! ! administered, revisiting the ESS1 households and an Dynamics of multidimensional poverty additional 1,500 urban households; the panel sample includes rural and small town households only. and wellbeing in Ethiopia The ESS uses a stratified, two-stage sampling scheme. Poverty can be viewed as taking many different forms, Enumeration areas (EAs) were randomly selected in ranging widely over a set of monetary (consumption or proportion to population size; 290 and 43 EAs were income) and nonmonetary dimensions (health and selected from rural and small town areas, respectively, education). Multidimensional poverty (MDP) is one and twelve households were chosen from each EA. such nonmonetary measure of wellbeing and, Tracking between waves was done at the household according to estimates derived from DHS data in 2011, level-- with a low attrition rate of 4.9%-- leading to a Ethiopia is the second poorest country in the world panel sample of 3,776 households. We further restrict using this measure. Using two waves of panel data the final analytical sample to exclude households that from the Ethiopia Socioeconomic Survey (ESS), we are missing information on any of the nine look at trends and dynamics of MDP in rural and small deprivations or real consumption per adult equivalent, town Ethiopia between 2012 and 2014.! for a final balanced sample of 3,197 households. Background!! Methods!! The Oxford Poverty and Human Development We used the OPHI methodology as a guide when Initiative’s (OPHI) Multidimensional Poverty Index constructing k; because the ESS is an extensive survey, (MPI) is a widely used indicator of multidimensional we were able to include nearly all OPHI-defined poverty, along with the corresponding weighted deprivations in our index. We incorporate three deprivation index (k). K aggregates information on a dimensions of wellbeing-- education, health, and living series of deprivations from three dimensions of standards-- with each dimension weighted to represent wellbeing, including health, education, and living one-third of the deprivation index, and each individual standards. While the body of literature on poverty indicator weighted equally within a given dimension. K dynamics is extensive, the majority of studies draw takes some value between 0 and 1, with 0 indicating no conclusions only about the dynamics of income- or deprivations and 1 signifying deprivation in every consumption-based poverty. However, recent indicator. literature documents that people who are identified as poor in the consumption space are often different In order to classify a household as poor or non-poor, from those who are multi-dimensionally poor. a minimum number of weighted dimensions are established and only those who are deprived in Data-- dimensions exceeding this value are considered poor. We analyze panel data from two waves of the ESS, a In this brief we use the traditional OPHI cutoff of collaboration between the Central Statistics Agency of k>=0.33 in order to make results comparable with Ethiopia (CSA) and the World Bank’s Living Standards external estimates of MDP. Measurement Study- Integrated Surveys of Agriculture (LSMS-ISA) project that collects multi-topic panel data Results-- at the household level. The ESS began in 2011 (ESS1), MDP is a widespread burden in Ethiopia, though it has during which 3,969 rural and small town households declined marginally from 92 percent in 2012 to 88 were surveyed. In 2013, a second wave (ESS2) was percent 2014.1 This decline is the result of nearly twice as many households exiting than entering poverty !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 This decline is statistically significant. during this period (see Table 1). However, 85 percent also become more educated and have acquired more of households are MD-poor at both points in time, and communication, transportation, and other assets of chronic MDP is disproportionately high among rural wellbeing. Positively, we do not observe statistically households; 85 percent of households in rural areas are significant deterioration in any single deprivation (see chronically poor compared to only 33 percent in small Table 2). towns. Table 2. Trends in MPI deprivations 2012 2014 2014- Table 1. MDP Dynamics, k=>0.33 (SEs) (SEs) 2012 Wave 1 Poor Not poor Poor Not poor Obs. 1a. >=1 child age 7-15 not in Wave 2 Poor Poor Not poor Not poor (hholds.) school 0.282 0.288 0.006 1b. No one in household has Total 85.2 3.4 7.0 4.5 3,197 >= 6 years of education 0.677 0.623 -0.054*** 2a. A child age 6-59 months is Rural 85.6 3.4 6.9 4.1 2,799 stunted 0.253 0.222 -0.031** Small 2b. No improved drinking 32.9 6.9 19.3 41.0 398 water access 0.503 0.388 -0.115*** Town 2c. No improved sanitation Observations are weighted to make results representative of all rural and access 0.410 0.424 0.014 small town individuals in Ethiopia. Balanced panel sample size includes 3a. No household electricity 0.925 0.904 -0.021*** 3,197 households in each wave. 3b. Household does not use solid cooking fuel 0.973 0.988 0.015 Figure 1 depicts changes in the distribution of k 3c. No finished floor in between 2012 and 2014. We observe mild household 0.972 0.968 -0.004 improvements across the distribution at the national 3d. Missing community or mobility/livelihood asset 0.612 0.550 -0.062*** level, with mass shifting to the left. We also find that Note: Difference significant at *p<0.1; **p<0.05; ***p<0.01. more individuals experienced a decline (improvement) Observations are weighted to make results representative of all rural in k than did those that accumulated deprivations over and small town individuals in Ethiopia. Balanced panel sample size the same period (47 vs. 33 percent), though most shifts includes 3,197 households in each wave. were minute and 20 percent of the population experienced no change at all in k. Discussion-&-Policy-Implications- More and more policymakers are looking to Figure 1. Distribution of k, 2012 and 2014 nonmonetary measures of poverty when assessing Cross-sectional distribution of k progress toward improvements in wellbeing. Understanding changes in MDP and the underlying 15 measure of deprivation at the individual level is crucial for developing effective and targeted policies. We find that certain facets of wellbeing are more susceptible to 10 Percent changes over a two-year period than others, and thus warrant policy-makers’ attention when attempting to 5 raise rural and small town households out of multi- dimensional poverty. Generous funding assistance for this research came from the 0 0 .2 .4 .6 .8 1 Weigthed deprivation index UK Department for International Development Ethiopia. ESS1 (2012) The findings outlined in this brief are drawn from: Seff, I. & ESS2 (2014) Jolliffe, D. (forthcoming) “Multidimensional poverty dynamics in Ethiopia: How do they differ from consumption-based Among the individual deprivations comprising k, poverty dynamics?” having no access to an improved source of drinking To access the ESS data: water saw the largest decline from 50.3 percent in 2012 hhtp://go.wordlbank.org/ZK2ZDZYDD0! to 38.8 percent in 2014. On average, households have !!!!!!!!!!!! ! !