for livestock, and increased livestock productivity ought to form part of the strategy for poverty reduction and agricultural growth (FAO, 2012)
2795\end{quotation}
2796In low income traditional economy, livestock form an integral part of chiefly smallholder diversified crop-livestock farming system and the live stock sector is the second most vital contributor of the agricultural economy. Therefore, livestock are central to livelihood of the poor, form an integral part of mixed farming system, generate employment opportunities, as a store of wealth, as a from of insurance, for gender empowerment by generating economics opportunities for women and boosting their bargaining power, improving the structure and fertility of soils, controlling insects and weeds, and enabling agro-industrial development. \\
2797 
2798According to the International Livestock Research Institute Report and Government Statistics(2016), Ethiopia did not unlock the potential despite its being a home to one of the largest livestock population in Africa and eight globally. It is believed that Ethiopia has approximately 50 million cattle, 50 million goats and sheep, bulk of horses, camels, donkey and chickens. The livestock sub sector  contributes for about 17$\%$ of the national GDP, 37$\%$ of agricultural GDP, 15$\%$ of export earning and 30$\%$ of agricultural employment play an essential role in national development. Mixed corp-livestock sub-sector  supports and sustains for more than 80$\%$ of the rural population and supply most of the counter's food.\\
2799 
2800As a result, a household is non-deprived in live stock if it has more than 3 small live stocks ( such as sheep, goat), more than 10 (chicken), one cattle, a horse, a donkey, a mule, and a camel and so on. Whereas, household is deprived in  live stock if it  has no more than 3 small live stocks ( such as sheep, goat), more than 10 (chicken), one cattle, a horse, a donkey, a mule, and a camel and so forth.
2801 \end{enumerate}
2802Table ($\#$) below summarizes Ethiopia's National MPI: Dimensions, Indicators, Cut-offs and Weights.
2803 \begin{landscape}
2804 \begin{table}
2805 \caption{The Dimensions, Indicators, Deprivation cutoffs and their respective weights for EMPI}
2806\label{table7}
2807 \begin{tabular}{p{3cm}p{6cm} p{14cm}p{2cm} }
2808\hline
2809\multicolumn{4}{c}{Table (6): Dimensions, Indicators, Cut-offs and Weights }\\
2810\hline
2811Dimensions and weight  &  Indicators &  Deprived if... & Relative Weight \\
2812\hline
2813Health(33.3$\%$) & Nutrition & any child under 5 years old with nutritional information is undernourished & 16.7$\%$\\
2814& Self reported Health Problems & any member of the household faces health problems  in the last two months from the interview date & 5.6$\%$\\
2815& Meals served daily & meals served daily including breakfast are taken on average less than 3 per day for adults (15+ years old) and less than 4 for children & 5.6 $\%$\\
2816& Activities of daily life & it has difficulties in the activities of daily life in seeing, hearing, walking, remembering, keeping self-care,communication; and hence absent from daily activities and cannot function  daily routines properly & 5.6$\%$\\
2817   \end{tabular}
2818   \end{table}
2819   \end{landscape}
2820   
2821   \begin{landscape}
2822 \begin{tabular}{p{3cm}p{6cm} p{14cm}p{2cm} }
2823\hline
2824\multicolumn{4}{c}{Table (7): Dimensions, Indicators, Cut-offs and Weights }\\
2825\hline
2826Dimensions and weight  &  Indicators &  Deprived if...    & Relative Weight \\
2827\hline
2828 Education (33.3$\%$)& School age children enrollment & any school age (7-13 years) child is not currently attending school & 11.1$\%$\\
2829          & Years of Education & no household member older than 15 has completed 5+ years of schooling & 11.1$\%$\\
2830    & School age children attendance & ah household is deprived if any school-age (7-17) child is out of school & 11.1$\%$\\ 
2831    & Literacy & a housed/member is deprived if more than 2 adult age (28-35  years) are illiterate & 11.1$\%$\\ 
2832
2833   \end{tabular}
2834   \end{landscape}  
2835   
2836    \begin{landscape}
2837 \begin{tabular}{p{3cm}p{6cm} p{14cm}p{2cm} }
2838\hline
2839\multicolumn{4}{c}{Table (7): Dimensions, Indicators, Cut-offs and Weights }\\
2840\hline
2841Dimensions and weight  &  Indicators &  Deprived if...    & Relative Weight \\
2842\hline
2843  Living Standard (33.3$\%$) & Access to improved electricity & if it gets electricity from  electricity from electrical batter, lantern, light from dry cell with switch, Kerosene light lamp (imported), Kerosene lamp (local kuraz), Candle/wax, Fire wood, and Other specify & 4.1$\%$\\
2844    & access to improved cooking fuels & it uses collecting fire wood, purchase fire wood, charcoal, crop residual/ leaves, dung/manure, sawdust and others & 4.1$\%$\\
2845      & access to improved drinking and running water & the household's sources of drinking water is water from kiosk/retailer, unprotected wall/spring, river/lake/pound and any other  & 4.1$\%$\\
2846      & access to improved sanitation facilities & the household uses pit latrine private ventilated, pit latrine, shared ventilated, pit latrine private not ventilated, pit latrine shared not ventilated, bucket, field, forest and others & 4.1$\%$\\
2847   \end{tabular}
2848   \end{landscape} 
2849The table below is the continuation of the Ethiopian MPI domains, indicators, poverty cut-offs and the respective weights assigned in normative decisions. The sources of indicators are basically the available information in the panel data set, from relevant literature, especially the global MPI, the basic facts, setting and matters on the Ethiopian poverty conditions, WHO, UNESCO Institute for statistics, World Bank, CSA and MoFED statistical archives and definitions\\
2850 
2851Based on that we developed four indicators in the health domain, three/four indicators in the education domain and eight indicators in the living standards domains that are believed to be better representatives of the issues under our study. 
2852%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2853 \begin{landscape}
2854 \begin{tabular}{p{3cm}p{6cm} p{14cm}p{2cm} }
2855%\hline
2856\multicolumn{4}{c}{Table (7): Dimensions, Indicators, Cutoffs and Weights }\\
2857\hline
2858Dimensions and weight  &  Indicators &  Deprived if...    & Relative Weight \\
2859\hline
2860Living Standard (33.3$\%$) & Housing conditions & the walls of the main dwelling are made from wood and mud, wood and thatch, wood only, stone only, block/non-plastered, and ree/bamboo; the roofs of the main dwelling are made from thatch, wood and made, reed/bamboo, and plastic canvas; the floor of the main dwelling  is made from mud/dung, reed/ bamboo and wood planks; and finally, it mkes use of traditional (non-removable) and it does not have at all & 4.1$\%$\\
2861      & overcrowding & more than three people live per room & 4.1$\%$\\
2862      & Assets & the household  has less than 3 small asses, no private cars and no private houses & 4.1$\%$\\
2863      & livestock & it has no more than 3 small live stocks (such as sheep, goats), more than 10 chickens, one cattle, an horse, a donkey, a mule, and a camel & 4.1$\%$\\
2864\end{tabular}
2865\end{landscape}
2866In brief, Ethiopian MPI Conceptual Framework Flow-Chart dimensions, indicators, their respective weight and their intertwined correlations  are presented in the figure 18 as follows: The Education Domain has three indicators adopted from literature that are relevant to the Ethiopian context. While the indicator years of schooling takes 50 percent of the weight normatively allotted to the  education dimension, the school age children school attendance and school age children enrollment rate take the other 50 percent weight assigned to the same dimension and it is equally shared between the two indicators.\\
2867 
2868 By the same token, 50 percent of the weight allotted to the health domain is assigned to the nutrition which is believed to be the critical indicator of this domain and is commonly used as the main indicator in many literature. Related to this, the indicators (meals served daily, difficulties in the activities of daily life and Self-reported health problems)  of the health dimension shares the other 50 percent of the weight assigned to this dimension is shared equally among the three indicators. \\
2869 
2870 
2871Corresponding, the most dominant dimension, the living standard dimension has eight indicators and the wight assigned to this domain is shared equally among each indicator. 
2872
2873\begin{landscape}
2874\begin{figure}
2875\caption{Ethiopian MPI Conceptual Framework Flow-Chart: Dimensions, Indicators, Weights and Intertwined Associations }
2876\label{Figure18}
2877\begin{center}
2878\includegraphics [scale=.76]{MPI_Dimensions_2017_Framework.PDF}%
2879\end{center}
2880\end{figure}
2881\end{landscape}
2882In comparison with the global MPI, the the Ethiopian MPI (EMPI)is congruent to the global MPI from he following perspectives:
2883\begin{itemize}
2884\item the health domain has the same malnutrition indicator. Nevertheless, the health dimension has three different indicators considered for this study. 
2885\item the education dimension has two similar indicators. However, the education has one more indicator which is school age children enrollment.\\
2886\item Living Standard dimension:\\
2887Considering the standard of living dimension, the EMPI takes into account all the global MPI living standard indicators plus some more indicator pertinent to our context research interest. Yet again, it ought to be noted that some dimension indicators in this study may be narrowly defied while other broadly defined as compared to the global MPI case. 
2888\end{itemize}
2889Considering the dimensions, indicators, deprivation cutoffs and their respective weights; the Ethiopian MPI though entirely adopted the global MPI technique and AF approach, it differs from the global MPI in many ways. To shed light on some key differences; 
2890\begin{itemize}
2891\item In the health domain, the ESS panel data sets do not provide any information on child morality, maternal mortality or any proxy variable of similar in nature. In consequence, the Ethiopian MPI has used some logical indicators (such as self-reported health problem, meals served daily and activities of the daily life)  of health dimension as we believed that these variables are better alternatives of the health domain given the data set wt our disposal. 
2892\item With regard to  the education dimension, the global MPI includes  years of schooling and school attendance. However, we included school age children enrollment, with some modification in modifying the variable label like years of education here is to mean years of schooling and school age children enrollment is parallel to school attendance. 
2893\item Besides, in the living standard dimension, we included more pertinent variables such as housing conditions is broadly defined to include the condition of the floor, roof, walls of the dwelling and the type of oven for cooking whereas only floor is considered in the global MPI.  Moreover, household assets and live stock are defined broadly  in order to contextualize to the Ethiopian houshold definition of livestock and assets. 
2894\end{itemize}
2895
2896 \subsection{Analysis of Missing Data}
2897  
2898In a survey, missing data is a common issue, and more often than not,and can be caused by many ways. Notably, respondents may refuse to answer a question because of privacy issues and the case at hand demands sensitive information by nature. To put it differently, the person conducting the survey does not understand the question. To put it in another way, the respondent might have answered,yet, the response s/he might have responded was not one of the options given. Or, possibly there was not enough time to complete the questionnaire or the households/person in the sampling space just lost interest, or maybe there is discrepancy in measurement and unit of account and so forth. Each questionnaire without a proper response is considered as a missing data. \\
2899
2900Not only survey data, but also research data are also recumbent to challenges of missing data. Besides, database also are pron to missing data problems. The repercussions of missing data could be  fickle since it is difficult to pinpoint the problem. Predictions may not be accurate when missing data are challenging since often the results are influenced and sometimes they are not. Besides, it is not always archaic whence missing data will cause a problem. As a matter of facts, each questions and/or variables not have same missing data. Also, each variables may have a small number of missing responses, yet, in combination, the missing could be enormous. Therefore, the data needs a through analysis on missing data to analyze the their impacts. \\
2901
2902Missing data may result in serious challenges. In the first place, most statistical procedures automatically eliminate cases with missing data. That means you may not have sufficient data to execute the analysis. Again, the analysis may run but results may not be statistically significant owing to small amount of input data. In the like manner, the results may be misleading if the case you analyze are not a random sample of all cases.\\
2903
2904A great deal of statistical procedures discard entire cases whenever they come across problems of missing data in any variables used in the analysis under consideration and predictions under such scenarios might be misleading even if each variable might have negligible  missing data, when analyzed  jointly with other missing variables, the aggregated observation shrinks drastically. Moreover, statistical estimation  (descriptive and regression) results may lead to misleading generalization since there is under representation and  biased outputs as a result of non-response error.\\
2905
2906To conclude, we have the view that the extent, effect and implication of missing data on the aver all analysis ought to be explored at the outset so that proper action could be taken before we dive into the detail analysis. The commonly used approaches of resolving problems of missing data are the statistical method of maximum likelihood with some underpinning assumptions such as assuming a model for the distribution of the data in the absence of missing data, and a model for the missing data mechanisms. As a point in case, he researcher might assume that the data are multivariate normal, and the missing data mechanism are is missing completely at random (the pattern of missingness is random and independent of the data values of the variables). According to the Institute for Digital Research and Education, University of California, Los Angeles (UCLA, 2015; website accessed on 25 April, 2017) a variable is said to be missing at random (MAR), according to the following definition:
2907\begin{quotation}
2908A variable is said to be missing at random if other variables (but not the variable itself) in the data set can be used to predict missingness on a given variable if neither the variable in the data set nor the unobserved value of the variable itself predict whether  a value will be missing. 
2909\end{quotation}  
2910Under the MAR presumption, which is highly associated to ignorability (a missing data mechanism is said to be ignorable if it is missing at random and the probability of missingness does not depend on the missing information itself and this assumption is required for optimal estimation of missing information),  here is that the probability of missingness does not depend on the true values after controlling for the observed variables.  Wealth of literature that deal on missing data, handling missing data, techniques and mechanisms with missing data ( such as Allison, 2002; Enders, 2010; ) suggested complete case analysis (list-wise deletion), available case analysis, mean imputation, single imputation, stochastic imputation, multiple imputation, maximum likelihood estimation and imputation, Expectation-Maximization(EM) as some of the plausible  resolutions to address the pervasive challenges of missing data.\\
2911 
2912Different statistical software code missing data differently. In Stata, for example, if the variable is numeric and if there is missing data, we see . (periods)in the dataset. If someone is working with string variables, the data will appear as  [blank]. Missing data values affect how Stata handles the data. Some common procedure are summarize, for each variable, the number of non-missing values are used; tabulate, by default, missing values are excluded and percentages are based on the number of non-missing values. If the missing options are employed on the tab command, the percentage are based on the total number of observations (non-missing and missing) and the percentage of missing values are reported in the table.\\
2913
2914Correlations,by default, correlations are computed based on the number of rows with non-missing data for the variables listed after the corr command (list-wise deletion of missing data). The Pwcorr command can be used to request that correlations be computed in a pairwise fashion, meaning that all of available data for each pair of variables will be used to compute the correlation. This means that a different number of observations may be used in the calculation of the correlation coefficients for each pair of variables. Finally, regression, is an observation is missing data for a variable in the regression model, that  observation is excluded from the regression (list-wise deletion of missing data).\\
2915
2916For this research we adopted the user-written command "mdesc" (notice that the command will not produce the percent of missing if 'missing' was coded as something other than (.)in the dataset) Stata module procedure to tabulate prevalence of missing values focusing on numeric and string variables. Mdesc, works with both numeric and string variables, produces a table with the number of missing values, total number of cases, and percent missing for each variable in the varlist (variable list). Thus, as a final check to see the total number of missing values we have for each variable. We believe that variables should not have high proportion of missing values at this stage and proper actions must be taken while proceeding with data analysis with high proportion of missing data (the command might need to be installed: write "findit mdesc" in the command window, and install it). With this context, the mdesc result is presented in the table below:\\
2917
2918\begin{table}
2919\caption{Analysis of Missing Data }
2920\label{table 8}
2921\begin{tabular}{p{9cm}|| p{2cm} p{2cm} p{2cm}}
2922\hline
2923\multicolumn{4}{c}{mdesc variable list Multidimensional Poverty Index}\\
2924\hline
2925Variable & Missing & Total & Percent Missing\\
2926\hline
2927Household member Deprived in Nutrition        & 5     & 39,775  & 0.01\\
2928Household member Deprived in Daily Functions  & 4,914 & 39,775  & 12.35\\
2929Household member Deprived in Daily Food Served & 62    & 39,775  & 0.15\\
2930Household member Deprived in Self-Reported Health & 2,652& 39,775 & 6.67\\
2931Household member Deprived in at least grade 5 years of Education  & 62   & 39,775& 0.16\\
2932Household School Age member Deprived in School attendance & 45  & 39,775 &  0.11\\
2933Household member deprived in School Enrollment Rate & 6,230& 39,775 & 15.66\\
2934Household Deprived in Electricity &  98 & 39,775   & 0.25\\
2935Household Deprived in Cooking Fuels & 246 & 39,775& 0.62\\
2936Household Deprived in Sanitation Facilities & 77   & 39,775  & 0.19
2937\end{tabular}  
2938\end{table}
2939%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2940\begin{tabular}{p{10cm}|| p{2.5cm} p{2cm} p{2.5cm}}
2941\hline
2942\multicolumn{4}{c}{mdesc variable list Multidimensional Poverty Index}\\
2943\hline
2944Variable & Missing & Total & Percent Missing\\
2945\hline
2946Household Deprived in improved access to water & 77 & 39,775 & 0.19\\
2947Household Deprived in Roof, Floor and Wall of the main dwelling & 77 & 39,775 & 0.19\\
2948Household Deprived in Overcrowding & 186 & 39,775& 0.47\\
2949Household Deprived in Assets & 0  & 39,775 & 0\\
2950Household Deprived in Livestock & 6,824 & 39,775&  17.16
2951\end{tabular}  \\
2952
2953As can be seen from the table above, apart from two multiple  deprivation indicators in the health domain (household member deprived in  the health functions and household member deprived in self-reported health); one indicator in the education domain (household member deprived in school enrollment) and one indicator in the living standard domain (household deprived in livestock) all other indicators have really small missing data and hence can be ignored without introducing a special mechanism for handling a missing data. \\
2954
2955\textit{CRAMER's V} \\
2956Cramer's V describes the correlation between indicators.Besides, it helps us to examine which indicators have high/low correlation and it enables us to identify redundancy. Research have divergence of views on how to interpret/decide based on the results of the correlation coefficient. Proponent views maintain that highly correlated indicators generate a robust measure and hence do not go for indicators that have low correlation  (Seth, 2012; Handbook of Composite Indicators; and OECD, 2008). Conversely, researchers who favored low association maintain that since high correlation illustrates circumlocution and hence redundant indicators ought to be dropped. The implication here is that the lower the redundancy the better acceptable multidimensional overlapping deprivations and poverty measurement. \\
2957  
2958The set of plausible decisions based on an analysis of associations include but not limited to, drop or modify weights based on highly associated indicators, blend some indicators into a sub-index, revisit the normative justification for considering those indicators, rearrange the categorization of indicators into dimensions, if indicators are highly associated, yet there is a solid normative or policy demand to treat both indicators, it is possible, however reconsider their weights and assign less weight on those indicators and more on the others. Or else, one might be dropped from the analysis. In cases where indicators have lower associations, but then each are independently crucial, then both can be considered in to the index with the presumption that each indicator contributes directly to poverty and well-being analysis  and so on.\\
2959
2960It focuses on ranges of dichotomous deprivations that ranges between 0 and 1; where 0,  stands for the lowest possible association between variables, and 1 stands for the largest possible associations. According to Alkire et al. (2010?) cross tab or contingency table (the basic instrument for demonstrating the relationship across indicators and basic ways to view a joint distribution) and a linked measure or a measure of similarity "P", correlation (or Cramer's V), principal component analysis (PCA) can be employed.  To shed light on the measure of similarity "P"-if two deprivations and poverty indicators are congruous, and if at least one of the marginal distributions $n_{1+},n_{+1} $ is $\neq$ 0 P is defined as:
2961\begin{equation}
2962P=\frac{n_{11}}{min[n_{1+},n_{+1}]}\in[0,1]
2963\end{equation}
2964Sources of information used by P include;
2965$n_{11}$ number of people who are multidimensionally poor in both indicators that is a joint distribution indicator.\\
2966
2967$n_{1+}$, $n_{+1}$ censored headcount ratios that indicate marginal distribution levels. Where P stands for the proportion of people who are deprived in the indicator with the lower raw headcount ratio that are also deprived in other indicators.
2968Our view is inline to the second view that favored low correlation since we argue that not only since high correlation illustrates tautology but also  embellishment of indicators will make targeting the poor difficult as disentangling and identifying the indicator becomes a very complex phenomena. Under those circumstances, in this MPI analysis we employed the correlation coefficient of the Cramer's V measure. It is the most common gauge of correlation between two nominal variables owing to its normalized range. To illuminate, the Cramer's V ranging for 0 to $\pm$1, that takes extreme values under statistical independence and complete correlation in the 2x2 case can be designated as:\\
2969
2970
2971\begin{equation}
2972V=\frac{n_{00}n_{11}-n_{01}n_{10}}{({n_{0+}n_{1+}n_{+0}n_{+1}})^{1/2}},\in[-1,1]
2973\end{equation} 
2974
2975
2976Where, $V^2$ is the mean square canonical correlation between two variables and a 2$\times$2 correlation coefficient/V could be viewed as the percentage of the maximum possible variation between two variables. Employing this technique the correlation coefficients of the Cramer's V are very small ranging from -0.2 to 0.32 supporting our view in favor of low correlation coefficient with that may help for targeting as identification of the deteriorated welfare-can be made and unnecessary repetitions are now simplified ( for the details results of the Cramer's V, you may refer to appendix 1 on page---). \\
2977
2978 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2979The Global Multidimensional Poverty Index (Global MPI) is a poverty measure that reflects the multiple deprivations that poor people face in the areas of education, health, and living standards. The Global MPI reflects both the incidence of multidimensional poverty (the proportion of people in a population who are multidimensionally poor) and its intensity (the average number of deprivations that each poor person experiences). It can be used to create a comprehensive picture of people living in poverty and allows for comparisons between countries, regions and the world, as well as within countries by ethnic group, urban/rural location, and other characteristics of households and communities." \\
2980
2981An MPI can be used to inform policy, leading to more cost-effective social programs that more efficiently target the needs of the poor. The structure of the Alkire Foster methodology has properties that make an MPI particularly useful for transparently informing policy. Among other things, it can be used to:
2982\begin{enumerate}
2983\item
2984    Produce official measures of multidimensional poverty
2985    \item     Compare incidence and intensity of poverty across countries
2986    \item     Compare sub-national groups, such as regions, urban/rural populations, and ethnic groups
2987  \item Compare composition of poverty by dimensions and indicators
2988    Report changes in poverty over time
2989(MPPN, 2016)
2990\end{enumerate}
2991Several countries have developed their own multidimensional poverty measures at the national or local level.\\
2992
2993
2994\newpage
2995 \section{ Results and Findings: Presentation and Analysis }
2996A society, household or an individual is considered as multidimensionally poor if they or one is deprived in at least one-third of the weighted indicators discussed in the conceptual framework discussed on page 165-170 on this study. \\
2997The proportion of that is multidimensionally destitute is considered as the incidence of poverty, or  simply the multidimensional headcount ratio denoted by \textit{(H}. The status depth of multidimensional poverty, or the average proportion of indicators at which indigent  people are deprived is epitomized as the intensity of poverty designated by \textit{(A)}. Finally, the multidimensional poverty index \textit{(MPI)} is computed by multiplying  the two components, namely, \textit{H} and \textit{A} across the poor and is give by $M_0=H{\times}A$. Consequently, it (MPI) shows both the share of poor people in the population and the extent to which they are deprived.\\
2998
2999If a society, a household or a person is deprived in 30-40 percent of  the weighted indicators that are considered in this study; then they are said to be in the vulnerability to poverty ration (indicated in the eighth column in the table below). However, if they are in 60 percent or more, in other words, $k=60$  percent or more, they are referred as being in the severity poverty ratio category. \\
3000
3001Those discerned as impecunious are deprived in at least one-third of the weighted domains and the indicators thereof portrayed on page 165-170 on this document. \\
3002
3003As once can see from table 9 and 10 in combination,  75.6$\%$ of Ethiopian households (78.5 million people) are multidimensionally poor. Moreover additional 18.2 million are vulnerable to poverty while more than 17 million Ethiopian are considered to be multidimensional poverty chronic population. The breadth of deprivation (intensity) across the MPI poor Ethiopians is 51.5$\%$. \\
3004
3005Finally, the share of the Ethiopian population that is multidimesionally poor, adjusted by the intensity of deprivations, is close to 39 percent. Profiling of Ethiopian household multidimensional poverty at the Regional states and city councils level is presented in the next section.\\
3006
3007 
3008\subsubsection{Profiling Multidimensional Poverty Index (MPI) in Ethiopia: at National/Federal and Regional level}
3009Employing the AF methodology and pertinent models specified fron equations (25) to (26), Ethiopia's National Poverty Index and Regional (Provincial) States under the Federal Administration; that is \textit{H}, \textit{A} and \textit{MPI}values are determined in order to achieve the first objective. \\
3010%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3011\begin{landscape}
3012\begin{figure}
3013\caption{Relationship Between Poverty Cutoffs(k) and Different     Poverty Indices}
3014\label{figure19}
3015\includegraphics[scale=0.72]{MPI_Povert_CutOffs(K)_Rellationship_2017.PDF}
3016\end{figure}
3017\end{landscape}
3018
3019The scatter chart and the bar chart on the top and bottom respectively, portray the relationship between MPI indices (given on the vertical axis) and the poverty cutoff (on the horizontal axis) with the blue color legend denoting incidence of poverty (H), the red color legend designating the poverty breadth (A), and te green color legend standing for the Adjusted Poverty Headcount ratio. \\
3020
3021As one can see from the figures above, as the poverty cutoffs (k) increases the proportion of the poor is going to decrease. Conversely, average intensity of MPI poverty (A) is positively correlated with poverty cutoffs (K). Moreover, it can be noticed that the incidence of poverty and the average intensity of poverty go in opposite direction with the poverty cutoffs. Therefore, determining the direction of the Adjusted Headcount ratio, which is basically the product of the two  needs deliberation of various scenarios. In case where, the magnitude of incidence of poverty out ways the magnitude of the intensity, then adjusted headcount ratio increase and vice versa. Another striking result we observe from the above figures is that when the poverty cutoff takes any value less than 20, the incidence headcount ratio (H) take large values close to 100$\%$.\\
3022
3023%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3024\begin{table}
3025\caption{Relationship Between MPI Indices and Poverty Cutoff}
3026\label{table9}
3027\begin{tabular}{p{4cm}|| p{4cm} p{4cm} p{4cm}}
3028\hline
3029\multicolumn{4}{c}{Relationship Between MPI Indices H, A and $M_0$ and Poverty Cut-offs (k) }\\
3030\hline
3031Poverty \\Cutoffs (k) & Incidence of Poverty (H)&  MPI Intensity  (A)  & Adjusted MPI($M_{0}$) \\
3032\hline
3033 10 &	 100$\%$ &   46$\%$ &	48$\%$ \\
3034 20	&    98$\%$	 &   48$\%$ &	48$\%$\\
3035 30	&    95$\%$	 &   49$\%$ &	48$\%$\\
3036 40	&    76$\%$	 &   51$\%$ &	39$\%$\\
3037 50	&    51$\%$	 &   57$\%$ &   29$\%$\\
3038 60	&    16$\%$	 &   64$\%$ &	10$\%$\\
3039 70	&    1$\%$	 &   75$\%$ &	1$\%$\\
3040 80	&    0    	 &   0$\%$  &	 0\\
3041 \hline
3042\end{tabular}
3043\end{table} 
3044%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3045
3046Recall that "k" in this study represents the percentage of multidimensional poverty cutoffs, that are commonly employed to determine whether a society, household or a person is under the multidimensional poverty or not. When k takes different values (10, 20, 30, 40...100), we study the variation in MPI indices at various values of 'k' and we set the plausible value of 'k' in order to scrutinize the MPI indices, poverty driving factors of each indicators, domain, dimensional and sub-group breakdowns. Moreover,  the Ethiopian households' poverty and poverty deprivation ratio vary as 'k' assumes different value. Looking at table 9 and figure 19 above, when 'k' takes smaller values (10, 20, 30), the incidence of poverty and adjusted headcount ratio are very big; where as when 'k' assumes larger values (80, 90, 100), they are very small and close to zero. Considering similar scenarios while constructing subjective poverty line in the conventional uni-dimensional poverty measurement and analysis; we considered logical but normative decision to set the value of 'k' some where in between and it is $k=40\%$ in this case.\\
3047 
3048As illustrated above, with the rise in the poverty cutoff (k), the average intensity across the multidimensional poor, i.e.,  \textit{A} is directly related; however, incidence of poverty \textit{H}, (blue color) in the scatter and bar chart presented in figure 19) is inversely related. In other words, with an increase of poverty cutoffs, \textit{H} and \textit{MPI}  display a decreasing trend while \textit{A} (red color in the scatter and bar chart)  reveals an increasing trend. It's observed that \textit{MPI} is the cross product of the incidence of poverty and the intensity of poverty takes a declining trend since the magnitude of the incidence of poverty out-ways the average intensity  \\
3049
3050By the same token, we observe inverse relationship between incidence of poverty and some level of just-adequate food energy intake k, which is used to determine the poverty line level of expenditure, z while constructing the poverty line based on the food energy intake method in the uni-dimensional poverty measurement and analysis (for details see Haugton and Khandker, 2009:p-54).\\
3051With this context, when ${H}=80\%$,  both \textit{H} and \textit{MPI} are zero indicating that there are no incidence of poverty and MPI poor Ethiopian households beyond this point.
3052
3053\begin{table}
3054\caption{Ethiopian Households MPI Poverty measures for the year 2013-2014 }
3055\label{table 10}
3056\begin{tabular}{p{2cm}|| p{4cm} p{3cm} p{2cm} p{2cm} p{2cm}}
3057\hline
3058\multicolumn{5}{c}{Adjusted Multidimensional Headcount $M_0$=H*A }\\
3059\hline
3060Index & MPI Coefficients & Std. Err & p  & [95$\%$ Conf. & Interval] \\
3061\hline
3062$M_{0}$ & 0.386***           & 0.001 &  0.00  & 0.383   & 0.389  \\
3063H       & 0.756***          & 0.003  &  0.00  & 0.745 & 0.756\\
3064A       & 0.515***          & 0.001  &  0.00  & 0.513 & 0.0.515\\
3065\hline
3066\end{tabular} 
3067\end{table}
3068Note:$ ***$ $=$significant at the 1\% level
3069%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3070In the year 2013-2014, the Ethiopian headcount ratio revealed that close to four in five (75.6$\%$) Ethiopian households were multidimensionally poor, with an average intensity of (51.5$\%$) amongst these MPI poor households. This in turn, results in a national MPI weighted score of (38.6\%). As one can see from the table 10 above, all estimated coefficients are significant at one percent significant level. However, if one looks more closely at the MPI indices, they are much higher the uni-dimensional monetary reports of absolute poverty incidence (for details see the null hypothesis proposition under the "Formulating Main Hypothesis of the Study".\\
3071
3072Looking at table 11 below, the first point to made is regarding MPI chronic population and population vulnerable to MPI at national and regional state levels. We have earlier discussed the MPI incidence, intensity, and  adjusted head count ratios in table 10. The additional information presented in  this table, therefore, is the MPI indices at regional level displayed for each regional states. For example, (Tigray: $H=71.5\%$, $A=52\%$ and $MPI=37.3\%$; Amhara: $H=72.5\%$, $A=51.5\%$ and $MPI=37.8\%$; Oromia: $H=78\%$, $A=52.6\%$ and $MPI=41.5\%$; and Oromia: $H=75.8\%$, $A=52\%$ and $MPI=39.5\%$; Rural:$H=76.8\%$, $A=51.5\%$ and $MPI=39.5\%$; Urban:$H=47.7\%$, $A=49.3\%$ and $MPI=23.5\%$ and so forth ) just to mention the MPI indices of the Federal Government  of Ethiopia major regional states. What the table also shows MPI chronic population and population vulnerable to MPI at both national and regional level. \\
3073
3074What table 11 also show is that head count population share (76\% of the sample) which can be roughly estimated to be 77, thousand of the Ethiopian total population. Correspondingly, Ethiopian households who are MPI chronic and vulnerable to poverty are extrapolated as 16, 896 and 18, 540 thousand respectively. 
3075 
3076%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3077\begin{landscape}
3078\begin{table}
3079\caption{Ethiopian Households MPI Poverty measures for 2014 ESS }
3080\label{table11}
3081\begin{tabularx}{\linewidth}{*8{C}}%
3082\toprule
3083 Region & \multicolumn{2}{c}{\thead{National and Regional\\ $M_{0}$ = (H*A)}} & \multicolumn{3}{c}{Population in the MPI} & \thead{MPI Chronic\\population} & \thead{Population\\ Vulnerable\\ to MPI} \\
3084\cmidrule(l{1.5em}r{1.5em}){2-3} \cmidrule(lr{1em}){4-6} \cmidrule(lr){7-7}\cmidrule(lr){8-8}
3085 & \makecell {Survey \\Years} & \makecell{Adjusted\\ Multidimensional\\Headcount} (Index) & \makecell{ Headcount\\(K=40\%)} & \makecell{Headcount\\(Pop share)\\of the\\ sample\\$n=39775$} & \makecell{Intensity of\\Deprivation \\(\%)} & \makecell{ In Sever\\ Poverty\\(K=60\%)} & \makecell { Vulnerable \\ to poverty\\(K=30\%)}\\
3086\midrule
3087Ethiopia & LSMS/2014 & 39\% & 75.6\% & 30,070 &  51.5\% & 16.5\% & 17.5\% \\
3088Tigray  & LSMS/2014 & 37.3\% & 71.5\% & 0.082 & 52\% & 15.5\% & 14.5\% \\
3089%Addis Ababa & LSMS/2014 & Y & 20 & 30 &100 & 140 & 120 \\
3090Afar     & LSMS/2014 & 38\% & 77\% & 0.038 &49\% & 17\% & 15.3\% \\
3091Amhara   & LSMS/2014 & 37.8\% & 72.5\% & 0.195 &51.5\% & 16.3\% & 15\% \\\addlinespace[2ex]
3092Oromia & LSMS/2014  & 41.2\% & 78\%    & 0.228 &52.6\% & 17.5\% & 15.6\% \\
3093Somali & LSMS/2014 & 43.5\% & 82.6\%     & 0.066 &51.7\% & 17.5 & 16.5\% \\
3094Benshangul{-Gumuz} & LSMS/2014 & 33.8\% & 67.5\% & 0.03 &50\% & 15\% & 13.6\%\\
3095SNNP   & LSMS/2014 &  39.5\% & 75.8\% & 0.262 & 52\% & 17\% & 15\% \\\addlinespace[2ex]
3096Gambela & LSMS/2014 & 33\% & 65\%      & 0.031 & 47.7\% & 14.5\% & 13\% \\
3097Harari & LSMS/2014 & 37.5\% & 74\%     &  0.032 & 50.5\% & 16.5\% & 14.8\% \\
3098Dire Dawa & LSMS/2014 & 40\% & 75\%     & 0.035 & 53.2\% & 16.7\% & 15\% \\\addlinespace
3099Rural & LSMS/2014 & 39.5\% & 76.8\%    & 0.94  & 51.5\% & 21\% & 23\% \\
3100Urban & LSMS/2014 & 23.5\% & 47.7\%    & 0.06  & 49.3\% & 15\% & 16.5\% \\
3101\bottomrule
3102\end{tabularx}
3103\end{table}
3104\end{landscape}
3105%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3106Likewise, MPI headcount poor population share for rural and urban Ethiopian household is 94$\%$ and 6$\%$. It may be deduced that the MPI headcount poor populations share is very small as 20.5\% of the total Ethiopian population live in urban areas. More importantly, MPI chronic population in rural and urban area of Ethiopia is estimated to be 21$\%$ and 15.6\% respectively. Besides, population vulnerable to MPI in rural and urban areas is worked out as 23$\%$ and 16.5$\%$. \\
3107
3108Considering Ethiopian Regional States MPI poor population share, MPI chronic population and population vulnerable to MPI for the major regional states of Tigray(8.2$\%$, 15.5$\%$ and 14.5$\%$); Amhara (19.5$\%$,  16.3$\%$ $\&$ 15$\%$); Oromia(23$\%$, 17.5$\%$ $\&$ 15.6$\%$) and Southern Nations, Nationalities and Peoples(SNNP)( 26.2$\%$, 17$\%$ $\&$ 15$\%$) respectively. 
3109\subsubsection{Comparing Ethiopia's MPI with Other One-Dimensional Monetary Measures}
3110Comparing the estimation results of the MPI induces with previous Ethiopia's monetary poverty statistics and MPI abstracts presented in the Oxford Poverty  and Human Development Institute (OPHI) Country Briefing, 2017 provide insightful perspectives and have profound importance. Looking at the MPI estimation results,  we observe that these results are smaller than the MPI results provided in the "Country Briefing, 2017" reports in which $H=87.35\%$, $A=64.65\%$ and $MPI=0.564$ (OPHI, 2017). Note that the OPHI,  2017 MPI reports are different from our results as the former employed the global MPI indicators, 2011 demographic and health survey data set (DHS) and relatively smaller poverty cutoff. \\
3111
3112However, the MPI indices estimation results of this study are much higher than Ethiopian monetary poverty reports presented World Bank reports (World Bank, 2014, 2015 and 2016) that range approximately from $38\%$ to 29$\%$. Using the one dimensional monetary measures of poverty, the Government's of Ethiopia Official reports range from 26$\%$ to 23$\%$ (MoFED, 2015 and 2016); which are very much lower than the results we presented above and even much lower than the World Bank report. 
3113
3114\subsection{Ethiopia's MPI Decomposed by Dimensions and Indicators}
3115Another vital feature of the MPI is that, given that the poor have been identified and MPI indices being estimated, it can further be decomposed it to its components of censored indicators. It can be demonstrated that when MPI is decomposed by population subgroups,one can look at the MPI in each group but also at the contribution of that group to the over all MPI.
3116\subsubsection{Percentage Contribution of Each Indicators to MPI ($M_0)$}
3117The next question after computing MPI indices, presenting, and profiling MPI estimation results and findings at national and regional level is what is driving the MPI poverty situations ? and what does each indicator contribute to the over all MPI?  As discussed above, the core advantage of the AF approach over other uni-dimensional monetary measures in particular and over other multidimensional measure in general is that the MPI can study at the contribution of each domains and indicators to the aggregate multidimensional poverty score and decompose this contribution to look at what is contributing more or less at a particular point in time. \\
3118
3119The three figures presented below stand for percentage of contribution of each indicators to the over all MPI ($M_0$) portrayed in pie chart, column graph and bar chart having the same message and addressing the same issue.\\
3120Looking at the figure below, the fist point that can be made is that of all the indicators considered, the largest contributors is the enrollment rate in the education dimension(18$\%$) which is again the second contributor of the three dimensions which contributes (28$\%$) as discussed below. In the light of the What more do these graphs display are the self-reported health indicator is the second largest contributor (contributing over 9$\%$) to the Ethiopian MPI. Moreover, MPI indicators in the living standard dimension, such as access to improved cooking fuels, sanitation facilities, housing conditions, and livestock are the third largest contributor indicators to the Ethiopian MPI. Whereas activities of daily life and years of education are the least contributors. \\
3121  
3122%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%  
3123\begin{figure}
3124\caption{Pie Chart: Percentage Contribution of each Indicator to Ethiopian MPI}
3125\label{Figure20}
3126\includegraphics[scale=0.90]{Pie_Chart_Contribution_of_Each_Indicator_MPI}
3127%\label{Figure15:PieChart}
3128\end{figure}
3129
3130%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3131
3132%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3133\begin{figure}
3134\caption{Alternative Presentation: Percentage Contribution of Each Indicator to MPI}
3135\label{Figure21}
3136\includegraphics[scale=.65]{MPI_Contributions_Indicators_2017.PDF}
3137\end{figure}
3138The contribution of each dimension  to the over all Ethiopian MPI is presented in the figures below. As it can be seen from the  figure below, the living standard dimension take the lions share contribution 58$\%$ to the over all adjusted MPI. Moving on, socioeconomic deriving factors, the education and health sector contribute 26$\%$ and 16$\%$ respectively. While it can be moderately claimed that Ethiopia has shown some positive progress in the health and education sectors, situations in the living standards remain in the worst scenarios. \\
3139\begin{figure}
3140\caption{Pie Chart: Percentage Contribution of each Domain to Ethiopian MPI}
3141\label{Figure22}
3142\includegraphics[scale=.75]{Pie_Chart_Domains_2017.PDF}
3143\end{figure}
3144%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3145%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3146\subsubsection{Dimensional Breakdown in the Ethiopian MPI: Uncensored VS. Censored Headcount Ration in }
3147As discussed earlier, another form of breakdown, that is dimensional breakdown in lieu of decomposition by population subgroup assists us to identify the deprivations experienced by the poor and measure and track back progress made in reducing these deprivations. The Ethiopian MPI uses 15 indicators (eight in the living standard domain, four in the health domain and three in the education domain)  with different relative waiting (relative weights give based on literature, consultancy with experts and based in normative judgments) to measure multidimensional poverty in Ethiopia.\\
3148 
3149The bar chart at the top presents the the uncensored headcount (UH) ration of the dimensions designating the population deprived in that give dimensions. Besides, it reports the censored headcount(CH) ratio of a dimension standing for the proportion of the population that is multidimensionally poo and deprived in that dimension at the same time. \\
3150
3151To illuminate more on this point, the green bins in the top bar graphs indicate the uncensored headcount ratio while the golden bins stands for the censored headcount ratio.  As noted in the methodology part, the uncensored headcount ratio reports similar information that could be obtained when the union approach is employed. Therefore, the green color in the top bar graph summarizes the weighted MPI average of the uncensored headcount ratio of a dimension which is the percentage of population who are deprived in each dimension.\\
3152\begin{landscape}
3153\begin{figure}
3154\caption{Bar Chart: Relationship  Between  Uncensored (green color)\\ and Censored (yellow color) Ethiopian MPI Indicators}
3155\label{Figure23}
3156\begin{landscape}
3157\includegraphics[scale=.75] {Uncensored_Censored_MPI_Indicators_Graph.PDF}
3158\end{landscape}
3159\end{figure}
3160\end{landscape}
3161%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3162Scanning the figure above, the bar graph on the top reveals that largest  percentage of the population are  deprived in the housing conditions (94$\%$), cooking fuels (93$\%$), electricity (92$\%$), livestock (92$\%$), and access to improved sanitation facilities (91$\%$) are among the top five dimensional contributors to both the percentage of population who are deprived in each dimension (Uncensored Headcount Ratio) and the percentage of population who are both multidimensionally por and deprived in that dimension (Censored Headcount Ratio) where the are denoted   by the yellow color summarizes weighted MPI average of the censored head count ratio. \\
3163
3164Furthermore, the spider diagram at the bottom shows the level of these same dimension wise decomposition of uncensored and censored headcount ratios. Recalling the discussions on dimensional breakdown under the methodology section, it's vital to explain the relationship between the $M_0$ and the uncensored headcount ratio when a union approach is employed to identify the poor. It can further be recalled that when the union approach is used, the censored headcount ratio for a dimension is its uncensored headcount ratio. Therefore, the Adjusted Headcount Ratio $M_0$  with the union approach is the weighted average of the uncensored headcount ratios.\\
3165
3166Moreover, looking at both figures above, the uncensored headcount ratio is often higher than the censored headcount ratio. Hence, the uncensored headcount ratio of a dimensions cannot be lower that its censored headcount ratio. Besides, the kinds of policy analysis that can be conducted employing the censored headcount ratio may be summarized as it help identify deprivations experienced by the poor and measure progress in reducing deprivations.  It could be also deduced that the uncensored headcount ratio reveals a very high deprivation and poverty levels. \\
3167 
3168Finally, making analysis of the uncensored headcount(UH) ratio of a dimensions which stands for the proportion of the population deprived in those dimensions against the censored headcount (CH)ratio of a dimension that represents the proportion of the population that is multidimensionally poo and deprived in those dimensions at the same time are very crucial for focused targeting aiming at reducing poverty at each  deprived dimension and for measuring deprivations and poverty.\\
3169%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3170\subsubsection{Censored Headcount Ratios (Deprivations) in each Ethiopia's MPI Indicators}
3171The figure below shows incidence of censored deprivation in each Ethiopia's MPI indicators. As can be seen from figure (bar chart) below, access to improved cooking fuels(88$\%$), access to electricity (85\%), access to improved housing/residence(85$\%$), self-reported health(82$\%$), and access to sanitation facilities (81$\%$) report five highest proportion of the population that is poor and deprived in each indicator, this is also referred to as the censored headcount ratios. It ought to be noticed that the proportion of the deprivation of non-poor people is precluded.  \\
3172    
3173The estimation results of incidence of censored deprivations, like the censored headcount ratio in the dimensional break down, report lower values that are values that lie in between the results of union and the intersection approach. besides, they are always smaller than the raw headcount ratios. It is logical to pose the question why do we use  censored and uncensored multidimensional poverty measurement and analysis? The precise answer is since the union approach give usually a very  high result whereas the intersection provides a very small result; it is very difficult for decision making in efforts to alleviate poverty and hence we need a relative accurate value which is in between the union and intersection approach.\\
3174 
3175It is important to make a distinction between the dimensional breakdown uncensored headcount ratios(UH) and censored headcount ratios and the raw headcount and censored headcount in the deprivations in each indicator. The former measures the contribution of each dimensions (hence called dimensional breakdown) to over all poverty, for example censored headcount ratio of a dimensions is the percentage of the population who are both multidimensionally poor and deprived in that dimension, whereas the latter (the censored headcount ratio of an indicator)gauges the proportion of the population that are poor and deprived in that indicator.  
3176
3177\begin{figure}
3178\caption{Censored Deprivation in each Ethiopian MPI Indicator}
3179\label{Figure24}
3180\includegraphics[scale=.85]{Censored_Deprivation_Indicators_2017.PDF}
3181\end{figure}
3182%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3183\newpage
3184%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%	
3185\subsubsection{ MPI Decomposition by Regions}   
3186The unique feature and status of MPI that make it more preferable among all other uni-dimensional monetary measurement of poverty measure and multiple measurement is that it can be decomposed by different population subgroups that may include (example region, age-group, gender, area/residence an so forth).\\
3187Since the subgroups are pertinent to a Ethiopia in general and to the four major regional states in particular; and the data employed definitely allows to compute MPI estimates in a representative way.\\
3188
3189Other that we considered in this study include, gender, age-group, urban vs rural and so on. Finally, to carryout the decomposition exercise, it is essential to check that the data employed is representative like in our case. \\
3190
3191The MPI can be decomposed by different populations subgroups, aiming at demonstrating  how the anatomy of multidimensional poverty differs among various regions or group. Figure 25 below reports the absolute indices regional (subgroup) decomposition of incidence of poverty (the blue color on the left hand side), MPI(the golden color in the middle) and population share(on the right hand side).\\
3192 
3193Among all the Regional States that constitute the Federal Democratic Republic of Ethiopia, we shall focus more on the four major regional states of Tigray, Amhara, Oromia and Southern Nations, Nationalities and Peoples. Considering the multi-clustered incidence of poverty, MPI and population share chart below,  Tigray is in the first column with (0.683, 0.352, and 0.082); Amhara in the third column with (0.737, 0.379 and 0.195); Oromia in the fourth column with (0.776, 0.405 and 0.228) and SNNP with (0.754, 0.388 and 0.262) the absolute figures representing regional incidence of poverty, MPI and population share respectively. From the visual inspection of this multi-clustered charts, it can be deduced that the incidence of poverty and MPI is higher in Oromia, followed by SNNP, Amhara and Tigray regional states respectively. \\
3194Considering all Federal States of Ethiopia, the poverty headcount ratio (H) and the Adjuste Headcount Ratio were high in Somali Regional State. However, Since we have best representative sample for the four major regional states (Amhara, Oromia, SNNP and Tigray), we will discuss the detail analysis and compare results among these regional states. It should also be noted that these regional states are home to more than 85 percent of the Ethiopian population. Considering these four major regional states,  incidence of headcount ratio and adjusted headcount ratio was high in the Oromia regional states where more than 37 million Ethiopin live. Despite the fact that SNNP has the highest population share in this particular time..\\
3195
3196Finally, as can be seen from figure 25 below, the MPI poverty profile of all Ethiopian Regional States, including the four major regional states, is presented for the specific survey year 2013-2014. \\
3197
3198%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3199
3200%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3201\newpage
3202\begin{landscape}
3203\begin{figure}
3204\caption{Absolute MPI Indices and Population share Decomposed by Regions(Subgroup)}
3205\label{Figure25}
3206\!\includegraphics[scale=1.2]{MPI_Decomposition_By-Region_2017.PDF}
3207\label{figure25}
3208\end{figure}
3209\end{landscape}
3210%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3211
3212 \begin{landscape}
3213 \begin{figure}
3214 \caption{Percentage Contribution of Each Region to MPI (H and MPI) Indices}
3215 \label{figure26}
3216 \! \includegraphics[scale=.90]{Percentage_regional_Contribution_2017.PDF}
3217  \end{figure}
3218 \end{landscape}
3219Taking a loot at figure 26 above, it reports the percentage contribution of each regional state to the headcount ratio and adjusted headcount ratio when multidimensional indices are decomposed by the same regional states of Ethiopia that we discussed earlier. The key difference between figure 25 and figure 26 is that while figure 25 reveal the absolute value of MPI indices when decomposed by regional states, figure 26 however presents the percentage contribution of each subgroup to the MPI indices. In this case, the subgroup decomposition is by regional states of Ethiopia\\
3220
3221As can be seen form figure 26 above, the Oromia regional state has the highest percentage contribution to the MPI indices, \textit{H} and \textit{MPI} followed by SNNP, Amhara and Tigray regional states respectively. Figure 26 presents more insightful and vital information for policy makers while crafting antipoverty programs aiming at tackling  overlapping multidimensional  deprivations and  poverty as it gives specific information that are crucial for targeting the deprived and the poor. \\
3222
3223In summary, figure 25 and 26 should not be understood as alternative way of presenting MPI indices just decomposed by regions as they provide different information. In other words, while figure 25, presents the MPI profile at each Ethiopian Regional states, figure 26, on the other hand displays the contribution of each regional to MPI indices that are highly importtant for decision making regarding alleviating MPI by policy makers. \\
3224
3225 \begin{landscape}
3226\begin{figure}
3227\caption{Bar Graph: Percentage Contribution of each Indicator to   Regional States MPI Indices}
3228\label{figure27}
3229\!\includegraphics[scale=0.96]{Regional01_decomposed_Indicators_Contribution_2017.PDF}
3230\end{figure}
3231\end{landscape}
3232Figure 27 above displays percentage contribution of each indicator to  regional states MPI indices. In other words, the percentage contribution of each MPI indicators identified in this study to the regional level multiple deprivations and poverty or percentage contribution of each indicators to MPI decomposed by subgroup-Ethiopian regional states. \\
3233
3234Likewise to the percentage contribution of each indicator to the Ethiopian  national/federal level MPI profile, school age children school enrollment rate, self-reported household health conditions, access to improved sanitation facilities,  access to improved electricity and cooking fuels, access to improved housing conditions, and school age children school attendance take the larger share in their percentage to the regional level MPI profile. \\
3235
3236Correspondingly, figure 28 below, reports percentage contribution of each dimensions (health, education  and living standard) to regional level MPI ($M_0$) poverty profile. Saying it differently, the percentage contribution of each domains to the MPI decomposed by regional state subgroups. Similar to the Ethiopian national MPI, the living standard dimensions has the largest percentage  contribution followed by the education dimensions and health dimensions.\\
3237
3238A worms-eye-view look at figure 28 reveals that multiple deprivation and poverty of SNNP (58.3\%), Amhara (57.1\%), Oromia(56.9\%) and Tigray (55.7\%), ranked from the largest to the relatively smallest, are contributed by the living standard   dimensions. Besides, considering the multiple interlinked deprivations and poverty of the regional states poverty profile, Tigray(27.8\%), Amhara(27.8\%), Oromia (26.7\%) and SNNP(25.1\%) ranked in order of their magnitude, is attributed to the percentage contribution education dimension. Finally, SNNP (16.6\%), Tigray(16.5\%), Oromia(16.4\%) and Amhara (15.1\%) is attributed to the health dimension respectively. 
3239\begin{center}
3240\begin{figure}
3241\caption{Percentage Contribution of Each Domain to Ethiopian Regional States MPI Indices}
3242\label{figure28}
3243\includegraphics[scale=0.85]{Domains_Decomposed_By_Regions.PDF}
3244\end{figure}
3245\end{center}
3246%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3247\newpage
3248\subsubsection{Decomposition by Age-group}
3249\begin{table}
3250\caption{Ethiopian Households MPI Decomposition by Subgroups - (MPI Decomposed by age$age_group$) for the year 2013-2014 }
3251\label{table 12}
3252\begin{tabular}{p{2cm}|| p{2cm} p{2cm} p{2cm} p{2cm} p{2cm} |p{2cm}}
3253\hline
3254\multicolumn{7}{c}{Ethiopian MPI Decomposition by Age Group}\\
3255\hline
3256MPI Indices & 1-15 years  & 16-29 years  & 30-45 years & 45-60 years & 65+ years  & Total\\
3257\hline
3258H   & 74.8\%        & 77.1\%  &  70\%  & 74.9\% & 76.9\%& 74.5\% \\
3259$M_{0}$ & 0.38       &  0.40 &  0.35  & 0.383   & 0.396 & 0.38 \\
3260Pop share  & 43.2\%   & 24.2\% &  18.2\%  & 11.1\% & 3.4\% & 1.00\\
3261\hline
3262\multicolumn{7}{c}{Percentage contribution of subgroups/age-groups to MPI Indices}\\
3263\hline
3264H      & 43.3\%     & 25\%  &  17.1\%  & 11.1\% & 3.5\%& 100\% \\
3265$M_{0}$ & 0.431    &  0.254 &  0.168  & 0.112   & 0.035 & 1.000 \\
3266\hline
3267\end{tabular} 
3268\end{table}
3269note: 1-15 years age (age-group$_1$), 16-29 years age (age-group$_2$, 30-45 years age (age-group$_3$), 45-60 years age (age-group$_4$ and 65+ years age (age-group$_5$).\\
3270
3271Looking at table 12 above, share of the poor people among the population or incidence of poverty at various age cohort, in other words, \textit{H} decomposed by various age-groups is ordered from highest to the lowest as age-group$_2(H=77.1\%)$, age-group$_5(H=76.9\%)$, age-group$_4(H=74.9\%)$, age-group$_1(H=74.8\%)$, and age-group$_3(H=70\%)$ respectively.\\
3272
3273Correspondingly, the adjusted headcount ratio $M_0$ is ranked as
3274age-group$_2(M_0=0.40)$, age-group$_5(M_0=0.396)$, age-group$_4(M_0=0.383)$, age-group$_1(M_)=0.380)$, and age-group$_3(M_)=0.350)$ respectively. It can be seen that the absolute age-group (subgroup) decomposition of \textit{H} and \textit{MPI} have the same order in the age cohort multifaceted distribution of deprivations and poverty.   \\ 
3275
3276Considering population share of each age-group, it can be ordered from the highest share to the lowest share as follows: (43.1$\%$, 24.2$\%$, 18.2$\%$, 11.1$\%$, and 3.4$\%$) for each age-groups, that is $age-group_1$, $age-group_2$, $age-group_3$, $age-group_4$, $age-group_5$ in accordance with. \\
3277
3278It is remarkable to note that the absolute MPI indices decomposition by age-group(subgroup) follow more or less similar distribution to the demographics structure and distribution of population according to age  structure where 0-14 years  accounts for 43.8 (male: 22, 430, 854/female 22, 317, 780), and 15-24 years accounts 20.5\% (male 10,197/ and female 10, 332, 643). In general, 65\% of the Ethiopia population is under 35 years old. As can be seen from figure 12, age-group$_1$ and age-group$_2$ (below 30 years old) cover more than 68.5$\%.$
3279
3280Uniquely, percentage contribution of each age-group (subgroup)to the incidence of poverty(headcount ratio, \textit{H}) and to the adjusted headcount ratio ($M_0$)written as 43.3\%, 25\%, 17.1\%, 11.1\% and 3.5\%. By the same token, for $(M_0)$ it is given in a similar order as 43.1\%, 25.4\%, 16.8\%, 11.2\% and 3.5 respectively. 
3281It is insightful to notice from table 12 above that very less that 15 years school-age children and very old people (65+ year) Ethiopians experience  the highest incidence of multidimensional poverty and adjusted MPI headcount as compared with other age categories. 
3282\begin{large}
3283Interpretation of each age-group, H and MPI and its contribution 
3284\end{large}
3285\begin{figure}
3286\caption{ MPI Indices and Percentage  Contribution of Each Indicator to MPI Decomposition by Age-group}
3287\label{Figure29}
3288\begin{center}
3289\includegraphics[scale=.75]{MPI_Decomposed_By-Agegroups_2017.PDF}
3290\label{figure29}
3291\end{center}
3292\end{figure}
3293Figure 29 above, portrays MPI Indices, their standard errors and confidence intervals (top figure) which closely similar to the MPI indices discussed earlier. Furthermore, i shows the percentage contribution of each indicators to MPI indices decomposed by age-groups. Identical to the reporting in the percentage of each indicator to the Ethiopian national and regional MPI indices; school age children enrollment role ranging between [16.5\%, 17\%, the smallest being the largest age-group followed by the second age whereas, the largest being the first age-group ],  self-reported health conditions ranging between [7.3\%,10\%, from the smallest age-group to the largest age-group] and access to improved cooking fuels, electricity, housing conditions are the next contributors to the MPI indices composition by age-groups. 
3294\newpage
3295\subsubsection{MPI Decomposition by Gender}
3296As the famous slogan goes, 'poverty has female face" so does MPI
3297www\\
3298xxx\\
3299zzz\\
3300... this  is very important topic to discuss and must be discussed in detail in more than two pages. 
3301....
3302Another subgroup decomposition of MPI with a critical implication is that decomposition by gender. As can be seen in figure 30 below (top), the multidimensional headcount ratio (multidimensional incidence of poverty) for females and female-headed households in Ethiopia is $H=73.7\%.$ Besides, for males and male-headed households is $H=76.4\%$ \\
3303Likewise, the population share for females and female-headed Ethiopian households is 49.4$\%$ whereas the population share for males and male-headed Ethiopian households is 50.6$\%$.\\
3304
3305Additionally, the percentage contribution of subgroup, in other words decomposition of MPI by gender, to the MPI indices is summarized in table 13 above. As can be seen from table 13, considering incidence of poverty, \textit{H}, it is 48.5\% for females and female headed households and  51.5\% for males and male-headed households.\\
3306 
3307Looking at the bottom part of figure 30 below, it presents the contribution of each MPi indicator to the MPI indices decomposed by gender.\\
3308It can be revealed from this graph that, the driving factors for the MPI decomposition by gender, to mention some among others are, like wise in the other findings reports explained above; enrollment rate is the main factor contribution 17.6\%, 17.5\% for females/female-headed and males/male-headed households respectively. \\
3309
3310To briefly shed light the other driving factors or percentage contribution of each indicators, is improved sanitation facilities (8.2\% and 8.1\%);  improved housing conditions (8.2\% and 8.1\%); household living stock (7.9\% and 7.9\%) for females/female-headed  and male/male-headed household respectively.
3311\begin{table}
3312\caption{Percentage Contribution of Subgroup to MPI Indices Decomposed by Gender ) for the year 2013-2014 }
3313\label{table 13}
3314\begin{tabular}{p{4cm}|| p{4cm} p{4cm} |p{2cm}}
3315\hline
3316\multicolumn{4}{c}{Percentage Contribution of Subgroup to MPI Indices Decomposed by Gender}\\
3317\hline
3318MPI Indices & Male & Female & Total\\
3319\hline
3320H   &    76.4\%     & 73.7\%  &  1.000 \\
3321$M_{0}$ & 0.517     &  0.483 & 1.000 \\
3322\hline
3323\end{tabular} 
3324\end{table}
3325\begin{landscape}
3326\begin{figure}
3327\caption{Absolute Indices Decomposition by Subgroup-Gender and Percentage Contribution of Each Subgroup}
3328\label{figure30}
3329\begin{center}
3330\includegraphics[scale=0.85]{MPI_Decomposition_By_Gender_2017.PDF}
3331\end{center}
3332\end{figure}
3333\end{landscape}
3334%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3335\subsubsection{MPI Decomposition by Area/Residents}
3336Another key quality of the MPI is again, once the multidimensional overlapping deprivations and  poor have been identified and the MPI indices have been calculated, the decomposition of the MPI by the human settlements of the poor as classified as rural or urban depending on the density of human created structure and resident people in a particular area. Roughly speaking, urban areas in the Ethiopian context can include small towns, towns and cities  while the rural areas include villages, hamlets and the country side in general. \\
3337
3338While rural area may develop randomly on the basis of natural vegetation and fauna available in a certain area, urban settlements are proper, planned settlements built up according to a process called urbanization. In most cases, rural areas are focused upon by governments, seats of heads of local government admins and grand development programs in the neighboring areas which in the process turned in to urban areas. \\
3339Similar to many parts of Sub Saharan African countries, while rural settlements are based more on natural resources and events, the urban population often receive the benefits of modern advancement, opportunities for  improved service delivery (education, health, credit, transport, employment, business, social interactions) and over all better standards of living. \\
3340
3341Correspondingly, urban areas in Ethiopia could be created through urbanization and are categorized by urban morphology as cities towns, conurbations or suburbs. In urbanism,contrast to rural areas like villages and hamlets, the urban sociology or urban anthropology it contrasts with the natural environments. Most of the urban area in Ethiopia (particularly in big cities) the house are dilapidated, tumbledown and slam houses are the major characteristics of the urban settlements and hence decomposition of MPI by area is core function of the study as it demands particular policy formulation while targeting the poor. \\
3342
3343It can be seen from figure 31 below that  the absolute MPI decomposition of by area reveals that the incidence of poverty \textit{(H)} in the rural ares is $H=76.8\%$ whilst in urban area, it is $H=47.7\%$. Corresponding, Absolute MPI contribution to subgroups or MPI decomposition by area is given as $M_0=0.395\%$ for rural areas and $M_0=0.235$ for urban areas. \\
3344Considering percentage contribution of the of subgroups/MPI decomposition by area shows that the incidence of poverty in the rural area contributes $H=96.2\%$ while the same contribution of the urban area is $H=3.9\%.$ However, the MPI percentage contribution in the rural area is $M_0=96.4\%$ while in urban areas it is $M_0=0.036.$
3345
3346Finally,regarding the contribution of each indicator to MPI indices decomposition by area, here as well, the school age children enrollment rate takes the lions share by contributing 19.4\% in rural areas and 17\% in urban areas. Besides, school age children school attendance contributes 11\% for rural and 7.5\% for uban areas; self-reported health contributes 9.1\% for rural and 9.6\% for urban; access to improved cooking fuels contributes 8.1\% for rural and 8.3\% for urban areas; access to improved sanitation facilities contributes 8.1\% for rural and 8.5\% for urban  just to shed light on percentage of contribution of some indicators when the MPI indices decomposition is performed by areas as rural and urban settlements. 
3347
3348\begin{landscape}
3349\begin{figure}
3350\caption{MPI Decomposition by Area/Residents}
3351\!
3352\label{Figure31}
3353%\begin{center}
3354\includegraphics[scale=0.90]{MPI_Decomposition_By_area_2017.PDF}
3355\label{figure31}
3356%\end{center}
3357\end{figure}
3358\end{landscape}
3359%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3360       \subsection{Mapping Ethiopia's National and Regional MPI Indices}
3361   
3362%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3363%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3364       
3365Poverty mapping are extremity vital for improving the targeting 
3366of antipoverty intervention programs. In drafting poverty alleviation packages, programs or projects, and disbursement of subsidies, resources utilization efficacy can be enhanced if the most needy group of the community re accurately targeted. 
3367 Besides, it is a cost saving approach as it reduces the leakage of resources to non-poor persons and it reduces the risk that poor persons will maybe overlooked by the program that was designed to assist the poor. Furthermore, poverty maps are very crucial for the government to articulate their antipoverty policies, strategies and objectives.\\
3368 
3369Correspondingly, poverty maps assist governments to articulate their policy objectives. To illuminate this, making resources allocation decisions on the basis of observed geographic poverty data in lieu of perceptive and normative rankings and ordering of regions can boost the transparency of government decision makings in their endeavor to reducing/eradicating poverty. Not only such data can  facilitate to curb the influence of special interest groups and vested interests in the allocation decision but also play a role for well-defined poverty maps in lending credibility to governments and donor partners in decision makings (World Bank Manual, 2015). \\
3370
3371According to the World Bank Manual (2015), the following four steps are recommended to create viable poverty maps:
33721.
33732.
33743.
33754.
3376 
3377Showing the spatial distribution of MPI indices (poverty headcount ratio and adjusted headcount ratio) by mapping helps to analyze the spatial distribution of MPI indices that in turn will facilitate and enable the multidimensional overlapping deprivation and targeting MPI chronic poor more effectively and rigorously. \\
3378
3379We argue that poverty has a geographical dimension. Geography, particularly the physical environment, plays a significant aptness on the state of poverty particularly in developing countries. However, this geographic dimension using the AF approach of multidimensional poverty analysis has not been given much attention particularly in he Ethiopian context. In an attempt to underscore its importance this study explores the level of multidimensional poverty maps which is an important value addition of this study. As explained in the methodology part under sub title "Creating Ethiopian Multidimensional Poverty Maps", multidimensional poverty maps demands data on the health dimension, education dimension, living standard dimensions and the (indicators thereof), in other words,  each indicators under every domain discussed earlier. \\
3380
3381The results obtained through this process could be slightly different variant of the MPI since creating MPI maps requires additional data. This procedure has a significant benefit as multidimensional maps do not suffer from specification errors and can further be dis-aggregated to any desired geographic and administration level, village/urban, up to specific household residents. \\
3382
3383According to Elbers et al.(2007), a geographically targeting  poverty reduction scheme activities based on districts or communities is significantly worse than household level targeting. Therefore, any further disaggregation of poverty imputations would be of great use to the policy makers and development practitioners targeting antipoverty programs aiming at  assisting the poor. Additionally, multidimensional poverty maps developed based on this technique are much easier to be produced than other methods that call on rigorous imputations and calculations.\\
3384 
3385Constructing process of the multidimensional poverty map commences with an appraisal of the level or profiles of MPI indices as showed in tables (10) and (11) above. However, the weight assignment procedures, indicator selection, dimensions and poverty cutoffs decision, identification aggregation and other pertinent concepts remain the same. \\
3386
3387
3388Figure 32 below shows the poverty headcount ratio (H) map for the MPI using the 2013-2014 survey data. The general arrangement of poverty is similar to the MPI poverty indices obtained via imputation. Considering the four major Regional States in Ethiopia (we single out these regional states owing to data representatives and more than 80\% population and political map coverage), MPI poverty head count ratio is much higher in Oromia and SNNP regional states while relatively lower in Amhara and Tigray regional states consecutively ranked from the relatively highest to the relatively lowest MPI poverty headcount ratio.\\
3389
3390Considering the other regions, MPi poverty headcount ratio is much high in Somali and Afar regional states. Whereas, it is relativelly much lower in Benshangul-Gumuz and Gambela People regional States. One vital difference with MPI estimates presented in from table (10)-(15) is that MPI poverty headcount ratio i the Afar regional state is now much higher as compared to results presented earlier. Besides, MPI povverty headcount ratio in Harari People and Dire-Dawa city administration cannot be easily seen due to their smallness. 
3391 \!
3392 \begin{landscape}
3393 \begin{figure}
3394 \caption{Poverty Headcount Ratio \textit{(H)} at regional level in 2013-2014}
3395 \label{figure32}
3396 \includegraphics[scale=1.5]{Graph_Incidence(H)_Poverty_2017.PDF}
3397 \end{figure}
3398 \end{landscape}
3399
3400\subsubsection{ Adjusted Poverty Ratio Subgroup Decomposition and Mapping Across Regional States}   
3401We succinctly discussed how overall poverty as can be disintegrated across various population subgroups, and mapped for visual policy analysis. \\
3402Figure 33 below, show the MPI poverty adjusted headcount ratio of Ethiopian regional states. 
3403Looking at figure 33 below, it can be deduced that the MPI poverty adjusted headcount ratio ($M_0$) is much higher in Oromia regional States followed by SNNP, Amhara and Tigray regional States. One important feature of this figure is that, it can be easily seen the difference in MPI poverty adjusted headcount ratio between Oromia and SNNP regional States (although they have closely similar MPI poverty headcount ratio in figure 32)mapping. \\
3404 
3405Another insightful result is that there is no much difference Amhara and Tigray regional states considering the adjusted headcount ratio mapping. Like wise, between SNNP and Amhara.\\
3406 Regarding the other regional, Somali has much higher MPI ppoverty adjusted headcount ratio mapping, while there is a vivid difference in this regard with Afar regional state. 
3407
3408\begin{landscape}
3409\begin{figure}
3410\caption{Adjusted Headcount Ratio ($M_0$) Disaggregated by Regional States}
3411\label{figure33}
3412\includegraphics[scale=1.4]{Graph_Eth_National_MPI(M0)_2017.PDF}
3413\end{figure}
3414\end{landscape}
3415%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3416Figure 34 below shows the union of MPI poverty incidence mapping and MPI adjusted headcount ratio at regional state level using MPI ($M_0$) with Bar Graphs relative to the size of MPI, incidence of MPI \textit{(H)} with Pie Chart over the regional states and dot mapping relative to the size of MPI  incidence ratio. 
3417%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3418\begin{landscape}
3419\begin{figure}
3420\caption{Union of MPI  Poverty Incidence \textit{(H)} and $M_0$ Mapping among Regional States of Ethiopia}
3421\label{figure34}
3422\includegraphics[scale=1.5]{Graph_Union_EthMPI_2017.PDF}
3423\end{figure}
3424\end{landscape}
3425\subsection{ Analysis of the Dominance Tests}
3426We contend that while there is an extensive use of the stochastic dominance analysis in poverty analysis and other research area in general, there is no empirical literature that scrutinizes the stochastic dominance analysis in the study area and limited application of this concept in multidimensional poverty analysis setting save for limited emerging literature recently, example Duclos et al.(2006a)  employed the multidimensional poverty, MPI analysis so as to obtain evidence on the sensitivity of spatial poverty ordering to choice of multidimensional poverty cutoff and MPI indices. Besides, in (2008), they claimed that a test for robust multidimensional poverty comparisons across countries and regional states can be carried out.\\
3427
3428In this study too, we computed sampling distribution of the MPI indices to allow statistical tests of difference in MPI poverty measures. 
3429
3430\subsubsection{Robustness, Sensitivity and Standard Errors }
3431Robustness tests are based on the coefficient of rank correlations Kendall tau-b, which measures the association between pairs, given the position that each takes when results are  sorted using different poverty indexes.\\
3432
3433These different poverty indexes can be obtained changing the weights indicators or the poverty cut-off (k). One common technique is employing difference weights and testing the MPI indices estimation results. One typical practice is testing for variations in weights that several MPI are computed keeping dimensions/indicators and deprivations cut-offs unchanged; only the weights are modified. Once all the MPI have been computed, figures by sub-national regions can be obtained and regions ranked. The Kendall tau-b coefficient can then be computed over the rankings.\\
3434
3435Another important feature is variations in the poverty cut-offs (k): several MPI are computed keeping the structure unchanged and also adjusting the k-value. Once all the MPI have been computed, figures by sub-national regions can be obtained and regions ranked. The Kendall tau-b coefficient can then be computed over the rankings. With this context, we conducted the dominance test to examine whether the over all poverty is larger or smaller in one regional state than another regional states of Ethiopia considering the nine Administration Regional States and two City Councils. Figure 34 below, depicts the result of dominance test among all Ethiopian regional states. As can be seen from figure 34 ,  it can be shown from the figure 34 below that although the Somali region strictly dominates all other regions, similar propositions cannot be made for the other regions as all regions below do not stochastically  dominate each other.  
3436\;\begin{landscape}
3437\begin{figure}
3438\caption{MPI indices Robustness Test for all Ethiopian Regions States}
3439\label{figure35}
3440\includegraphics[scale=1.8]{Graph_MPI_Robustness_All_Regions.PDF}
3441\end{figure}
3442\end{landscape}
3443Figure 36 below shows, MPI poverty incidence dominance test among five major Ethiopian Regional States. To tackle the problem stated under figure 35 above, we reduced the regional state in to five (choose these five regions as they cover over 90\% of the geographical and political map and the Ethiopian population). \\
3444
3445It can be inferred from this that the MPI in Oromia region is higher that the other major regions (Tigray, Amhara and SNNP) in Ethiopia. In other words, MPI poverty incidence in Oromia regional state strictly dominates, MPI poverty incidence in SNNP; SNNP dominates, MPI poverty incidence in Amhara; and Amhara's MPI poverty incidence dominates, MPI poverty incidence in Tigray and vice versa. 
3446 \begin{landscape}
3447 \begin{figure}
3448 \caption{MPI Poverty Incidence Dominance test among Five Major Ethiopian Regional States}
3449 \label{figure36}
3450 \includegraphics[scale=2.5]{Graph_Dominance_Test_Poverty(H)_2017.PDF}
3451\end{figure}
3452\end{landscape}
3453\subsubsection{Ethiopian National and regional MPI Poverty Incidence Test of Statistical Inference, Confidence Interval(CI) and Standards Errors (SE)}
3454Inferential Statistical statistics such as standard errors (SE) and confidence intervals (CI) examine with inference bout populations based on the behavior of sample. We conducted both SEs and CIs test as the dominance test alone cannot provide sufficient support for accepting or not accepting the null hypothesis stated at the national level and regional level. 
3455
3456Table 14 below depicts the CI and SE test results of difference in MPI poverty (adjusted headcount ratio $(M_0)$  across Ethiopian regional states.  All results are significant at 1\% level of significance. \\
3457
3458Moreover, table 15 below presents tests of Statistical Inference of MPI $(M_0)$ results differences among Ethiopian regional states. Both estimation results showed below assist us to decide whether to accept or not accept the null hypothesis we stated above and achieve the last specific objectives this study.
3459%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3460\begin{landscape}
3461\begin{table}
3462\caption{Test Results of MPI Differences between Ethiopian Regional State Administrations}
3463\label{table 14}
3464\begin{tabular}{p{5cm}|| p{3.5cm} p{5cm} p{3.5cm} p{3.5cm}}
3465\hline
3466\multicolumn{5}{c}{MPI($M_0$) TEST  OF MPI DIFFERENCES BETWEEN REGIONS }\\
3467\hline
3468Variable &    Mean & Linearized Std. Err   & [95$\%$ Conf. & Interval] \\
3469\hline
3470MPI(M0), K=40    &                &         &        & \\
3471Tigray           &  0.368***     & 0.0213  & 0.326  & 0.4096\\
3472Afar             &  0.375***     & 0.0122  & 0.331  & 0.4014\\
3473Amhara           & 0.3774***     & 0.01218 & 0.3535 & 0.4014\\
3474Oromia           & 0.4195***     & 0.01220 & 0.3955  & 0.4435\\
3475Somali           & 0.4372***     & 0.01515 &  0.4075 & 0.4670\\
3476Gambela          & 0.3301***     & 0.04424 & 0.2430  & 0.4171\\
3477SNNP             &  0.4040***    & 0.01599 & 0.3726  & 0.4354\\
3478Benshangul-Gumuz & 0.3047****    & 0.03083 & 0.2440  & 0.3654\\
3479Harari           & 0.3905***     & 0.0420  & 0.3078  & 0.4731 \\
3480Diredawa         & 0.4112***      & 0.0449  & 0.3227  & 0.4997\\
3481\hline
3482\end{tabular} 
3483\end{table}
3484\end{landscape}
3485%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3486
3487\begin{landscape}
3488\begin{table}
3489\caption{Tests of Statistical Inference of MPI $(M_0)$ Results Differences Between Regions}
3490\label{table15}
3491\begin{tabular}{p{5cm}|| p{9cm} p{2.5cm}}
3492\hline
3493\multicolumn{3}{c}{Test of Statistical Inference of MPI $(M_0)$ Differences between Regions}\\
3494\hline
3495Test  & Adjusted Wald test & Prob$>$F \\
3496\hline
3497\_b[Tigray]$=$ \_b[Oromia]     & [Mean-MPI]Tigray-[Mean-MPI]Oromia $=$ 0  &   0.035 \\
3498\_b[Amhara]$=$ \_[Oromia] & [Mean-MPI]Amhara-[Mean-MPI]Oromia $=$ 0 & 0.0153 \\
3499\_b[Tigray]$=$ \_b[Somali]     & [Mean-MPI]Tigray-[Mean-MPI]Somali $=$ 0  &   0.0104 \\
3500\_b[Amhara]$=$ \_b[Somali]     & [Mean-MPI]Amhara-[Mean-MPI]Somali $=$ 0  &   0.0038 \\
3501\\
3502\hline
3503\end{tabular}
3504\end{table}
3505\end{landscape} 
3506%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3507 \pagebreak
3508 \begin{thebibliography} {}
3509 \bibliographystyle{Harvard}
3510 \bibitem[OPHIWP101]%website
3511 author: Xiaolin Wang*, Hexia Feng**, Qingjie Xia*** and Sabina Alkire\\
3512  \textit{On the Relationship between Income Poverty and Multidimensional Poverty in China}
3513 \end{thebibliography}
3514%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3515\newpage
3516\section{Estimation Results Interpretations, and  Discussions}
3517\textit{Discussion}:\\
3518The discernment of multidimensional headcount ratio/incidence denoted by ${H=75.6}$ percent can be comprehended as the percentage of people who are poor. Moreover, the expository of the breadth of deprivation (intensity), which  is the average deprivation score,  or percentage of dimensions in which poor people are derived or deprivation score experienced by people in multidimensional poverty, \textit{A} is 51.5 percent.\\
3519 
3520Therefore,the $ M_{0}=H{\times} A $ is the product of the two component discussed above, or the share of the population that is multidimensionally poor, adjusted by the intensity of the deprivation is given by  $M_0=0.386$ can be interpreted as Ethiopian household percentage of deprivations poor people experience, as a share of the possible deprivations that would be experienced if all people were deprived and poor in all dimensions. 
3521%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3522\subsubsection{Implications of Contribution of each Indicator to Ethiopia's MPI}
3523The implications there of for policy makers while targeting the multidimensional poor, evaluating and monitoring policy implementations by exploring at what is influencing the multidimensional deprivations and poverty  conditions in Ethiopia and craft strategies on how to assist how policies can be coordinated to a focused target while responding to such poverty situations.\\
3524
3525Correspondingly, by taking a look at such poverty situations, it can help for policy makers to prioritize the best fits among the envelops of antipoverty social programs while fighting against poverty focusing at a particular indicator at an individual level. As a result, this is considered to be specific feature of MPI and hence its plus benefits over other measure. \\
3526%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3527\subsection{ Examining the Difference Between MPI Incidence of Poverty and Monetary Absolute Poverty measurement in Ethiopia}
3528The main hypothesis we are interested as discussed  in the methodology section was that to assess the MPI incidence of poverty and monetary , the main hypothesis that we are most interested in this study was stated in the research hypothesis denoted by $H_a$  is the multidimensional headcount ratio in Ethiopia does not lie between the ranges of 38.7$\%$ - 29.6$\%$.  In fact, it is much higher than these figures. The same is for urban and rural poverty figures.\\
3529The other hypothesis is the null hypothesis, denoted by $H_0$ which states that the mean poverty headcount ratio in Ethiopian household lie between 38.7$\%$ - 29.6$\%$ nationally. Besides, this null hypothesis asserts that the  absolute poverty headcount for urban and rural residents  ranges from 35.1$\%$ -25.7 $\%$ and from 39.3$\%$-30.4$\%$ respectively.\\
3530
3531${H_0}$: the population mean ${\mu}$ poverty headcount ratio in Ethiopian household lie between 38.7$\%$ - 29.6$\%$.\\
3532${H_a}$: ${\neq}{\mu}$.\\
3533Considering the law of lager numbers and the central limit theorem, we assumed that as $n{\rightarrow}{\infty},$
3534$({\widehat{M_0}}-M_0)$  $\overrightarrow{d}$ 
3535Normal $(0, \, \dfrac{\delta^2_0}{n}),$ where $\delta^2_0={\in[\widehat{c_i}}(k)-M_0]^2$ is the population variance of $M_0$ (Alkire et. al.,  2015: p-244).\\
3536
3537As can be seen in table 14 and table 15,  we rejected the proposed null hypothesis that states there is no significance difference between the national absolute poverty incidence measured using the monetary poverty measurement and MPI headcount ratio or incidence of poverty \textit{H} since the MPI incidence of poverty in Ethiopia is much higher than the monetary poverty measurement absolute incidence of poverty in Ethiopia and it is significant at 1\% level of significance. \\
3538
3539wwwww\\
3540aaaa\\
3541xxxxx\\
3542(More discussion needed here)\\
3543
3544%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3545\subsubsection{MPI Poverty Indices (H\, and $M_0$) Difference Across Ethiopia Regional States }
3546     
3547     \begin{center}
3548    \begin{frame}
3549    Oromia
3550    \end{frame}
3551    \end{center}
3552The Oromia Regional state has been plugged with a continuous upheavals  and crises since November, 2015. Although, not the root cause,  it was sparked by the federal government's integrated plan to expand the capital-Addis Ababa by allegedly seizing land from the surrounding Oromo farmers and dissecting the Oromia Regional State Administration. Consequently, it was witness the highest causalities, some reported suggesting that more than 780 civiliaces were killed, uncountable injuries, migration and psychological terror by government security forces crackdowns, police brutality and using excessive forces that created uncertainty in the horn of Africa in general and in Ethiopia in particular.  The great Oromo Ethiopians take the lions share of the Ethiopian population but also are the home to most of the national wealth produce and natural, human and physical resources. Besides, it contributes the lions share to the over all economy, resources supply and is the hub of business and incubation centers. 
3553
3554Albeit, these being the facts,  it is one  the MPI poorest region as compared with the four big Administrative Regions (Tigrai, Amhara, SNNP and Oromia) with regional representative data. This has a serious policy implications. If we consider distribution of  of the biggest social security in Ethiopia, the productive safety net program, percent of households received assistance by region and place of residence, the Oromia regional stated received 0.9, which is lower than the national level, 3\% and much the three major regional states, Amhara, 4.2\%, SNNP, 3.3\%, Tigrai 11.8\% and other regions, 3.6\%. \\
3555
3556Considering the second null hypothesis that focuses on the difference MPI  indices across Ethiopian regional states that was stated as follws:  suppose  given  two regions, say Oromia and Tigray. The population achievement matrix are denoted by $X_1$ and $X_2$ respectively. We seek to test the null hypothesis $H_0$: $M_{0.1}-M_{0.2}=0,$ which implies that the poverty in region Oromia is not significantly different from the poverty in region Tigray. With regard to any of the three alternatives: i) $H_a : M_{0.1}-M_{0.2}{\neq}0,$ which means that one of the two regions is significantly poorer than the other; or ii) $H_a : M_{0.1}-M_{0.2}>0, $  which means that region Oromia is significantly poorer than region Tigray; or iii) $H_a : M_{0.1}-M_{0.2}<0,$ which mean that region Tigray is significantly poorer than region Oromia and so on. \\
3557For the first alternative, we need to conduct a two-tailed test, while for the other two alternatives, we should perform a one-tailed test.\\
3558
3559Based on the results presented in table 15 above, we do not accept the null hypothesis that states there is no significant difference in adjusted headcount ratio in Ethiopian regional states as there is a significant in adjusted head count ratio between Tigray and Oromia at 5\% level of significance;  MPI adjusted headcount between Amhara and Oromia at 5\% level of significance; MPI adjusted headcount ratio difference  between Tigray and Somali at 5\% level of significance and MPI adjusted headcount ratio between Amhara and Somali at 1\% level of significant. \\
3560
3561Conversely, the MPI adjusted headcount ratio among the five major regions remains insignificant and we cannot  reject the null hypothesis under such scenarios. 
3562aaaaa\\
3563bbbbbbb\\
3564ccccccc\\
3565zzzzzzz\\
3566(More discussions here...)\\
3567
3568%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3569 \section{ Conclusions and Implication for Policy and/or Further Research }
3570    aaaaa\\
3571    wwwww\\
3572    hhhhhh\\
3573    xxxxxx\\
3574    
3575    \section{Conclusions}
3576    xxxxxx\\
3577    wwwwww\\
3578    hhhhhhhh\\
3579            
3580    \section{Implications for Policy and/or further Research }
3581    wwwwwwwwww\\
3582    aaaaaaaaaaaa\\
3583    zzzzzzzzzzzzzz\\
3584    hhhhhhhhhhhhhhhh\\
3585    
3586%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3587\bibliographystyle{plain}
3588\bibliography{abbrvnat}
3589\bibliography{literature/library}
3590\listoffigures{Figure1}
3591\listoftables
3592
3593\printindex
3594
3595\end{document}