Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

Srinivas Goli; Nagendra Kumar Maurya; Moradhvaj; Prem Bhandari

doi:10.1177/2158244019837458

Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

Goli, Srinivas; Maurya, Nagendra Kumar; Moradhvaj,; Bhandari, Prem 2019-03-13 00:00:00 This article examines the extent of regional inequality in multidimensional poverty in Nepal using the nationally representative Nepal Demographic Health Survey (2011) data. The authors present a more robust method of multidimensional poverty index (MPI), particularly in terms of the procedure of estimation and aggregation of the indicators as compared with previous studies. The findings suggest that despite the relatively better economic progress and a considerable reduction in education and health poverty, there is a wide inequality across the regions. Far less has been achieved in the case of reducing the standard of living poverty, that is, wealth poverty and inequalities across the regions. The article finds that global MPI tends to inflate poverty estimates in the case of Nepal. It also suggests that development policies and poverty reduction programs in Nepal must aim to reduce multidimensional poverty, of which deprivation in education, health and basic amenities must be an integral component, along with their efforts to improve economic growth and reduce income poverty. Keywords multidimensional poverty, Nepal, MPI, DHS, regional differentials & Human Development Initiative [OPHI], 2013; Rangarajan Introduction & Dev, 2015; Reyles, 2010; Rogan, 2016; Wang & Wang, During the last one decade, Nepal has gone through a major 2016; World Bank, 2014; Yu, 2013). Nepal uses the concept political transition. Abolition of monarchy, the establishment of absolute poverty and has followed its own definition: of a Federal Democratic Republic, the election of Constituent According to which a person earning less than 1 US$ a day is Assembly in 2008 (and reelection in 2014) and the adoption termed as poor. By this definition, the latest official figures of the new constitution in September 2015 are the landmarks suggest that more than 35% of the total population is living in the political history and economic planning of Nepal. The below the poverty line in Nepal. However, the application of country is making every effort to move out of an extended the concept of relative poverty is virtually absent in Nepal political transition, and is also aiming to become a developed (Alkire, Adriana, & Roche, 2013; Central Bureau of Statistics country in the world by 2022 (United Nations Nepal, 2014). [CBS], 2011; Nepal Human Development Report [NHDR], The ambitious journey of transition from a comparatively 2014; Uematsu, Shidiq, & Tiwari, 2016; World Bank, 2016). less developed country to a developing and then to a devel- Earlier, to achieve million development goals (MDGs), oped nation demands a concerted effort for a holistic and now, to achieve sustainable development goals (SDGs), approach to the development. This is not possible without a Nepal has been fairly active in investing in social policies significant reduction in the incidence of both absolute and including poverty reduction and active society engagement relative multidimensional poverty in the country. (Government of Nepal, 2011; Uematsu et al., 2016). This has Measurement of poverty itself is a complex and highly debat- able issue. Social scientists in different countries have Jawaharlal Nehru University, New Delhi, India adopted different dimensions to determine poverty status of 2 University of Lucknow, India the populations (see, Angulo, Díaz, & Pardo, 2016; Bader, University of Michigan, Ann Arbor, USA Bieri, Wiesmann, & Heinimann, 2016; Dhongde & Haveman, Corresponding Author: 2017; Dutta, 2015; Guio, Gordon, & Marlier, 2012; Guio Srinivas Goli, Assistant Professor, Center for the Study of Regional et al., 2016; Hanandita & Tampubolon, 2016; Montoya & Development, Jawaharlal Nehru University, New Delhi 110067, India. Teixeira, 2017; Nowak & Scheicher, 2017; Oxford Poverty Email: sirispeaks2u@gmail.com Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open resulted in a significant reduction in poverty (Uematsu et al., Guio et al., 2012; Guio et al., 2016; Hanandita & Tampubolon, 2016). For instance, according to an estimate, the percentage 2016; Mohanty et al., 2017; Montoya & Teixeira, 2017; of the multidimensional poor in Nepal has dropped signifi- Nowak & Scheicher, 2017; Rangarajan & Dev, 2015; Rogan, cantly from 64.7% to 44.2% between 2006 and 2011, that is, 2016; Tsui, 2002; Wang & Wang, 2016; World Bank, 2014; by 4.1 percentage points per year (Alkire et al., 2013). In Yu, 2013). A majority of these studies used methods designed fact, the country has been able to reduce the national poverty for global MPI calculations (Alkire & Foster, 2007; Alkire, much faster than its neighboring countries such as India, Foster, & Santos, 2011; Alkire & Santos, 2013). Pakistan, and Bangladesh (Drèze & Sen, 2013). In spite of Advancing the existing methodology of selection of considerable progress in poverty reduction in recent years, parameters in multidimensional poverty, Guio et al. (2012) Nepal remains one of the poorest countries in the world. and Guio et al. (2016) proposed an analytical framework for With a human development index (HDI) of 0.548 in 2014, developing robust material deprivation indicators for the Nepal is ranked 145th out of 187 countries listed in the whole population in the context of the European Union. They United Nations Development Programme (UNDP; 2015). carried out systematic item-by-item analyses at national lev- The National Living Standards Survey (NLSS) conducted in els to identify material deprivation criteria, which satisfacto- 2010-2011 reported that more than 30% of Nepalese live on rily meet suitability, validity, reliability, and additive criteria less than US$14 per person per month using the income- across the European Union. There have been some efforts to based poverty estimation. However, this figure rises to 44.2% include multiple indicators in measuring poverty in Nepal in the case of multidimensional poor. Furthermore, there is a (Alkire et al., 2013; CBS, 2013; Mitra, 2016; NHDR, 2014). large inequality in the prevalence of poverty within the However, we understand that these Nepal-specific estimates nation. Although the overall poverty rate for Nepal is 30%, of multidimensional poverty have methodological limita- this figure rises to 45% in the mid-western region and to 46% tions both in terms of estimation procedures and the use of in the far-western region (NLSS 2010-2011). Thus, a indicators and their units of measurement. This study is an national-level figure often obscures the within-country attempt to refine the measure of multidimensional poverty inequality in poverty (Uematsu et al., 2016). both regarding its dimensionality and estimation procedure and to fill this gap in the literature, specifically in the context of Nepal. We provide a theoretical background and the evi- Background and Rationale dence of multidimensional poverty measures used in previ- Measuring poverty is a complicated process (Sen, 1979; ous studies below. Then, we describe data, our proposed Townsend, 1954, 1971, 1979). Early efforts of measuring method and its advantage over the existing method, indica- poverty involved unidimensional indicators based on income tors of alternative dimensions, and the difference in the pro- or consumption expenditure (Abel-Smith & Townsend, cedure used to estimate the multidimensional poverty in this 1965; Atkinson, 1987; Atkinson, 1970; Bosanquet, 1903; study as compared with other studies in the context of Nepal Clark, Hemming, & Ulph, 1981; Hagenaars, 1987; Kakwani, (Alkire et al., 2013; CBS, 2013; Mitra, 2016; NHDR, 2014; 1980; Ravallion, 1998; Ravallion & Huppi, 1991; Sen, 1976, Uematsu et al., 2016). 1981, 1987, 1989; Thon, 1979; Townsend, 1954, 1971, 1979). Later, it was recognized that no single indicator alone Measuring Poverty: Past Efforts and could capture the multiple aspects of poverty (Foster, Greer, Our Approach & Thorbecke, 1984; Townsend, 1979). Poverty is much more than having a low income or low consumption expenditure Debates on measuring poverty were intensified in the 1970s, (Anand & Sen, 1997; Bourguignon & Chakravarty, 2003; but these discussions were mainly about measuring income Sen, 1970; Townsend, 1954). poverty and defining the poverty line. During this period, the Realizing the significance of multiple indicators, there identification of the poor was exclusively by family-size- have been some efforts to include multiple indicators in mea- adjusted household income, concerning a specified income suring poverty. The first multidimensional measure can be poverty line. Some contributions are worth mentioning here, traced back to Townsend (1979), and the underpinnings of for example, Townsend (1954, 1971, 1979), Sen (1970, the multidimensional poverty index (MPI) were set out by 1972, 1973, 1992, 2000b), Bardhan (1970, 1971), Dasgupta, Foster et al. (1984). The global MPI was designed in 2010 by Sen, and Starrett (1973), Gordon and Townsend (1990). Sen the OPHI and the UNDP using different indicators to deter- (1976), in his seminal article “Poverty: An Ordinal Approach mine poverty beyond income-based measures (OPHI, 2013). to Measurement,” has emphasized the theoretical soundness This MPI replaced the previous human poverty index in sub- of the income poverty measurement. He has suggested an sequent human development reports of the world countries. ordinal approach based on ordinal axioms for measuring Following this, numerous studies in many countries have poverty. However, he admitted that such an approach is dif- used various procedures to estimate multidimensional pov- ficult to replicate in reality, as required data may not be avail- erty of individuals and households (Angulo et al., 2016; able. From the mid-1970s, it was recognized that poverty is Bader et al., 2016; Dhongde & Haveman, 2017; Dutta, 2015; much more than just having a little income (Townsend, Goli et al. 3 2010). During this time, the “basic needs approach,” social societies and their state of affairs (Alkire, Foster, & Santos, exclusion, and “capability approaches” gained prominence 2011). Thus, a key criticism of global MPI used by UNDP is in complementing the process of identification of the poor or a lack of examination of its suitability within countries, deprived populations. Studies have shown that income does dimensional structure reliability and validity and measure- not represent the nonmonetary multidimensional depriva- ment invariance apart from the ambiguity in their household- tions of households (such as lack of access to nutritious food, level aggregation procedure of individual and household health services, quality education, potable water, livable indicators. Some of these are discussed in detail in the house, sanitation facilities, electricity, basic information and method section of this study (also see Guio et al., 2012; Guio more), and thus fails to identify the poor correctly (Sen, et al., 2016). A study by Betti et al. (2018) on simplified 1970, 1972, 1973, 1976; Townsend, 1979). Consequently, Jacknife variance estimates for measures of MPI is a seminal researchers have introduced various nonmonetary measures contribution to the literature. of deprivation, supplementing these multidimensional analy- In Nepal, to the best of our knowledge, the present exer- ses with monetary measures to create a better overall picture cise is the third attempt to measure MPI at a disaggregated of poverty (Foster et al., 1984; Townsend, 1979). For level. The first attempt was by Alkire and her team at OPHI instance, Townsend (1979) highlighted “relative depriva- who estimated MPI for 104 countries including Nepal tion” by which he meant an absence or inadequacy of those (Alkire, Roche, Santos, & Seth, 2011). In this study, they diets, amenities, standards, services, and activities, which are included years of schooling and school attendance for the common or customary in society. This could be understood education dimension, child mortality and nutritional status as the initial debate on a multidimensional aspect of poverty for the health dimension, and cooking fuel, sanitation, water, and deprivation. His measure of deprivation included a list of electricity, and floor and asset ownership for the standard of 60 indicators of the standard of living. The indicators cov- living. They estimated MPI by place of residence and devel- ered diet, clothing, fuel and light, house and housing ameni- opmental region (OPHI, 2013). Their findings suggested that ties, and so on. Sen (2000a), endorsing the need to take a 44.2% of people could be classified as poor in 2011, if pov- multidimensional approach to poverty, writes “Human lives erty was assessed multidimensional. This estimate was much are battered and diminished in all kinds of different ways, higher than the income-based poverty (25.2%) of NLSS and the first task is to acknowledge that deprivations of very (2010-2011). The second study is the NHDR (2014) released different kinds have to be accommodated within a general by the Government of Nepal in collaboration with UNDP. In overarching framework” (p. 18). Thus, a poverty measure this report, human poverty index was estimated (31.1%) based on multiple indicators is more robust than the one using the percentage of people who were not expected to sur- measured based solely on income poverty (Betti, Gagliardi, vive beyond age of 40 years for a long and healthy life; adult & Verma, 2018; Deaton, 2013; Drèze & Sen, 2013; Gordon literacy rate for knowledge; and the percentage of people & Townsend, 1990; Guio et al., 2012; Guio et al., 2016). This without access to safe water, percentage of malnourished suggests that the poverty debate has moved from a unidimen- children below 5 years of age and deprivation in economic sional (income) to a multidimensional approach, which led provisioning for a decent standard of living. Human poverty to the estimation of MPI. index was estimated separately for a rural–urban place of Foster et al. (1984) proposed a systematic methodology residence, five development regions, three ecological zones, for estimating multidimensional poverty for the first time. subregions, and districts. Their approach fueled the debate of measuring poverty and In this study, as used in previous approaches, our measure provided a framework for decomposing poverty. They of multidimensional poverty includes three basic dimensions showed new poverty measure that is additively decompos- of human development—health, education, and economic able with population share weights. Although their work was status. However, our estimates of MPI differ significantly one of the most cited in poverty literature due to its notable from the NHDR (2014) and Alkire, Roche, and Vaz (2014) methodological contribution for decomposing the indicators regarding inclusion and coverage of indicators, their estima- of measuring poverty, it however, did not contribute much in tion procedure, and aggregation of each dimension of MPI. measuring multidimensional poverty. Later on, Alkire and First, we include net enrolment ratio instead of gross enrol- Foster (2007) made a significant methodological ment ratio under education dimension. Second, wealth index improvement. is used to measure economic status as compared with indi- In 2010, Alkire and Santos developed MPI for various vidual standard of living indicators. Finally, each indicator countries for the first time under OPHI. Multidimensional under different dimensions has been measured at the indi- poverty comprises of a number of indicators showing the vidual level rather than at household level. We also carried deprivations experienced by people—such as poor health, out a rigorous screening of variables to identify multiple lack of education, inadequate living standard, lack of income, deprivation items, which adequately meet suitability, valid- disempowerment, poor quality of work and threat from vio- ity, and reliability of the assumed latent dimensions of MPI. lence (Alkire & Foster, 2011a, 2011b). These indicators may We have explained in detail below about the robustness and vary depending upon the context of countries, cultures or inclusiveness of our method of estimation of MPI over 4 SAGE Open NHDR (2014) and Alkire, Roche, et al. (2011) in the context deprivations. Thus, it was important to confirm whether they of Nepal, which is also applicable in the global context. This were suitable to and representative of a single latent dimen- study has largely benefited from the conceptual and method- sion of the standard of living or not. For each of these three ological advancements made in recent studies (e.g., Guio aspects, item selection criteria were not only drawn from the et al., 2012; Guio et al., 2016; Notten & Mendelson, 2016). literature on theoretical basis, but also empirically tested. We also disaggregated multidimensional poverty estimates The items that have successfully passed principles of reli- by place of residence, developmental region, ecological ability were then used in aggregation to MPI (Table 1). In zone, and subregions. addition to this, we have also tested for validity. However, in absence of income data in the same data set, we have assessed it through the macrolevel correlation between individual Data MPI components and income poverty levels (Figure 1). This study has used data from a nationally representative Nepal Demographic Health Survey (DHS) collected in 2010- Education 2011. The sample was designed to provide estimates of most key variables for the 13 ecodevelopment regions and three NHDR (2014) used adult literacy rate for measuring educa- ecological regions (for further details on sampling design, tional deprivation. This measure has been severally criticized see Ministry of Health and Population [MOHP], New ERA, for being not a true indicator of measuring educational depri- & ICF International, 2012). The survey covered a nationally vation (Smith, 1992; UNDP, 1993). Therefore, studies are representative sample of 10,826 households, which yielded increasingly using years of schooling and school enrolment completed interviews with 12,674 women aged 15 years to to measure educational deprivation. Alkire, Roche, et al. 49 years in all selected households and with 4,121 men aged (2011) used both years of schooling and school enrolment as 15 years to 49 years in every second household. What is of indicators of educational deprivation. While the present particular interest to this study is that these data provide study also uses years of schooling and child enrolment ratio, detailed information on mortality, nutrition, sociodemo- but unlike Alkire, Roche, et al. (2011) who used gross enrol- graphic characteristics, access to basic amenities, and house- ments, we have used net enrolments. We used net enrolment hold assets. Furthermore, we have used income poverty and to overcome the limitations of gross enrolment ratio, which multidimensional poverty information from CBS (2011) and could be more than 100% and often fails to measure the true NHDR (2014), respectively, in the validity test of MPI items enrolment ratio. Reliability analyses for the items included in chosen for this study. the dimension of education suggest a high reliability coeffi- cient (RC = 0.8078), meaning that items considered under education dimension probably measure the same underlying Measure of Multidimensional latent concept “education.” Poverty: Rethinking Dimensions and Computation Health We have estimated MPI using information from indicators of three main dimensions—education, access to health, and While NHDR (2009) used percent of people not expected to standard of living. However, within these dimensions, our survive beyond 40 years of age as an indicator of health sta- indicators are not the same as used by Alkire, Roche, et al. tus, both Alkire, Roche, et al. (2011) and we used two indica- (2011) and NHDR (2014). tors of health: under-five mortality status and nutritional The dimensions and indicators within each dimension are status. presented in Table 1. We have described the measurement of Under-five mortality is measured as whether a child expe- dimensions and indicators below. The previous studies have rienced death before his or her fifth birthday. Our other mea- used both theoretical and empirical criteria in identifying sures of health are the nutritional status of the child and the suitable items to be included in the MPI (Dickes, 1989; mother. The most convenient and standard method of mea- Eurostat, 2002; Guio, 2009; Perry, 2002; Whelan, 1993). In suring nutritional status of a child is his or her physical this study, suitability is theoretically derived as aptness to growth or the weight-for-age. This index provides informa- reflect low well-being of people across the different prov- tion about growth and body composition and is measured in inces of Nepal. Each indicator within a specific dimension terms of standard deviation units (Z scores) from the median was selected based on reviews of extensive literature in the of the reference population. In addition, we have also consid- context of Nepal. For example, in the case of the first two ered height-for-age of children below 5 years of age to repre- dimensions, suitable items were selected qualitatively that sent chronic undernutrition. Children whose standard score represent a single latent dimension: education or health. (Z score) of weight-for-age (underweight) and height-for-age However, the next dimension, standard of living, was derived (stunting) is below minus two standard deviations (–2 SD) based on factor analyses because there were a large number from the median of the reference population are considered of items that represented different kinds of material as malnourished. For analysis, a dichotomous variable Goli et al. 5 Table 1. Dimensions, Indicators, and Measures of MPI and Their RCs. Reliability analyses NHDR Alkire, Roche, Santos, Number Dimensions Indicators (2014) and Seth (2011) Present study—Measures RC of items Education Adult literacy rate Yes — — — 0.8078 2 Years of schooling — Yes Yes If completed less than 5 years of schooling School enrollment — Yes Yes If a child (aged 6-10 years) is not enrolled in school Health People not expected Yes — — — 0.7220 4 to survive beyond the age of 40 years Under-five mortality — Yes Yes If died before the fifth birthday Nutritional status — Yes Yes If child is underweight (child and woman) (–2 SD from median of reference population) If child has stunted growth (–2 SD from median of reference population) If a woman has BMI < 18.5 kg/m Standard of Malnourished children Yes — Yes — living below the age of 5 years Safe drinking water Yes Yes Yes — Sanitation — Yes Yes — Electricity — Yes Yes — Flooring — Yes Yes — Cooking fuel — Yes Yes — Assets ownership — Yes Yes — Wealth index — — Yes It is a composite index of 0.5796 30 30 household amenities and assets MPI Education (two items), 0.7453 36 health (four items), and wealth Index (30 items) Note. MPI = multidimensional poverty index; NHDR = Nepal Human Development Report; RC = reliability coefficient; BMI = body mass index. This indicator in the present study was used as a part of health dimension. These indicators are being used in the construction of wealth index, a widely known proxy measure of economic status. whether an under-five child’s growth was below –2 SD Wealth Index (coded 0) or above –2 SD (coded 1) for weight-for-age and Household wealth index is used as an indicator of standard height-for-age was created. of living. Wealth index is commonly used in DHS country- Mother’s nutritional status is measured by body mass level studies and considered as a most reliable summary index (BMI). The BMI is categorized into three components: measure (Rutstein, Jhonson, & Gwatkin, 2000). For the pur- thin (BMI < 18.5), normal (BMI between 18.5 and 24.9), pose of this study, wealth index computed in Nepal DHS and obese (BMI > 25). In this study, we considered thin (2011) has been considered as an indicator of standard of women with a BMI less than 18.5 as malnourished. Our indi- living. The coverage of wealth index (MOHP, New ERA, & cators are same as taken by the Alkire, Roche, et al. (2011). ICF International, 2012; Rutstein, 1999, 2008) is much In the case of the dimension of health, reliability analyses for more than the six standards of living indicators (i.e., access the items included reveal a high RC (.7220), meaning that to safe water, sanitation, electricity, flooring, cooking fuel, items considered under health dimension probably measure and assets ownership) used by Alkire, Roche, et al. (2011). the same underlying latent concept “health.” 6 SAGE Open Figure 1. Correlation matrix for the different components of MPI with income-based poverty. Note. MPI = multidimensional poverty index. The wealth index in its current form, which takes a better dividing the ranking into five equal categories, each com- account of urban–rural differences in the scores and indica- prising 20% of the population. In this study, bottom two tors of wealth in the Nepal DHS (2011), is created in three quintiles, poorest and poorer (40%), form the poor. This clas- steps. First, a subset of indicators common to urban and rural sification is widely accepted in demographic research areas was used to create wealth scores for households in both because it has been repeatedly proven in the DHS data–based areas (for a detailed list of indicators see MOHP, New ERA, analyses that these two quintiles show deprived status in sev- & ICF International, 2012; Rutstein, 2008). Categorical vari- eral demographic indicators than other three quintiles ables were transformed into dichotomous (0-1) indicators. (Rutstein, 2008). These indicators were then examined using principal compo- The wealth index is particularly valuable in countries, nent analysis (PCA) method to produce a common factor which lack reliable data on income and expenditures, which score for each household. Second, separate factor scores are the traditional indicators used to measure household eco- were produced for households in urban and rural areas using nomic status (Chakraborty, Fry, Behl, & Longfield, 2016). For an area-specific indicator. Third, the separate area-specific developing countries such as Nepal, with extreme geographi- factor scores were combined to produce a nationally appli- cal difficulties, high poverty incidence, and not so strong sta- cable wealth index by adjusting area-specific scores through tistical information system, wealth index may serve a better a regression on the factor scores. This three-step procedure purpose for measuring economic status (standard of living) permits greater adaptability of the wealth index in both urban than direct income (Gwatkin et al., 2007; Rutstein & Johnson, and rural areas. The resulting combined wealth index has a 2004). However, the recent studies (Guio et al., 2012; Guio mean of zero and a standard deviation of one. et al., 2016) have suggested for testing the reliability and Once the index is computed, national-level wealth quin- validity of dimensional structures and suitability of items used tiles (from lowest to highest) were obtained by assigning the in the construction of material deprivation index such as household score to each de jure household member, ranking wealth index. Following them, we have performed the reli- each person in the population by his or her score, and then ability analyses for the 30 items used in the wealth index Goli et al. 7 construction. The results of reliability analyses suggest, reasons. First, when we estimate child nutritional and child although slightly less, but a satisfactory RC (0.58) between the mortality indicators with the entire sample of child popula- items selected, meaning they correlate to an assumed latent tion as a denominator instead of households, it will overcome concept of “standard of living” in the study. the limitation of excluding the households, which do not have under-five children at the time of the survey. Second, this process avoids the duplication of deprivation in terms of Estimation of MPI health and education across household members. In the case Compared to the initial period of Townsend (1979) and of the index of standard of living, we have estimated it at the Foster et al. (1984), the measure of multidimensional pov- household level and generalized the household index value erty has been advanceing in terms of its dimensional spread to all the household members living in that particular house- and the number of items used in each dimension. In particu- hold by exporting this variable to person file. lar, the introduction of the DHS, a homogeneous and reliable While estimating the composite index, the UNDP (2010) household survey over 80 developing countries, has facili- human development report used geometric mean (GM) to tated the researchers to construct the index on unit-level data obtain HDI. We believe that this method of estimating GM of sets using a wider number of indicators related to both house- each dimension of indices (education, health and standard of holds and individuals. The global MPI becomes widely used living) is a better procedure to derive MPI rather than the measure since 2010 after its regular publication in human simple mean as used by Alkire and Foster (2007) and Alkire, development reports of UNDP. Alkire, Roche, et al. (2011) Roche, et al. (2011). Because GM makes sure that a low have estimated global MPI using three dimensions: educa- achievement in one dimension is not linearly compensated tion, health, and living standard. As per this approach, a per- by high achievement in another dimension. The GM reduces son on each indicator is identified as deprived or not deprived the level of substitutability between dimensions and at the using information for any one household member. Then, it is same time ensures that 1% decline in the index, say, life aggregated across all the household members. This criterion expectancy at birth has the same impact on the MPI as 1% of identifying poor and calculating MPI in their method has decline in education or income index. serious drawbacks. Researchers have often raised questions Thus, we estimate MPI using GM method by the about measurability and aggregation process of indicators in following: the method of MPI calculation proposed by Alkire and her colleagues (Rangarajan & Dev, 2015). For instance, if one MPI = EPIH ×× PI SLPI , (1) () person in a household is undernourished that does not mean all household members are undernourished. Similarly, if one child has not attended school for 5 years or more that does where EPI is education poverty index, HPI is health poverty not mean that other children did not go to school. Moreover, index, and SLPI is the standard of living poverty index. For in the case of under-five mortality, if a household does not comparison, we also estimated MPI using simple mean by have under-five children, such households will not be the following: included in that particular dimension of MPI. According to Nepal DHS (2011), the proportion of such households was as () EPI+HPI+SLPI (2) MPI = . high as 11%. This is the main reason that Rangarajan and Dev (2015) have not estimated MPI using Alkire, Roche, et al.’s (2011) identification approach for neighboring devel- oping country “India.” Rangarajan and Dev (2015) also Thus, our methodology is different from Alkire, Roche, questioned the aggregation method of individual- and house- et al. (2011) and NHDR (2014) in terms of indicators used, hold-level indicators by Alkire, Roche, et al. (2011). Some of estimation and aggregation approach. NHDR (2014) used the recent studies by Guio et al. (2016) and Notten and traditional indicators to measure human poverty index, Mendelson (2016) have made conceptual and methodologi- which is severely criticized by the scholars now (Dhongde cal advancements in the construction of material deprivation & Haveman, 2017; Dutta, 2015; Gordon, 2000; Mitra, 2016; index like one which similar to MPI. Rogan, 2016; Rutstein, 2008; Wang & Wang, 2016). For We concur with Rangarajan and Dev (2015), Guio et al. instance, NHDR (2014) used adult literacy rate for educa- (2016), and Notten and Mendelson (2016), and therefore, tion; percentage of people not expected to survive beyond made some modifications in Alkire and others’ criterion of 40 years of age (life expectancy) for health and percentage aggregation of dimension indices to overcome the above- of people not having access to safe drinking water and child said limitations. In particular, we have estimated indicators undernourishment for the standard of living. The adult lit- at the population level for rural–urban, ecological zones and eracy rate is replaced by the mean years of schooling at subnational level instead of household level in all three UNDP HDI because it does not depict the current scenario dimensions, namely, health, education, and standard of liv- of educational attainment in a country or state. Therefore, ing. This method has merit over the previous method for two UNDP includes school enrollments, which represent current 8 SAGE Open scenario. We considered both mean years of schooling and Differences were also observed across ecological zones net school enrollment ratio. Among health indicators, (namely, mountain, hill, and Terai). A large number of people NHDR (2014) fails to take more sophisticated and sensitive across the ecological regions had less than 5 years of schooling. measures such as child mortality and nutrition. Standard of The figure varies from 48.0% in hills to 59.1% in the mountain living poverty is measured by deprivation in access to drink- areas. Surprisingly, children not enrolled (age 6-10 years) ing water, which is acceptable; but the inclusion of child reflects the reverse situation. The highest number of children undernourishment cannot be justified to be included in the not enrolled (age 6-10 years) found to be the highest in Terai standard of living poverty unless they had tested for its suit- region (10.63%) and lowest in mountain region (1.40%). ability and reliability in this dimension. It also fails to incor- By development regions, the highest proportion of indi- porate several other important variables under the dimension viduals with less than 5 years of education were reported in of the standard of living: sanitation, electricity, housing, and the mid-western development region (55.5%) followed by many others. Therefore, our indicator, especially in the case far-western region (54.4%). The lowest proportion was of the standard of living poverty (measured as wealth index), found in the eastern and western development region which incorporates multiple household assets, is a better (47.7%). Except for central region (13.50%), in other regions indicator. children not enrolled (age 6-10 years) were below 10%. The From methodology (dimensional structure and aggregation far-western region recorded the lowest (2.60%) number of process) point of view, we differ with Alkire, Roche, et al. children not enrolled (age 6-10 years). (2011) on three points: number and type of indicators, their Further disaggregation based on ecodevelopmental region estimation procedure, and process of aggregation of dimen- suggests that individuals had less than 5 years of education in sion indices used in the computation of MPI. In terms of indi- the central mountain (63.7%), western mountain (60.5%), cators, we differ only at three places. First, we have used net far-western hills (56.8%), and central Terai (61.2%) are con- enrollment ratio rather than gross enrollment ratio in the edu- siderably high. On the contrary, the proportion of children cation dimension, an additional indicator height-for-age under not enrolled in school was the highest in central Terai nutritional status of children under health dimension and (19.1%) followed by eastern Terai (6.9%) and western Terai wealth index based on 30 assets instead of Alkire and col- (5.6%). These results show that there is variation in educa- leagues’ standard of living index based on six indicators. tional poverty by rural–urban residence, development Second, we also differ with Alkire and colleagues in terms of regions, and ecodevelopmental subregions. estimation procedure of individual indicators. They have esti- mated deprivation basically at a household level even for indi- Distribution of indicators by geographic regions: Health. By place cators, which conventionally supposed to be estimated at the of residence, among health poverty indicators, as far as individual level. However, we estimated each indicator based under-five mortality is concerned, both rural (4.7%) and on the nature of indicator and considered the unit of analysis urban (4.0%) areas still face a dearth of health facilities as accordingly, that is, if it is an individual kind of variable then reflected in high under-five mortality. Gaps are noticeable in measured at the individual level or if it is household variable the case of nutritional indicators also. More than one third of then measured at the household level then the score distributed total children in rural areas are poor regarding nutritional sta- among the household members. tus whereas in urban areas around one fourth of total children are poor. However, as compared to children, nutritional sta- tus of adult females is slightly better. As per our estimates, Results 18.5% females in rural areas and 13.5% females in urban areas have poor nutritional status. Descriptive Statistics By ecological regions, we found maximum adult females Table 2 presents descriptive statistics of indicators covered with BMI < 18.5 in the Terai region (22.5%), but children under each dimension of MPI by place of residence, ecologi- with poor nutritional status were maximum in the mountain cal regions, developmental regions, and ecodevelopment region (height-for-age 52.8%; weight-for-age 35.5%). subregions. The descriptive statistics show a large difference Under-five mortality was the highest in the mountain region in each of the specific indicators of all three dimensions of (6.4%). As against this, hills were found to be the least poor MPI, namely, education, health, and standard of living. in terms of select health indicators. Among developmental regions, broadly mid- and far- Distribution of indicators by geographic regions: Education. By western region show the greatest extent of deprivations on an place of residence, about 54.4% people do not complete 5 average in all health indicators. In the case of under-five years of schooling in rural areas as compared with 32.2% in mortality and adult females with BMI < 18.5 found to be urban areas. On the contrary, children between 6 years and highest in the far-western region (mortality < 5%-6.60%; 10 years of age not enrolled are 7.7% in rural areas as com- and BMI < 18.5%-23.6%). On the contrary, height-for-age pared with 3.1% in urban areas. The result shows a huge gap and weight-for-age are the highest in the mid-western region in rural–urban educational poverty. (height for age 49.0%; weight for age 36.2%). 9 Table 2. Descriptive Statistics of Indicators, Nepal, 2011. Education Health Standard of living Children not enrolled Wealth status poorest Less than 5 year of (aged 6-10 years; n = Under-five death (n = Height for age Weight for age and poor wealth schooling (n = 36,908) 6,305) 5,306) (moderate; n = 2,392) (moderate; n = 2,392) BMI < 18.5 (n = 6,179) quintile (n = 49,791) 95% CI % 95% CI 95% CI 95% CI 95% CI 95% CI 95% CI Background characteristics % LL UL LL UL % LL UL % LL UL % LL UL % LL UL % LL UL Place of residence Urban 32.2 31.0 33.5 3.1 2.0 4.6 4.0 2.6 6.1 25.4 20.1 31.5 15.2 11.1 20.6 13.5 11.4 15.9 6.3 5.7 6.9 Rural 54.4 53.8 54.9 7.7 7.0 8.4 4.7 4.2 5.4 40.6 38.6 42.7 29.1 27.2 31.0 18.5 17.5 19.6 44.6 44.2 45.1 Ecological zone Mountain 59.1 57.2 61.1 1.4 0.6 2.9 6.4 4.5 9.2 52.8 45.6 59.8 35.5 29.0 42.6 16.0 12.7 19.9 71.8 70.3 73.3 Hill 48.0 47.2 48.8 3.7 3.0 4.5 4.6 3.8 5.6 40.6 37.5 43.7 25.4 22.8 28.3 11.8 10.6 13.2 52.5 51.8 53.2 Terai 52.8 52.1 53.5 10.6 9.6 11.7 4.5 3.8 5.3 36.2 33.6 38.9 28.4 26.0 31.0 22.5 21.1 24.0 25.5 24.9 26.0 Development region Eastern 47.7 46.6 48.7 5.1 4.1 6.4 4.7 3.6 6.0 35.9 32.1 39.9 25.0 21.6 28.7 16.2 14.4 18.1 34.6 33.8 35.5 Central 53.7 52.8 54.6 13.5 12.0 15.0 4.5 3.6 5.6 36.0 32.7 39.4 27.6 24.6 30.8 19.7 18.0 21.5 32.2 31.5 32.9 Western 47.7 46.6 48.8 3.4 2.5 4.6 4.1 3.1 5.6 37.3 33.0 42.0 22.1 18.5 26.2 13.7 11.9 15.6 35.7 34.8 36.6 Mid-western 55.5 54.0 57.0 4.2 3.1 5.8 4.2 3.0 5.8 49.0 43.9 54.2 36.2 31.5 41.3 18.4 15.7 21.4 61.2 60.0 62.4 Far-western 54.4 52.8 56.1 2.6 1.6 4.1 6.6 4.9 8.9 45.6 39.7 51.5 32.5 27.2 38.3 23.6 20.5 27.1 57.9 56.6 59.3 Ecodevelopment subregion Eastern mountain 51.8 48.0 55.5 1.3 0.3 5.4 5.4 2.3 12.0 45.3 31.6 59.8 23.2 13.2 37.6 9.4 5.3 16.4 65.4 62.3 68.4 Central mountain 63.7 60.3 67.0 1.4 0.3 5.9 3.9 1.4 10.2 47.1 32.5 62.2 35.9 22.8 51.4 14.4 9.2 21.8 60.0 57.0 62.9 Western mountain 60.5 57.4 63.5 1.4 0.5 4.3 7.9 5.1 12.2 58.5 48.7 67.7 41.0 31.8 50.8 21.8 16.0 28.8 84.3 82.4 86.1 Eastern hill 51.1 49.3 52.9 3.1 1.9 5.1 5.3 3.5 7.9 43.8 36.6 51.2 27.6 21.5 34.6 11.2 8.7 14.4 61.8 60.2 63.3 Central hill 40.6 39.2 42.0 4.7 3.3 6.6 3.6 2.3 5.7 28.2 22.8 34.4 20.3 15.5 26.0 10.9 8.9 13.4 32.8 31.6 34.0 Western hill 48.9 47.4 50.3 1.7 1.0 3.1 4.8 3.3 6.8 36.0 30.5 41.8 15.9 12.0 20.7 8.1 6.3 10.2 48.8 47.5 50.0 Mid-western hill 53.9 51.6 56.1 5.5 3.7 8.2 3.4 2.0 5.8 50.3 42.8 57.8 36.0 29.1 43.6 17.0 13.3 21.6 72.9 71.2 74.5 Far-western hill 56.8 54.0 59.6 3.8 2.1 7.0 6.7 4.2 10.6 57.3 48.1 66.1 40.3 31.7 49.6 22.6 17.3 28.8 79.4 77.5 81.3 Eastern Terai 45.3 44.0 46.7 6.9 5.4 8.9 4.2 3.0 5.9 30.7 26.2 35.7 24.0 19.8 28.7 19.7 17.2 22.4 16.1 15.2 16.9 Central Terai 61.2 60.0 62.3 19.1 17.0 21.4 5.0 3.9 6.4 38.5 34.4 42.8 30.2 26.3 34.3 25.9 23.5 28.5 28.9 28.0 29.8 Western Terai 46.0 44.3 47.7 5.6 3.9 7.9 3.2 1.8 5.4 39.7 32.4 47.4 32.6 25.8 40.1 20.9 17.8 24.4 17.5 16.4 18.6 Mid-western Terai 55.9 53.7 58.1 3.7 2.1 6.3 3.4 1.9 6.2 42.3 34.3 50.7 31.8 24.6 40.1 20.0 16.0 24.7 41.5 39.6 43.4 Far-western Terai 51.0 48.7 53.3 2.0 1.0 4.1 6.0 3.7 9.7 30.3 22.6 39.4 24.7 17.6 33.5 23.6 19.4 28.4 36.2 34.4 38.1 Nepal 51.2 50.7 51.7 7.2 6.3 8.1 4.7 4.1 5.3 39.2 37.3 41.2 27.8 26.1 29.6 17.8 16.9 18.8 39.6 39.1 40.0 Source. Author’s calculations based on Nepal DHS 2011 data. Note. BMI = body mass index; LL = lower limit; UL = upper limit; Nepal DHS = Nepal Demographic Health Survey. 10 SAGE Open In the case of ecodevelopmental subregion, western moun- followed by hills and Terai regions. Surprisingly, educational tain region has experienced the highest number of under-five poverty is the highest in the Terai region (0.237). Again, deaths (7.9%) and underweight children (41.0%). However, stark differences can be noticed in the case of wealth poverty. central Terai region had the highest number of adult females About three fourth people are poor in mountain region as with BMI < 18.5 (25.9%). Under-five mortality was the least compared with only one half and one fourth in hill and Terai in the western Terai (3.20%), whereas underweight children regions, respectively. and adult females with BMI < 18.5 were the lowest in the western hill. Thus, there is no single subregion, which has MPI across development regions. The results of MPI by the performed better in all health indicators. development region reveal that mid-western (0.262) and far- western (0.247) regions have the highest incidence of pov- Distribution of indicators by geographic regions: Standard of living. By erty, whereas, western region (0.188) has the least. place of residence, huge poverty differences are evident in rural– Educational poverty is the highest in the central region urban areas in the case of wealth status. The share of poorer and (0.269) followed by the eastern region (0.1560). Here again, poorest in terms of wealth status is more than 7 times in rural the educational poverty is found to be the lowest (0.119) in areas than urban areas. About 6.3% people are found to be in the far-western region, but in contrast to it, this region has the poorer and poorest wealth quintile in urban areas as against of highest health poverty (0.219) and second-highest standard 44.6% in rural areas. of living poverty (0.579). Similar trends for enrolment were By ecological regions, high variation was noticed in the case also observed under different survey reports (Asia-Pacific of wealth status. Around 71.8% people are poor in mountain Cultural Center for UNESCO [ACCU], 2001; Ministry of region followed by hills (52.5%) and Terai region (25.5%). Health and Population et al., 2012; NLSS, 2011). However, Wealth poverty is 3 times higher in mountain region than Terai. differences in health poverty among other regions than far- In the case of developmental regions, the incidence of western are not very significant. But pronounced divergence wealth poverty was just double in the mid- and far-western can be noticed again in the case of poverty in terms of stan- regions as compared with other developmental regions. dard of living. Poverty in terms of standard of living is nearly About 60% people were poor in terms of wealth in mid- and double in mid-western (0.612) and far-western (0.579) far-western regions as compared with around 35% in other regions as compared with other three regions. Furthermore, three regions. This indicates the unequal distribution of differences can also be seen in the overall MPI score by wealth across developmental regions. developmental regions. Things are not different in ecodevelopment subregions. The degree of variation in wealth poverty can be imagined from MPI across ecodevelopment subregions. Table 3 presents the the number of people belonging to poor and poorest wealth MPI and relative rank of subregions. Far-western hill is the quintile ranges from 16.1% (eastern Terai) to 84.3% (western poorest (0.305) among all followed by mid-western hill mountain). Far-western hills (79.4%) and mid-western hills (0.283). Trends in the individual dimensions suggest greater (72.9%) were other two subregions with huge wealth poverty. variations in the case of educational and standard of living All Terai-related subregions exhibit low wealth poverty. poverty. Poverty differences between least and most educa- tionally poor are 0.260, whereas in the standard of living poverty are 0.682. However, differences are much less in the MPI case of health poverty, which is only 0.131. Standing of sub- Table 3 provides estimates of MPI by place of residence, regions in terms of individual dimensions also differs signifi- ecological regions, development regions and subregions. cantly. For instance, central Terai (0.342) and estern mountain (0.082) in education, western mountain (0.254) and western MPI by rural–urban place of residence. Overall MPI estimate hill (0.122) in health and western mountain (0.843) and east- for rural areas (0.254) is about 2.5 times greater than in urban ern Terai (0.161) in the standard of living poverty are the areas (0.091). While 25% people in the rural areas are expe- most and least poor regions, respectively. riencing multidimensional poverty, this proportion is only 10% in the urban areas. Specifically, the results show that Income-Based Poverty Versus Multidimensional rural areas are at a disadvantageous position in comparison Poverty with urban areas in all three dimensions, namely, education, health and wealth status. The highest difference in rural– In this section, we compare the income-based poverty mea- urban poverty was found in wealth status. Poverty in terms of sure with MPI. Poverty is the lack of resources over time, wealth status in rural areas (0.446) is nearly 7 times higher whereas MPI is a consequence of deprivation in various than their urban counterparts (0.063). resources. There is a complementary and dynamic relation- ship between these two indicators. In developed countries MPI across ecological regions. Estimates by ecological region context, considerable variation were observed between reveal that MPI is the highest in the mountain region income-based poverty estimates and MPI, although the 11 Table 3. Multidimensional Poverty by Place of Residence, Nepal, 2011. Standard of living poverty Multidimensional poverty Education poverty index (EPI) Health poverty index (HPI) index (SLPI) index (MPI) 95% CI 95% CI 95% CI 95% CI Background characteristics Region Score LL UL Score LL UL Score LL UL Score LL UL Rank Place of residence Urban 0.10 0.08 0.12 0.12 0.09 0.16 0.06 0.06 0.07 0.09 0.07 0.11 2 Rural 0.20 0.19 0.21 0.18 0.17 0.19 0.45 0.44 0.45 0.25 0.24 0.27 1 Ecological region Mountain 0.09 0.06 0.13 0.21 0.17 0.26 0.72 0.70 0.73 0.24 0.19 0.29 1 Hill 0.13 0.12 0.15 0.15 0.14 0.17 0.53 0.52 0.53 0.22 0.20 0.24 2 Terai 0.24 0.22 0.25 0.18 0.16 0.20 0.26 0.25 0.26 0.22 0.21 0.23 2 Development Eastern 0.16 0.14 0.18 0.16 0.14 0.19 0.35 0.34 0.36 0.21 0.19 0.23 4 region Central 0.27 0.25 0.29 0.17 0.15 0.20 0.32 0.32 0.33 0.25 0.23 0.26 2 Western 0.13 0.11 0.15 0.15 0.12 0.18 0.36 0.35 0.37 0.19 0.17 0.21 5 Mid-western 0.15 0.13 0.18 0.19 0.16 0.23 0.61 0.60 0.62 0.26 0.23 0.30 1 Far-western 0.12 0.09 0.15 0.22 0.18 0.26 0.58 0.57 0.59 0.25 0.21 0.29 2 Ecodevelopment Eastern mountain 0.08 0.04 0.17 0.15 0.08 0.26 0.65 0.62 0.68 0.20 0.13 0.31 8 subregion Central mountain 0.09 0.04 0.20 0.18 0.10 0.29 0.60 0.57 0.63 0.22 0.13 0.33 6 Western mountain 0.09 0.05 0.17 0.25 0.19 0.33 0.84 0.82 0.86 0.27 0.20 0.36 3 Eastern hill 0.13 0.10 0.16 0.16 0.12 0.21 0.62 0.60 0.63 0.23 0.19 0.28 5 Central hill 0.14 0.11 0.17 0.12 0.09 0.16 0.33 0.32 0.34 0.18 0.15 0.21 10 Western hill 0.09 0.07 0.12 0.12 0.09 0.16 0.49 0.48 0.50 0.18 0.15 0.21 10 Mid-western hill 0.17 0.14 0.21 0.18 0.13 0.24 0.73 0.71 0.75 0.28 0.24 0.34 2 Far-western hill 0.15 0.11 0.20 0.24 0.18 0.32 0.79 0.78 0.81 0.30 0.25 0.37 1 Eastern Terai 0.18 0.15 0.20 0.16 0.13 0.19 0.16 0.15 0.17 0.16 0.14 0.19 13 Central Terai 0.34 0.32 0.37 0.20 0.17 0.23 0.29 0.28 0.30 0.27 0.25 0.29 3 Western Terai 0.16 0.13 0.19 0.17 0.13 0.22 0.18 0.16 0.19 0.17 0.14 0.20 12 Mid-western Terai 0.14 0.11 0.19 0.17 0.13 0.24 0.42 0.40 0.43 0.22 0.17 0.27 6 Far-western Terai 0.10 0.07 0.15 0.18 0.13 0.25 0.36 0.34 0.38 0.19 0.15 0.24 9 Total 0.19 0.18 0.20 0.17 0.16 0.19 0.40 0.39 0.40 0.24 0.22 0.25 Source. Author’s calculations based on Nepal DHS 2011 data. Note. LL = lower limit; UL = upper limit; Nepal DHS = Nepal Demographic Health Survey. 12 SAGE Open Table 4. Income-Based Poverty Versus Multidimensional Poverty in Nepal, 2011. Poverty (%) Background Multidimensional Multidimensional Income-based Human poverty index, 2010- a a b c characteristics Region poverty (AM) 2011 poverty (GM) 2011 poverty (2010-2011) 2011, by NHDR (2014) Place of residence Urban 12.83 9.11 15.5 18.5 Rural 32.96 25.38 27.4 34.0 Ecological region Mountain 43.24 23.91 42.3 38.5 Hill 32.98 22.08 24.3 29.2 Terai 26.70 22.13 23.4 33.0 Development Eastern 27.15 20.59 22.3 29.2 region Central 29.25 24.63 21.7 31.5 Western 26.85 18.82 21.4 27.2 Mid-western 39.33 26.20 31.7 36.6 Far-western 37.83 24.71 45.7 34.8 Subregion Eastern mountain 37.59 20.13 — 30.7 Central mountain 39.29 21.51 — 37.5 Western mountain 49.18 26.99 — 29.3 Eastern hill 36.96 23.35 — 30.2 Central hill 23.73 17.70 — 24.7 Western hill 30.10 17.58 — 25.6 Mid-western hill 43.09 28.26 — 38.2 Far-western hill 47.14 30.50 — 42.1 Eastern Terai 20.62 16.48 — 29.5 Central Terai 31.32 26.90 — 39.4 Western Terai 22.47 16.89 — 29.7 Mid-western Terai 31.89 21.81 — 32.5 Far-western Terai 27.95 18.75 — 28.4 Total 30.39 23.64 25.2 31.1 a b c Source. Author’s estimates based on Nepal DHS 2011 data. CBS (2010-2011), Nepal. NHDR, 2014. Note. AM = arithmetic mean; GM = geometric mean; NHDR = Nepal Human Development Report; Nepal DHS = Nepal Demographic Health Survey; CBS = Central Bureau of Statistics. relationship is strictly not linear. Both Peter Townsend (1979) poverty by CBS (2010-2011). Our poverty estimates for the and Mack and Lansley (1985) used the relationship between urban area (9.11%) show lesser poverty as compared with income and deprivation to choose their MPI items (cited in CBS 2010-2011 (15.5%) and NHDR (2014; 18.5%) but in Guio et al., 2016). But existing literature in developing coun- the case of rural poverty, the differences are not very signifi- tries such as Nepal hardly gives any evidence on the nature cant between our estimates (25.38%) and CBS 2010-2011 of the relationship between income poverty and multidimen- (27.4%). However, rural poverty shown by NHDR (2014) is sional poverty. This study fills this decisive gap. Table 4 about 10% higher than our MPI (GM) estimates. In the case compares estimates of multidimensional poverty estimates of classification based on ecological regions, mountain from our study (by the arithmetic mean method and GM region showed the highest poverty by all measures, but high method) with income-based poverty estimates by CBS variability in estimates across the methods is visible. Only (2011) and Nepal human poverty estimates by NHDR (2014). 23.91% people are experiencing multidimensional poverty Our estimates of overall MPI (GM) are similar to income- as per MPI (GM) whereas the corresponding figure by MPI based poverty. For instance, 23.64% people are poor as per (arithmetic mean) is 43.24%, by CBS 2010-2011 is 42.3% our estimates, and the corresponding figure by income-based and 38.5% by NHDR (2014). The same is the case with hill measure is 25.2%. This provides a kind of validation to our and Terai regions. A similar contrast in terms of variability of results. estimates by different methods is also visible in development Meanwhile, MPI estimates by the arithmetic mean are regions. similar to NHDR (2014) estimates. As per NHDR (2014), In the case of subregions, there is a change in relative posi- poverty is 31.1%, whereas as per our estimates, it is 30.39%. tion of subregions in terms of least and most multidimen- Rural–urban poverty differences by MPI (GM) are about sional poor. As per MPI (GM), far-western hill subregion 15%, which is around 12% in the case of income-based (30.50%) is the region with the highest multidimensional Goli et al. 13 44% in Alkire et al. (2013) in case of Nepal. However, we have no doubt in the fact that there is scope to improve the measure further given the availability of data, especially, on the new forms of assets and direct income information in the same data set. The estimates at subnational level suggest that geographi- cal location still works as a major determining factor in pov- erty as mountain region has the highest multidimensional poverty. The disparity between rural and urban poverty is significant. Despite substantial progress in the reduction of income poverty in recent years, multidimensional poverty in rural areas of Nepal remains slightly higher. It may have seri- ous implications because more than 80% of the population in Nepal still lives in rural areas. However, a careful observa- Figure 2. Relationship between MPI and income-based poverty. tion of different dimensions of multidimensional poverty Source. Figure generated based on the estimates showed in Table 4. shows that it is the rural deprivation in the standard of living, Note. MPI = multidimensional poverty index. which is a major contributor to rural–urban differences in the total MPI. These results are in tune with some of the previous poor and eastern Terai (16.48%) is the subregion with least studies (Bhurtel, 2013; Jerve, 2001; NHDR, 2009; Wagle, multidimensional poor. On the contrary, far-western hill 2007), which also noted the poor standard of living in rural (42.1%) and central hill subregions (24.7%) were the most areas, which contributes majorly to multidimensional rural and least multidimensional poor, respectively, as per NHDR poverty in the country. (2014). However, the education status seems to be not related to Overall, compared to developed countries, in Nepal, we the geographical location as our study indicates the highest found less variation in the levels of income-based poverty net enrollment in the mountain region followed by hills and and multidimensional poverty (Figure 2). This may be Terai regions, whereas the NHDR (2014) shows that hills because with an increase in absolute income levels with eco- have the lowest educational poverty. The educational pov- nomic growth, the relative disparities in material access erty seems to be indicator sensitive where NHDR (2014) increases. With the introduction of every new technology poverty index takes illiteracy rate into consideration; the and resources, richer tend to access faster than their counter- present study has taken years of schooling and net enroll- part, thus, leads to rising disparities, which also give rise to ment ratio, which are better indicators than adult literacy and increase in relative deprivation or poverty (Deaton, 2013). gross enrollment ratio. Similar contrasting results in educa- However, in low-income countries still, the overall material tional poverty are also observed by development regions and access relative to developed countries is less. Therefore, subdevelopment regions. These results assume greater there is less gap between incomes-based poverty and multi- importance in the context where Nepal government fixed dimensional deprivation. their targets of educational attainment under development plans in terms of enrollment ratios, not literacy; our results provide more robust basis for effective policy planning and Discussion and Conclusion implementation. The results are also robust because regions Unlike several recent studies on multidimensional poverty, with a high number of children not enrolled (6-10 years of which used global MPI procedure of estimation, we have age) coincide with the high number of people attending less proposed an alternative approach for measuring MPI. In this than 5 years of schooling. approach, poverty in terms of each indicator has been mea- It is also noticeable that educationally poor regions (Terai) sured at the individual level rather than at household level. are the regions, which border with the most backward states Moreover, the aggregation procedure used in this study is (Uttar Pradesh and Bihar) of the neighboring country “India” similar to the procedure used in HDI calculation. As pointed regarding education and economic progress. There may be out in the methods section, this procedure overcomes the some common explanations and linkages behind the low limitations of global MPI used by UNDP-HDR reports. Our development process in the regions, which require deeper results are methodologically robust in terms of both numbers empirical investigation (Samuels, Nino-Zarazua, Wagle, of indicators taken for each dimension and method of aggre- Sultana, & Sultana, 2011; Srinivasan, 2012). One possible gation. Our methodological improvements show that current explanation is that migration is very high in the Terai region methodology of estimating multidimensional poverty par- (49% highest among all regions), especially, the male migra- ticularly in the case of Nepal tends to deflate poverty num- tion (49.8%), which is also the case with neighboring Indian bers by a significant margin. For instance, multidimensional states (Nepal DHS, 2011). Nepalese leave their homes at an poor by GM aggregation method in this study is 24% against early age in expectation of earning a livelihood and, thus, do 14 SAGE Open not give due attention to educational attainment (Samuels features still are important determinant of multidimensional et al., 2011). poverty. This also shows that the government policies in Although interregion differences can be noticed in health Nepal mainly focus on the economic growth and employ- poverty, it is comparatively lesser than educational and stan- ment generation and ignore to bridge the gap between rich dard of living poverty. The heavy presence of international and poor. At the same time, the government also failed to agencies, that is, International Center for Integrated remove income inequalities as stated in Tenth Plan document Mountain Development (ICMOD), World Health (NHDR, 2014; Wagle, 2007). Organization (WHO), and United Nations Children’s Fund, In terms of comparability across the methods, it is impor- and so on and the introduction of many health-related pro- tant to note that CBS (2011) estimates of poverty are essen- gressive schemes by the Nepal government in the last two tially income-based poverty, which is comparable with our decades have brought the much-needed improvement in standard of living poverty estimates rather than overall MPI health status across all the regions, especially in the disad- estimates. Income is the source as well as the outcome of the vantageous regions such as mountain (Bentley, 1995; Engel wealth of a household. Past wealth helps in income genera- et al., 2013). Engel et al. (2013) noted, tion and thus generated income helps in wealth accumula- tion. Nonetheless, this feeding loop is based on the condition A consistent policy focus and sustained financial commitment that income level must be above their consumption expendi- by the government and donors throughout the past two decades, ture level. The income level of two households or individu- including substantial increases in funding for maternal health als may be same, but their wealth level may differ since the early 1990s, has allowed for widespread improvements significantly. Income level often does count for generational in access to medical services, particularly in remote areas. accumulation and transfer of wealth. Therefore, income may show the lesser incidence of poverty than wealth. This Thus, Nepal’s experience in the health sector can provide is exactly the case with Nepal. Poverty regarding wealth important lessons for other developing countries, especially, status showed the relatively higher incidence of poverty the South Asian Association of Regional Cooperation among different geographical areas (Figure 3). Even, coun- (SAARC) nations struggling with high levels of maternal try-level poverty in terms of wealth status differs signifi- mortality and poor health facilities, mainly within a circum- cantly than income poverty. Our measure of standard of stance of difficult terrain and high-income poverty. living poverty is comprehensive as it is based on wealth Regarding wealth status, poverty levels are higher across all index, which has been prepared by including multiple the regions. The high poverty in terms of wealth status indi- household assets. The estimates from this study show the cates poor conditions of housing, sanitation, electricity, drink- greater extent of wealth poverty compared with OPHI ing water, and other basic amenities. Wealth poverty contributes (2013). However, multidimensional poverty levels are much to significant differences in the overall multidimensional pov- less than wealth index–based poverty levels, which suggest erty of different place of residence and regions. Wealth poverty that Government of Nepal is operating a number of social in rural areas is 7 times higher than urban areas, whereas it is 3 security measures in the form of provision of basic educa- times higher in mountain than Terai region. Similarly, it is tion and health services to all the populations, which might almost double in the far-western and mid-western region as work as counteracting to wealth-based deprivation. compared with other developmental regions. Stark differences In conclusion, this study had examined the extent of multi- were also observed across the subregional classification. These dimensional poverty in Nepal disaggregated by geographic findings from the study are corroborated by other studies, regions. It adopted the more robust method of MPI compared which also noted the significant rise in income and wealth with global MPI of UNDP, particularly regarding indicators, inequalities during last three decades across the rural–urban their definitions, dimensional structure, and aggregation pro- regions (NHDR, 2014; Wagle, 2007; World Institute for cedure than that of the previous studies. It also took into con- Development Economics Research [WIDER], 2005; World sideration the latest methodological improvements in Bank, 2006). Bhurtel (2013) argued that it implies two things: calculating deprivation index measures by Guio et al. (2012), Guio et al. (2016), and Notten and Mendelson (2016). To con- First, the labour share of the national income has declined over clude, the findings of the study suggest that although Nepal time while the share of capital has rapidly increased. The dull has experienced a decent economic progress and a consider- growth of agriculture and stagnant manufacturing has mainly able reduction in education and health poverty with a consid- contributed to the growing economic inequalities. Secondly, the erable increase in wealth inequalities across the regions, government (Nepal) has failed to take fiscal measures to reduce overall MPI remains high. A far less has been achieved in the income inequality. Monetary measures such as providing cheap and easy credit to the poor have been largely ineffective. case of reducing the standard of living poverty, that is, wealth poverty and inequalities across the regions. Thus, the article High incidence of the standard of living poverty in the suggests that development policies and poverty reduction pro- mountain areas is also a reflection of poor provision of basic grams in Nepal must aim to reduce multidimensional poverty needs of life as well as a confirmation of the fact that natural by geographic regions, of which deprivation in education, Goli et al. 15 Initiative, Oxford Department of International Development, Queen Elizabeth House, University of Oxford. Alkire, S., Roche, J. M., & Vaz, A. (2014). Multidimensional poverty dynamics: Methodology and results for 34 countries (OPHI research in progress). Retrieved from https://ophi.org. uk/ophi-research-in-progress-41a-2/ Alkire, S., & Santos, M. E. (2010, July). Acute multidimensional poverty: A new index for developing countries (OPHI Working Papers 38). Oxford, UK: University of Oxford. Alkire, S., & Santos, M. E. (2013). A multidimensional approach: Poverty measurement & beyond. Social Indicators Research, 112, 239-257. Anand, S., & Sen, A. (1997). Concepts of human development and poverty! A multidimensional perspective. In Poverty and human development: Human development papers (pp. 1-20). New York, Figure 3. Relationship between wealth and income poverty. NY: United Nations Development Programme. Retrieved from Note. Figure is constructed based on the estimates of wealth poverty from https://scholar.harvard.edu/sen/publications/concepts-human- Table 2 and income poverty from Table 4. development-and-poverty-multidimensional-perspective Angulo, R., Díaz, Y., & Pardo, R. (2016). The Colombian multidi- mensional poverty index: Measuring poverty in a public policy health, and basic amenities must be an integral component, context. Social Indicators Research, 127, 1-38. along with their efforts to improve economic growth and Asia-Pacific Cultural Centre for UNESCO. (2001). Regional gen- reduce income and wealth-based poverty. der and ethnic disparity in education. Tokyo, Japan: Author. Retrieved from http://www.accu.or.jp/litdbase/literacy/nrc_ Declaration of Conflicting Interests nfe/eng_bul/BUL18.pdf Atkinson, A. B. (1987). On the measurement of poverty. The author(s) declared no potential conflicts of interest with respect Econometrica: Journal of the Econometric Society, 55, 749-764. to the research, authorship, and/or publication of this article. Atkinson, A. B. (1970). On the measurement of inequality. Journal of Economic Theory, 2, 244-263. Funding Bader, C., Bieri, S., Wiesmann, U., & Heinimann, A. (2016). A dif- The author(s) received no financial support for the research, author- ferent perspective on poverty in Lao PDR: Multidimensional ship, and/or publication of this article. poverty in Lao PDR for the years 2002/2003 and 2007/2008. Social Indicators Research, 126, 483-502. ORCID iD Bardhan, P. K. (1970). On the minimum level of living and the rural poor. Indian Economic Review, 5, 129-136. Moradhvaj https://orcid.org/0000-0002-8798-385X Bardhan, P. K. (1971). On the minimum level of living and the rural poor: A further note. Indian Economic Review, 6, 78-87. References Bentley, H. (1995). The organisation of health care in Nepal. Abel-Smith, B., & Townsend, P. (1965). The poor and the poorest. International Journal of Nursing Studies, 32, 260-270. London, England: G. Bell. Betti, G., Gagliardi, F., & Verma, V. (2018). Simplified jackknife Alkire, S., Adriana, C., & Roche, J. M. (2013, February). variance estimates for fuzzy measures of multidimensional Multidimensional poverty index 2013: Brief methodologi- poverty. International Statistical Review, 86, 68-86. cal note and results. Oxford, UK: Oxford Poverty & Human Bhurtel, B. P. (2013, October 17). Rich man’s world as the wealth Development Initiative, Oxford Department of International gap grows in Nepal. The Nation, The Kathmandu Post, Asia Development, Queen Elizabeth House, University of Oxford. News Network. Retrieved from http://nationmultimedia.com/ Alkire, S., & Foster, J. (2007). Counting and multidimensional pov- opinion/RICH-mans-world-as-the-wealth-gap-grows-in- erty measures (OPHI Working Paper Series, No. 7). Oxford, Nepal-30217256.html UK: Oxford Poverty & Human Development Initiative, Bosanquet, H. (1903). The “poverty line.” Charity Organisation University of Oxford. Revised in January 2008. Review, 8, 321-325. Alkire, S., & Foster, J. (2011a). Counting and multidimensional Bourguignon, F., & Chakravarty, S. R. (2003). The measurement poverty measurement. Journal of Public Economics, 95, of multidimensional poverty. Journal of Economic Inequality, 476-487. 1, 25-49. Alkire, S., & Foster, J. (2011b). Understandings and misunder- Central Bureau of Statistics. (2011). Poverty in Nepal. Kathmandu, standings of multidimensional poverty measurement. Journal Nepal: National Planning Commission Secretariat, Government of Economic Inequality, 9, 289-314. of Nepal. Available from www.cbs.gov.np Alkire, S., Foster, J., & Santos, M. E. (2011). Where did identifica- Chakraborty, N. M., Fry, K., Behl, R., & Longfield, K. (2016). tion go? Journal of Economic Inequality, 9, 501-505. Simplified asset indices to measure wealth and equity in health Alkire, S., Roche, J. M., Santos, M. E., & Seth, S. (2011). programs: A reliability and validity analysis using survey data Multidimensional poverty index 2011: Brief methodological from 16 countries. Global Health: Science and Practice, 4, note. Oxford, UK: Oxford Poverty & Human Development 141-154. 16 SAGE Open Clark, S., Hemming, R., & Ulph, D. (1981). On indices for the mea- national poverty strategies (pp. 89-120). New York, NY: surement of poverty. The Economic Journal, 91, 515-526. United Nations Development Programme. Dasgupta, P., Sen, A., & Starrett, D. (1973). Notes on the measure- Kakwani, N. (1980). On a class of poverty measures. Econometrica: ment of inequality. Journal of Economic Theory, 6, 180-187. Journal of the Econometric Society, 48, 437-446. Deaton, A. (2013). The great escape: Health, wealth, and the ori- Mack, J., & Lansley, S. (1985). Poor Britain. London, England: gins of inequality. Princeton, NJ: Princeton University Press. George Allen & Unwin. Dhongde, S., & Haveman, R. (2017). Multi-dimensional depriva- Ministry of Health and Population, New ERA, & ICF International. tion in the US. Social Indicators Research, 133, 477-500. (2012). Nepal Demographic and Health Survey 2011. Dickes, P. (1989). Pauvreté et conditions d’existence: théories, Kathmandu, Nepal: Ministry of Health and Population. modèles et measures [Poverty and living conditions: Theories, Mitra, S. (2016). Synergies among monetary, multidimensional and models and measures]. Walferdange: Centre d’études de popu- subjective poverty: Evidence from Nepal. Social Indicators lations de pauvreté et de politiques socio-économiques. Research, 125, 103-125. Drèze, J., & Sen, A. (2013). An uncertain glory: India and its con- Mohanty, S. K., Agrawal, N. K., Mahapatra, B., Choudhury, D., tradictions. Princeton, NJ: Princeton University Press. Tuladhar, S., & Holmgren, E. V. (2017). Multidimensional Dutta, S. (2015). Identifying single or multiple poverty trap: An poverty and catastrophic health spending in the mountainous application to Indian household panel data. Social Indicators regions of Myanmar, Nepal and India. International Journal Research, 120, 157-179. for Equity in Health, 16(1), 21. Engel, J., Glennie, J., Adhikari, S. R., Bhattarai, S. W., Prasai, D. P., & Montoya, Á. J. A., & Teixeira, K. M. D. (2017). Multidimensional Samuels, F. (2013). Nepal’s story understanding improvements poverty in Nicaragua: Are female-headed households better in maternal health. London, England: Overseas Development off? Social Indicators Research, 132, 1037-1063. Institute. Retrieved from https://www.odi.org/sites/odi.org.uk/ Nepal Human Development Report. (2009). Nepal Human files/odi-assets/publications-opinion-files/8624.pdf Development Report 2009: State transformation and human Eurostat. (2002). Income, poverty and social exclusion: Second development. United Nations Development Programme. report. Luxembourg: Office for Official Publications of the Retrieved from http://www.refworld.org/docid/4a8a72542. European Communities. html Foster, J., Greer, J., & Thorbecke, E. (1984). A class of decom- Nepal Human Development Report. (2014). National Human posable poverty measures. Econometrica: Journal of the Development Report 2014: Beyond geography unlock- Econometric Society, 52, 761-766. ing human potential. Kathmandu, Nepal: National Planning Gordon, D. (2000). The scientific measurement of poverty: Recent Commission. Retrieved from http://hdr.undp.org/sites/default/ theoretical advances. In J. Bradshaw & R. Sainsbury (Eds.), files/nepal_nhdr_2014-final.pdf Researching Poverty (pp. 37-58). Aldershot, UK: Ashgate. Nepal Living Standard Survey. (2011). Nepal-Living Standards Gordon, D., & Townsend, P. (1990). Measuring the poverty line. Survey 2010-2011, third round. Kathmandu, Nepal: Central Radical Statistics, 47, 5-12. Bureau of Statistics, National Planning Commission Secretariat, Government of Nepal. (2011, November). Nepal Living Standards Government of Nepal. Survey 2010/2011 (Statistical Report Vol. I). Kathmandu, Notten, G., & Mendelson, M. (2016). Using low income and mate- Nepal: Central Bureau of Statistics, National Planning rial deprivation to monitor poverty reduction. Ottawa, Ontario, Commission Secretariat, Government of Nepal. Canada: Caledon Institute of Social Policy. Guio, A.-C. (2009, February 10). What can be learned from deprivation Nowak, D., & Scheicher, C. (2017). Considering the extremely indicators in Europe. Paper presented at the Indicator Subgroup poor: Multidimensional poverty measurement for Germany. of the Social Protection Committee. Retrieved from https:// Social Indicators Research, 133, 139-162. www.bristol.ac.uk/poverty/downloads/keyofficialdocuments/ Oxford Poverty & Human Development Initiative. (2013). Nepal Deprivation%20indicators%20in%20Europe.pdf country briefing—Multidimensional poverty index data bank. Guio, A.-C., Gordon, D., & Marlier, E. (2012). Measuring material Oxford, UK: Oxford Poverty & Human Development Initiative, deprivation in the EU: Indicators for the whole population and University of Oxford. Retrieved from www.ophi.org.uk/multi- child-specific indicators (Eurostat methodologies and working dimensional-poverty-index/mpi-country-briefings/ papers). Luxembourg: Office for Official Publications of the Perry, B. (2002). The mismatch between income measures and European Communities. direct outcome measures of poverty. Social Policy Journal of Guio, A.-C., Marlier, E., Gordon, D., Fahmy, E., Nandy, S., & New Zealand, 19, 101-127. Pomati, M. (2016). Improving the measurement of material Rangarajan, C., & Dev, M. S. (2015). Counting the poor: deprivation at the European Union level. Journal of European Measurement and other issues. Economic and Political Weekly, Social Policy, 26, 219-333. 50(2), 70-74. Gwatkin, D. R., Rutstein, S., Johnson, K., Suliman, E., Wagstaff, A., Ravallion, M. (1998). Poverty lines in theory and practice (Vol. & Amouzou, A. (2007). Socio-economic differences in health, 133). Washington, DC: World Bank Publications. nutrition, and population. Washington, DC: World Bank. Ravallion, M., & Huppi, M. (1991). Measuring changes in poverty: Hagenaars, A. (1987). A class of poverty indices. International A methodological case study of Indonesia during an adjust- Economic Review, 28, 583-607. ment period. The World Bank Economic Review, 5, 57-82. Hanandita, W., & Tampubolon, G. (2016). Multidimensional pov- Reyles, D. Z. (2010). The human development index (HDI). Oxford, erty in Indonesia: Trend over the last decade (2003–2013). UK: Oxford Poverty & Human Development Initiative. Social Indicators Research, 128, 559-587. Rogan, M. (2016). Gender and multidimensional poverty in South Jerve, A. M. (2001). Rural-urban linkages and poverty analysis. Africa: Applying the global multidimensional poverty index In A. Grinspun (Eds.), Choices for the poor: Lessons from (MPI). Social Indicators Research, 126, 987-1006. Goli et al. 17 Rutstein, S. O. (1999). Wealth versus expenditure: Comparison Townsend, P. (Ed.). (1971). The concept of poverty. London, between the DHS wealth index and household expenditures England: Heinemann. in four departments of Guatemala. Calverton, MD: ORC Townsend, P. (1979). Poverty in the United Kingdom. Macro. Harmondsworth, UK: Penguin Books. Rutstein, S. O. (2008, October). The DHS wealth index: Approaches Townsend, P. (2010). The meaning of poverty. The British Journal for rural and urban areas (The DHS Working Papers, No. 60). of Sociology, 61(Suppl. 1), 85-102. Demographic and Health Research, United States Agency for Tsui, K. Y. (2002). Multidimensional poverty indices. Social International Development. Retrieved from https://dhsprogram.com/ Choice and Welfare, 19, 69-93. pubs/pdf/WP60/WP60.pdf Uematsu, H., Shidiq, A. R., & Tiwari, S. (2016, September). Trends Rutstein, S. O., & Johnson, K. (2004). The DHS wealth index (DHS and drivers of poverty reduction in Nepal: A historical perspec- Comparative Reports No. 6). Calverton, MD: ORC Macro. tive (Policy Research Working Paper No. 7830). Washington, Retrieved from https://dhsprogram.com/pubs/pdf/cr6/cr6.pdf DC. World Bank. Retrieved from https://openknowledge. Rutstein, S. O., Johnson, K., & Gwatkin, D. (2000, March). Poverty, worldbank.org/handle/10986/25146 health inequality, and its health and demographic effect. United Nations Development Programme. (1993). Human develop- Presented at the annual meeting of the Population Association ment report 1993. New York, NY: United Nations Development of America, Los Angeles, CA. Programme, Oxford University Press. Samuels, F., Nino-Zarazua, M., Wagle, S., Sultana, T., & Sultana, M. United Nations Development Programme. (2010). Human develop- M. (2011, November). Vulnerabilities of movement: Cross-border ment report 2010–20th anniversary edition. The real wealth mobility between India, Nepal and Bangladesh (Background of nations: Pathways to human development. New York, NY: note). London, England: Overseas Development Institute. Author. Sen, A. (1970). Interpersonal aggregation and partial comparabil- United Nations Development Programme. (2015). Human devel- ity. Econometrica: Journal of the Econometric Society, 38, opment report 2015: Work for human development. New 393-409. York, NY: United Nations Development Programme, Oxford Sen, A. (1972). Interpersonal comparison and partial comparabil- University Press. ity: A correction. Econometrica: Journal of the Econometric United Nations Nepal. (2014). 12-point understanding reached Society, 40(5), 959. between the Seven Political Parties and Nepal Communist Sen, A. (1973). Poverty, inequality and unemployment: Some Party (Maoists). Available from http://www.un.org.np/ conceptual issues in measurement. Economic and Political Wagle, U. R. (2007, October 11). Economic inequality in the Weekly, 8, 1457-1464. “democratic” Nepal: Dimensions and implications. Presented Sen, A. (1976). Poverty: An ordinal approach to measurement. at the Himalayan Policy Research Conference, Madison, WI. Econometrica: Journal of the Econometric Society, 44, Retrieved from https://ejournals.unm.edu/index.php/nsc/ 219-231. article/view/642/739 Sen, A. (1979). Issues in the measurement of poverty. The Wang, Y., & Wang, B. (2016). Multidimensional poverty measure Scandinavian Journal of Economics, 81, 285-307. and analysis: A case study from Hechi City, China. Springer Sen, A. (1981). Ingredients of famine analysis: Availability and Plus, 5, Article 642. entitlements. The Quarterly Journal of Economics, 96, 433-464. Whelan, B. (1993). Non-monetary indicators of poverty. In Sen, A. (1987). Gender and cooperative conflicts (No. 1342). J. Berghman & B. Cantillon (Eds.), The European face of social Helsinki, Finland: World Institute for Development Economics security: Essays in honour of Herman Deleeck (pp. 24-42). Research. Aldershot, UK: Avebury. Sen, A. (1989). Food and freedom. World Development, 17, 769-781. World Bank. (2006). World development indicators 2006. Sen, A. (1992). Poverty reexamined. Cambridge, MA: Harvard Washington, DC: Author. University Press. World Bank. (2014). Poverty and inequality measures in practice: Sen, A. (2000a). A decade of human development. Journal of A basic reference guide with Stata examples. Washington, DC: Human Development, 1, 17-23. Author. Sen, A. (2000b). Social exclusion: Concept, application, and scru- World Bank. (2016). Moving up the ladder: Poverty reduction and tiny (Social Development Papers). Office of Environment and social mobility in Nepal. Retrieved from https:// openknowledge. Social Development, Asian Development Bank. Retrieved worldbank.org/handle/10986/25173 from http://hdl.handle.net/11540/2339 World Institute for Development Economics Research. (2005). Smith, P. (1992). Measuring human development. Southampton, World income inequality database (Version 2.0a). Helsinki, UK: University of Southampton. Finland: World Institute for Development Economics Research, Srinivasan, P. V. (2012, November). Regional cooperation and United Nations University. integration through cross-border infrastructure development in Yu, J. (2013). Multidimensional poverty in China: Findings based South Asia: Impact on poverty (South Asia Working Paper Series, on the CHNS. Social Indicators Research, 112, 315-336. 14). Asian Development Bank. Retrieved from https://www.adb. org/publications/regional-cooperation-and- integration-through- Author Biographies cross-border-infrastructure-development Thon, D. (1979). On measuring poverty. Review of Income and Srinivas Goli is an assistant professor of population studies at JNU, Wealth, 25, 429-439. New Delhi. He did his master’s and PhD in population studies from Townsend, P. (1954). Measuring poverty. The British Journal of IIPS, Mumbai, India. He is teaching fertility studies, family plan- Sociology, 5, 130-137. ning, family demography, gender and quantitative methods in social 18 SAGE Open science research at JNU. His research areas are formal demography, New Delhi and IIPS, Mumbai respectively. His research interests family demography, ageing, inequalities in health, nutrition and gen- are mortality, health care, health financing and gender & develop- der status and its social determinants. Currently, he is involved in ment. He has published four journal articles in highly reputed inter- number of collaborative projects in India and abroad. national journals. Nagendra Kumar Maurya is an assistant professor at the Prem Bhandari is social researcher at the Population Research Department of Applied Economics, University of Lucknow, Center, University of Michigan, Ann Arbor, Michigan. He did his Lucknow, India. He did his master’s and PhD in Applied Economics PhD in rural sociology and demography from the Pennsylvania from University of Lucknow, India. His research interest includes State University. His research interest includes social research state finances, public economics, human development, macroeco- methods, socio-economic and cultural determinants of migration, nomics, poverty and inequality. fertility, and population health, rural social change; population and Moradhvaj is a doctoral student of population studies at JNU, New environment, sociology of agriculture, and rural development Delhi. He did master and MPhil in population studies from JNU, planning. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png SAGE Open SAGE http://www.deepdyve.com/lp/sage/regional-differentials-in-multidimensional-poverty-in-nepal-rethinking-kM5875No0K

Loading next page...

References (60)

H. Bentley (1995)
The organisation of health care in Nepal.
International journal of nursing studies, 32 3
On the minimum level of living and the rural poor
Álvaro Montoya, Karla Teixeira (2017)
Multidimensional Poverty in Nicaragua: Are Female-Headed Households Better Off?
Social Indicators Research, 132
Aldi Hagenaars (1987)
A Class of Poverty Indices
International Economic Review, 28
Amartya Sen (1981)
Ingredients of famine analysis: availability and entitlements.
The quarterly journal of economics, 96 3
The mismatch between income measures and direct outcome measures of poverty
N. Chakraborty, K. Fry, Rasika Behl, K. Longfield (2016)
Simplified Asset Indices to Measure Wealth and Equity in Health Programs: A Reliability and Validity Analysis Using Survey Data From 16 Countries
Global Health: Science and Practice, 4
The meaning of poverty
A. Sen (2000)
A Decade of Human Development
Journal of Human Development, 1
J. Foster, J. Greer, E. Thorbecke (1984)
A Class of Decomposable Poverty Measures
Econometrica, 52
M. Rogan (2016)
Gender and Multidimensional Poverty in South Africa: Applying the Global Multidimensional Poverty Index (MPI)
Social Indicators Research, 126
Poverty, inequality and unemployment: Some conceptual issues in measurement
M. Ravallion (1998)
Poverty lines in theory and practice
S. Alkire, J. Foster, María Santos (2011)
Where did identification go?
The Journal of Economic Inequality, 9
The “poverty line.”
A. Sen (1970)
Interpersonal Aggregation and Partial Comparability
Econometrica, 38
S. Alkire, J. Foster (2010)
Counting and Multidimensional Poverty Measurement
Shatakshee Dhongde, R. Haveman (2017)
Multi-Dimensional Deprivation in the U.S.
Social Indicators Research, 133
S. Clark, R. Hemming, D. Ulph (1981)
On Indices for the Measurement of Poverty
The Economic Journal, 91
P. Townsend (1962)
The meaning of poverty. 1962.
The British journal of sociology, 61 Suppl 1
Rich man’s world as the wealth gap grows in Nepal
The scientific measurement of poverty: Recent theoretical advances
Wulung Hanandita, G. Tampubolon (2015)
Multidimensional Poverty in Indonesia: Trend Over the Last Decade (2003–2013)
Social Indicators Research, 128
A. Sen (1972)
Interpersonal Comparison and Partial Comparability: A Correction
Econometrica, 40
On the minimum level of living and the rural poor: A further note
F. Ness (1971)
An Uncertain Glory
Change: The Magazine of Higher Learning, 3
C. Singer (2001)
Human Development Report 2000
Journal of Government Information, 28
R. Angulo, Yadira Díaz, Renata Pardo (2016)
The Colombian Multidimensional Poverty Index: Measuring Poverty in a Public Policy Context
Social Indicators Research, 127
Non-monetary indicators of poverty
Jiantuo Yu (2013)
Multidimensional Poverty in China: Findings Based on the CHNS
Social Indicators Research, 112
Food and freedom
Measuring the poverty line
Yanhui Wang, B. Wang (2016)
Multidimensional poverty measure and analysis: a case study from Hechi City, China
SpringerPlus, 5
P. Townsend (2020)
Measuring Poverty
Welfare Law
E. Swanson (2006)
World development indicators 2007
M. Ravallion, M. Huppi (1991)
Measuring changes in poverty : a methodological case study of Indonesia during an adjustment period
The World Bank Economic Review, 5
A. Sen (1976)
Poverty: An Ordinal Approach to Measurement
Econometrica, 44
Anne-Catherine Guio, Eric Marlier, D. Gordon, Eldin Fahmy, S. Nandy, Marco Pomati (2016)
Improving the measurement of material deprivation at the European Union level
Journal of European Social Policy, 26
J. Breman (2015)
Down and Out in Ahmedabad
Economic and Political Weekly, 50
Rural-urban linkages and poverty analysis
Christoph Bader, Sabin Bieri, U. Wiesmann, A. Heinimann (2016)
A Different Perspective on Poverty in Lao PDR: Multidimensional Poverty in Lao PDR for the Years 2002/2003 and 2007/2008
Social Indicators Research, 126
S. Dutta (2015)
Identifying Single or Multiple Poverty Trap: An Application to Indian Household Panel Data
Social Indicators Research, 120
A. Atkinson (1987)
ON THE MEASUREMENT OF POVERTY
Econometrica, 55
F. Bourguignon, S. Chakravarty (2003)
The Measurement of Multidimensional Poverty
The Journal of Economic Inequality, 1
S. Alkire, J. Foster (2011)
Understandings and misunderstandings of multidimensional poverty measurement
The Journal of Economic Inequality, 9
A. Sen (1979)
Issues in the Measurement of Poverty
The Scandinavian Journal of Economics, 81
Ho-Su Sin, S. Namekawa (2013)
The great escape
Epigenetics, 8
G. Betti, F. Gagliardi, V. Verma (2018)
Simplified Jackknife Variance Estimates for Fuzzy Measures of Multidimensional Poverty
International Statistical Review, 86
D. Thon (1979)
On Measuring Poverty
Review of Income and Wealth, 25
P. Dasgupta, A. Sen, D. Starrett (1973)
Notes on the measurement of inequality
Journal of Economic Theory, 6
Victor Lippit (2000)
Human Development Report, 1997
Economic Development and Cultural Change, 48
Daniel Nowak, C. Scheicher (2017)
Considering the Extremely Poor: Multidimensional Poverty Measurement for Germany
Social Indicators Research, 133
Counting the poor: Measurement and other issues
A. Atkinson (1970)
On the measurement of inequality
Journal of Economic Theory, 2
Concepts of human development and poverty! A multidimensional perspective
N. Kakwani (1980)
On a Class of Poverty Measures
Econometrica, 48
S. Mitra (2016)
Synergies Among Monetary, Multidimensional and Subjective Poverty: Evidence from Nepal
Social Indicators Research, 125
Sanjay Mohanty, Nand Agrawal, B. Mahapatra, D. Choudhury, S. Tuladhar, E Holmgren (2017)
Multidimensional poverty and catastrophic health spending in the mountainous regions of Myanmar, Nepal and India
International Journal for Equity in Health, 16
K. Tsui (2002)
Multidimensional poverty indices
Social Choice and Welfare, 19
S. Alkire, María Santos (2013)
A Multidimensional Approach: Poverty Measurement & Beyond
Social Indicators Research, 112

Publisher: SAGE
Copyright: Copyright © 2022 by SAGE Publications Inc, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
ISSN: 2158-2440
eISSN: 2158-2440
DOI: 10.1177/2158244019837458
Publisher site: See Article on Publisher Site

Abstract

This article examines the extent of regional inequality in multidimensional poverty in Nepal using the nationally representative Nepal Demographic Health Survey (2011) data. The authors present a more robust method of multidimensional poverty index (MPI), particularly in terms of the procedure of estimation and aggregation of the indicators as compared with previous studies. The findings suggest that despite the relatively better economic progress and a considerable reduction in education and health poverty, there is a wide inequality across the regions. Far less has been achieved in the case of reducing the standard of living poverty, that is, wealth poverty and inequalities across the regions. The article finds that global MPI tends to inflate poverty estimates in the case of Nepal. It also suggests that development policies and poverty reduction programs in Nepal must aim to reduce multidimensional poverty, of which deprivation in education, health and basic amenities must be an integral component, along with their efforts to improve economic growth and reduce income poverty. Keywords multidimensional poverty, Nepal, MPI, DHS, regional differentials & Human Development Initiative [OPHI], 2013; Rangarajan Introduction & Dev, 2015; Reyles, 2010; Rogan, 2016; Wang & Wang, During the last one decade, Nepal has gone through a major 2016; World Bank, 2014; Yu, 2013). Nepal uses the concept political transition. Abolition of monarchy, the establishment of absolute poverty and has followed its own definition: of a Federal Democratic Republic, the election of Constituent According to which a person earning less than 1 US$ a day is Assembly in 2008 (and reelection in 2014) and the adoption termed as poor. By this definition, the latest official figures of the new constitution in September 2015 are the landmarks suggest that more than 35% of the total population is living in the political history and economic planning of Nepal. The below the poverty line in Nepal. However, the application of country is making every effort to move out of an extended the concept of relative poverty is virtually absent in Nepal political transition, and is also aiming to become a developed (Alkire, Adriana, & Roche, 2013; Central Bureau of Statistics country in the world by 2022 (United Nations Nepal, 2014). [CBS], 2011; Nepal Human Development Report [NHDR], The ambitious journey of transition from a comparatively 2014; Uematsu, Shidiq, & Tiwari, 2016; World Bank, 2016). less developed country to a developing and then to a devel- Earlier, to achieve million development goals (MDGs), oped nation demands a concerted effort for a holistic and now, to achieve sustainable development goals (SDGs), approach to the development. This is not possible without a Nepal has been fairly active in investing in social policies significant reduction in the incidence of both absolute and including poverty reduction and active society engagement relative multidimensional poverty in the country. (Government of Nepal, 2011; Uematsu et al., 2016). This has Measurement of poverty itself is a complex and highly debat- able issue. Social scientists in different countries have Jawaharlal Nehru University, New Delhi, India adopted different dimensions to determine poverty status of 2 University of Lucknow, India the populations (see, Angulo, Díaz, & Pardo, 2016; Bader, University of Michigan, Ann Arbor, USA Bieri, Wiesmann, & Heinimann, 2016; Dhongde & Haveman, Corresponding Author: 2017; Dutta, 2015; Guio, Gordon, & Marlier, 2012; Guio Srinivas Goli, Assistant Professor, Center for the Study of Regional et al., 2016; Hanandita & Tampubolon, 2016; Montoya & Development, Jawaharlal Nehru University, New Delhi 110067, India. Teixeira, 2017; Nowak & Scheicher, 2017; Oxford Poverty Email: sirispeaks2u@gmail.com Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open resulted in a significant reduction in poverty (Uematsu et al., Guio et al., 2012; Guio et al., 2016; Hanandita & Tampubolon, 2016). For instance, according to an estimate, the percentage 2016; Mohanty et al., 2017; Montoya & Teixeira, 2017; of the multidimensional poor in Nepal has dropped signifi- Nowak & Scheicher, 2017; Rangarajan & Dev, 2015; Rogan, cantly from 64.7% to 44.2% between 2006 and 2011, that is, 2016; Tsui, 2002; Wang & Wang, 2016; World Bank, 2014; by 4.1 percentage points per year (Alkire et al., 2013). In Yu, 2013). A majority of these studies used methods designed fact, the country has been able to reduce the national poverty for global MPI calculations (Alkire & Foster, 2007; Alkire, much faster than its neighboring countries such as India, Foster, & Santos, 2011; Alkire & Santos, 2013). Pakistan, and Bangladesh (Drèze & Sen, 2013). In spite of Advancing the existing methodology of selection of considerable progress in poverty reduction in recent years, parameters in multidimensional poverty, Guio et al. (2012) Nepal remains one of the poorest countries in the world. and Guio et al. (2016) proposed an analytical framework for With a human development index (HDI) of 0.548 in 2014, developing robust material deprivation indicators for the Nepal is ranked 145th out of 187 countries listed in the whole population in the context of the European Union. They United Nations Development Programme (UNDP; 2015). carried out systematic item-by-item analyses at national lev- The National Living Standards Survey (NLSS) conducted in els to identify material deprivation criteria, which satisfacto- 2010-2011 reported that more than 30% of Nepalese live on rily meet suitability, validity, reliability, and additive criteria less than US$14 per person per month using the income- across the European Union. There have been some efforts to based poverty estimation. However, this figure rises to 44.2% include multiple indicators in measuring poverty in Nepal in the case of multidimensional poor. Furthermore, there is a (Alkire et al., 2013; CBS, 2013; Mitra, 2016; NHDR, 2014). large inequality in the prevalence of poverty within the However, we understand that these Nepal-specific estimates nation. Although the overall poverty rate for Nepal is 30%, of multidimensional poverty have methodological limita- this figure rises to 45% in the mid-western region and to 46% tions both in terms of estimation procedures and the use of in the far-western region (NLSS 2010-2011). Thus, a indicators and their units of measurement. This study is an national-level figure often obscures the within-country attempt to refine the measure of multidimensional poverty inequality in poverty (Uematsu et al., 2016). both regarding its dimensionality and estimation procedure and to fill this gap in the literature, specifically in the context of Nepal. We provide a theoretical background and the evi- Background and Rationale dence of multidimensional poverty measures used in previ- Measuring poverty is a complicated process (Sen, 1979; ous studies below. Then, we describe data, our proposed Townsend, 1954, 1971, 1979). Early efforts of measuring method and its advantage over the existing method, indica- poverty involved unidimensional indicators based on income tors of alternative dimensions, and the difference in the pro- or consumption expenditure (Abel-Smith & Townsend, cedure used to estimate the multidimensional poverty in this 1965; Atkinson, 1987; Atkinson, 1970; Bosanquet, 1903; study as compared with other studies in the context of Nepal Clark, Hemming, & Ulph, 1981; Hagenaars, 1987; Kakwani, (Alkire et al., 2013; CBS, 2013; Mitra, 2016; NHDR, 2014; 1980; Ravallion, 1998; Ravallion & Huppi, 1991; Sen, 1976, Uematsu et al., 2016). 1981, 1987, 1989; Thon, 1979; Townsend, 1954, 1971, 1979). Later, it was recognized that no single indicator alone Measuring Poverty: Past Efforts and could capture the multiple aspects of poverty (Foster, Greer, Our Approach & Thorbecke, 1984; Townsend, 1979). Poverty is much more than having a low income or low consumption expenditure Debates on measuring poverty were intensified in the 1970s, (Anand & Sen, 1997; Bourguignon & Chakravarty, 2003; but these discussions were mainly about measuring income Sen, 1970; Townsend, 1954). poverty and defining the poverty line. During this period, the Realizing the significance of multiple indicators, there identification of the poor was exclusively by family-size- have been some efforts to include multiple indicators in mea- adjusted household income, concerning a specified income suring poverty. The first multidimensional measure can be poverty line. Some contributions are worth mentioning here, traced back to Townsend (1979), and the underpinnings of for example, Townsend (1954, 1971, 1979), Sen (1970, the multidimensional poverty index (MPI) were set out by 1972, 1973, 1992, 2000b), Bardhan (1970, 1971), Dasgupta, Foster et al. (1984). The global MPI was designed in 2010 by Sen, and Starrett (1973), Gordon and Townsend (1990). Sen the OPHI and the UNDP using different indicators to deter- (1976), in his seminal article “Poverty: An Ordinal Approach mine poverty beyond income-based measures (OPHI, 2013). to Measurement,” has emphasized the theoretical soundness This MPI replaced the previous human poverty index in sub- of the income poverty measurement. He has suggested an sequent human development reports of the world countries. ordinal approach based on ordinal axioms for measuring Following this, numerous studies in many countries have poverty. However, he admitted that such an approach is dif- used various procedures to estimate multidimensional pov- ficult to replicate in reality, as required data may not be avail- erty of individuals and households (Angulo et al., 2016; able. From the mid-1970s, it was recognized that poverty is Bader et al., 2016; Dhongde & Haveman, 2017; Dutta, 2015; much more than just having a little income (Townsend, Goli et al. 3 2010). During this time, the “basic needs approach,” social societies and their state of affairs (Alkire, Foster, & Santos, exclusion, and “capability approaches” gained prominence 2011). Thus, a key criticism of global MPI used by UNDP is in complementing the process of identification of the poor or a lack of examination of its suitability within countries, deprived populations. Studies have shown that income does dimensional structure reliability and validity and measure- not represent the nonmonetary multidimensional depriva- ment invariance apart from the ambiguity in their household- tions of households (such as lack of access to nutritious food, level aggregation procedure of individual and household health services, quality education, potable water, livable indicators. Some of these are discussed in detail in the house, sanitation facilities, electricity, basic information and method section of this study (also see Guio et al., 2012; Guio more), and thus fails to identify the poor correctly (Sen, et al., 2016). A study by Betti et al. (2018) on simplified 1970, 1972, 1973, 1976; Townsend, 1979). Consequently, Jacknife variance estimates for measures of MPI is a seminal researchers have introduced various nonmonetary measures contribution to the literature. of deprivation, supplementing these multidimensional analy- In Nepal, to the best of our knowledge, the present exer- ses with monetary measures to create a better overall picture cise is the third attempt to measure MPI at a disaggregated of poverty (Foster et al., 1984; Townsend, 1979). For level. The first attempt was by Alkire and her team at OPHI instance, Townsend (1979) highlighted “relative depriva- who estimated MPI for 104 countries including Nepal tion” by which he meant an absence or inadequacy of those (Alkire, Roche, Santos, & Seth, 2011). In this study, they diets, amenities, standards, services, and activities, which are included years of schooling and school attendance for the common or customary in society. This could be understood education dimension, child mortality and nutritional status as the initial debate on a multidimensional aspect of poverty for the health dimension, and cooking fuel, sanitation, water, and deprivation. His measure of deprivation included a list of electricity, and floor and asset ownership for the standard of 60 indicators of the standard of living. The indicators cov- living. They estimated MPI by place of residence and devel- ered diet, clothing, fuel and light, house and housing ameni- opmental region (OPHI, 2013). Their findings suggested that ties, and so on. Sen (2000a), endorsing the need to take a 44.2% of people could be classified as poor in 2011, if pov- multidimensional approach to poverty, writes “Human lives erty was assessed multidimensional. This estimate was much are battered and diminished in all kinds of different ways, higher than the income-based poverty (25.2%) of NLSS and the first task is to acknowledge that deprivations of very (2010-2011). The second study is the NHDR (2014) released different kinds have to be accommodated within a general by the Government of Nepal in collaboration with UNDP. In overarching framework” (p. 18). Thus, a poverty measure this report, human poverty index was estimated (31.1%) based on multiple indicators is more robust than the one using the percentage of people who were not expected to sur- measured based solely on income poverty (Betti, Gagliardi, vive beyond age of 40 years for a long and healthy life; adult & Verma, 2018; Deaton, 2013; Drèze & Sen, 2013; Gordon literacy rate for knowledge; and the percentage of people & Townsend, 1990; Guio et al., 2012; Guio et al., 2016). This without access to safe water, percentage of malnourished suggests that the poverty debate has moved from a unidimen- children below 5 years of age and deprivation in economic sional (income) to a multidimensional approach, which led provisioning for a decent standard of living. Human poverty to the estimation of MPI. index was estimated separately for a rural–urban place of Foster et al. (1984) proposed a systematic methodology residence, five development regions, three ecological zones, for estimating multidimensional poverty for the first time. subregions, and districts. Their approach fueled the debate of measuring poverty and In this study, as used in previous approaches, our measure provided a framework for decomposing poverty. They of multidimensional poverty includes three basic dimensions showed new poverty measure that is additively decompos- of human development—health, education, and economic able with population share weights. Although their work was status. However, our estimates of MPI differ significantly one of the most cited in poverty literature due to its notable from the NHDR (2014) and Alkire, Roche, and Vaz (2014) methodological contribution for decomposing the indicators regarding inclusion and coverage of indicators, their estima- of measuring poverty, it however, did not contribute much in tion procedure, and aggregation of each dimension of MPI. measuring multidimensional poverty. Later on, Alkire and First, we include net enrolment ratio instead of gross enrol- Foster (2007) made a significant methodological ment ratio under education dimension. Second, wealth index improvement. is used to measure economic status as compared with indi- In 2010, Alkire and Santos developed MPI for various vidual standard of living indicators. Finally, each indicator countries for the first time under OPHI. Multidimensional under different dimensions has been measured at the indi- poverty comprises of a number of indicators showing the vidual level rather than at household level. We also carried deprivations experienced by people—such as poor health, out a rigorous screening of variables to identify multiple lack of education, inadequate living standard, lack of income, deprivation items, which adequately meet suitability, valid- disempowerment, poor quality of work and threat from vio- ity, and reliability of the assumed latent dimensions of MPI. lence (Alkire & Foster, 2011a, 2011b). These indicators may We have explained in detail below about the robustness and vary depending upon the context of countries, cultures or inclusiveness of our method of estimation of MPI over 4 SAGE Open NHDR (2014) and Alkire, Roche, et al. (2011) in the context deprivations. Thus, it was important to confirm whether they of Nepal, which is also applicable in the global context. This were suitable to and representative of a single latent dimen- study has largely benefited from the conceptual and method- sion of the standard of living or not. For each of these three ological advancements made in recent studies (e.g., Guio aspects, item selection criteria were not only drawn from the et al., 2012; Guio et al., 2016; Notten & Mendelson, 2016). literature on theoretical basis, but also empirically tested. We also disaggregated multidimensional poverty estimates The items that have successfully passed principles of reli- by place of residence, developmental region, ecological ability were then used in aggregation to MPI (Table 1). In zone, and subregions. addition to this, we have also tested for validity. However, in absence of income data in the same data set, we have assessed it through the macrolevel correlation between individual Data MPI components and income poverty levels (Figure 1). This study has used data from a nationally representative Nepal Demographic Health Survey (DHS) collected in 2010- Education 2011. The sample was designed to provide estimates of most key variables for the 13 ecodevelopment regions and three NHDR (2014) used adult literacy rate for measuring educa- ecological regions (for further details on sampling design, tional deprivation. This measure has been severally criticized see Ministry of Health and Population [MOHP], New ERA, for being not a true indicator of measuring educational depri- & ICF International, 2012). The survey covered a nationally vation (Smith, 1992; UNDP, 1993). Therefore, studies are representative sample of 10,826 households, which yielded increasingly using years of schooling and school enrolment completed interviews with 12,674 women aged 15 years to to measure educational deprivation. Alkire, Roche, et al. 49 years in all selected households and with 4,121 men aged (2011) used both years of schooling and school enrolment as 15 years to 49 years in every second household. What is of indicators of educational deprivation. While the present particular interest to this study is that these data provide study also uses years of schooling and child enrolment ratio, detailed information on mortality, nutrition, sociodemo- but unlike Alkire, Roche, et al. (2011) who used gross enrol- graphic characteristics, access to basic amenities, and house- ments, we have used net enrolments. We used net enrolment hold assets. Furthermore, we have used income poverty and to overcome the limitations of gross enrolment ratio, which multidimensional poverty information from CBS (2011) and could be more than 100% and often fails to measure the true NHDR (2014), respectively, in the validity test of MPI items enrolment ratio. Reliability analyses for the items included in chosen for this study. the dimension of education suggest a high reliability coeffi- cient (RC = 0.8078), meaning that items considered under education dimension probably measure the same underlying Measure of Multidimensional latent concept “education.” Poverty: Rethinking Dimensions and Computation Health We have estimated MPI using information from indicators of three main dimensions—education, access to health, and While NHDR (2009) used percent of people not expected to standard of living. However, within these dimensions, our survive beyond 40 years of age as an indicator of health sta- indicators are not the same as used by Alkire, Roche, et al. tus, both Alkire, Roche, et al. (2011) and we used two indica- (2011) and NHDR (2014). tors of health: under-five mortality status and nutritional The dimensions and indicators within each dimension are status. presented in Table 1. We have described the measurement of Under-five mortality is measured as whether a child expe- dimensions and indicators below. The previous studies have rienced death before his or her fifth birthday. Our other mea- used both theoretical and empirical criteria in identifying sures of health are the nutritional status of the child and the suitable items to be included in the MPI (Dickes, 1989; mother. The most convenient and standard method of mea- Eurostat, 2002; Guio, 2009; Perry, 2002; Whelan, 1993). In suring nutritional status of a child is his or her physical this study, suitability is theoretically derived as aptness to growth or the weight-for-age. This index provides informa- reflect low well-being of people across the different prov- tion about growth and body composition and is measured in inces of Nepal. Each indicator within a specific dimension terms of standard deviation units (Z scores) from the median was selected based on reviews of extensive literature in the of the reference population. In addition, we have also consid- context of Nepal. For example, in the case of the first two ered height-for-age of children below 5 years of age to repre- dimensions, suitable items were selected qualitatively that sent chronic undernutrition. Children whose standard score represent a single latent dimension: education or health. (Z score) of weight-for-age (underweight) and height-for-age However, the next dimension, standard of living, was derived (stunting) is below minus two standard deviations (–2 SD) based on factor analyses because there were a large number from the median of the reference population are considered of items that represented different kinds of material as malnourished. For analysis, a dichotomous variable Goli et al. 5 Table 1. Dimensions, Indicators, and Measures of MPI and Their RCs. Reliability analyses NHDR Alkire, Roche, Santos, Number Dimensions Indicators (2014) and Seth (2011) Present study—Measures RC of items Education Adult literacy rate Yes — — — 0.8078 2 Years of schooling — Yes Yes If completed less than 5 years of schooling School enrollment — Yes Yes If a child (aged 6-10 years) is not enrolled in school Health People not expected Yes — — — 0.7220 4 to survive beyond the age of 40 years Under-five mortality — Yes Yes If died before the fifth birthday Nutritional status — Yes Yes If child is underweight (child and woman) (–2 SD from median of reference population) If child has stunted growth (–2 SD from median of reference population) If a woman has BMI < 18.5 kg/m Standard of Malnourished children Yes — Yes — living below the age of 5 years Safe drinking water Yes Yes Yes — Sanitation — Yes Yes — Electricity — Yes Yes — Flooring — Yes Yes — Cooking fuel — Yes Yes — Assets ownership — Yes Yes — Wealth index — — Yes It is a composite index of 0.5796 30 30 household amenities and assets MPI Education (two items), 0.7453 36 health (four items), and wealth Index (30 items) Note. MPI = multidimensional poverty index; NHDR = Nepal Human Development Report; RC = reliability coefficient; BMI = body mass index. This indicator in the present study was used as a part of health dimension. These indicators are being used in the construction of wealth index, a widely known proxy measure of economic status. whether an under-five child’s growth was below –2 SD Wealth Index (coded 0) or above –2 SD (coded 1) for weight-for-age and Household wealth index is used as an indicator of standard height-for-age was created. of living. Wealth index is commonly used in DHS country- Mother’s nutritional status is measured by body mass level studies and considered as a most reliable summary index (BMI). The BMI is categorized into three components: measure (Rutstein, Jhonson, & Gwatkin, 2000). For the pur- thin (BMI < 18.5), normal (BMI between 18.5 and 24.9), pose of this study, wealth index computed in Nepal DHS and obese (BMI > 25). In this study, we considered thin (2011) has been considered as an indicator of standard of women with a BMI less than 18.5 as malnourished. Our indi- living. The coverage of wealth index (MOHP, New ERA, & cators are same as taken by the Alkire, Roche, et al. (2011). ICF International, 2012; Rutstein, 1999, 2008) is much In the case of the dimension of health, reliability analyses for more than the six standards of living indicators (i.e., access the items included reveal a high RC (.7220), meaning that to safe water, sanitation, electricity, flooring, cooking fuel, items considered under health dimension probably measure and assets ownership) used by Alkire, Roche, et al. (2011). the same underlying latent concept “health.” 6 SAGE Open Figure 1. Correlation matrix for the different components of MPI with income-based poverty. Note. MPI = multidimensional poverty index. The wealth index in its current form, which takes a better dividing the ranking into five equal categories, each com- account of urban–rural differences in the scores and indica- prising 20% of the population. In this study, bottom two tors of wealth in the Nepal DHS (2011), is created in three quintiles, poorest and poorer (40%), form the poor. This clas- steps. First, a subset of indicators common to urban and rural sification is widely accepted in demographic research areas was used to create wealth scores for households in both because it has been repeatedly proven in the DHS data–based areas (for a detailed list of indicators see MOHP, New ERA, analyses that these two quintiles show deprived status in sev- & ICF International, 2012; Rutstein, 2008). Categorical vari- eral demographic indicators than other three quintiles ables were transformed into dichotomous (0-1) indicators. (Rutstein, 2008). These indicators were then examined using principal compo- The wealth index is particularly valuable in countries, nent analysis (PCA) method to produce a common factor which lack reliable data on income and expenditures, which score for each household. Second, separate factor scores are the traditional indicators used to measure household eco- were produced for households in urban and rural areas using nomic status (Chakraborty, Fry, Behl, & Longfield, 2016). For an area-specific indicator. Third, the separate area-specific developing countries such as Nepal, with extreme geographi- factor scores were combined to produce a nationally appli- cal difficulties, high poverty incidence, and not so strong sta- cable wealth index by adjusting area-specific scores through tistical information system, wealth index may serve a better a regression on the factor scores. This three-step procedure purpose for measuring economic status (standard of living) permits greater adaptability of the wealth index in both urban than direct income (Gwatkin et al., 2007; Rutstein & Johnson, and rural areas. The resulting combined wealth index has a 2004). However, the recent studies (Guio et al., 2012; Guio mean of zero and a standard deviation of one. et al., 2016) have suggested for testing the reliability and Once the index is computed, national-level wealth quin- validity of dimensional structures and suitability of items used tiles (from lowest to highest) were obtained by assigning the in the construction of material deprivation index such as household score to each de jure household member, ranking wealth index. Following them, we have performed the reli- each person in the population by his or her score, and then ability analyses for the 30 items used in the wealth index Goli et al. 7 construction. The results of reliability analyses suggest, reasons. First, when we estimate child nutritional and child although slightly less, but a satisfactory RC (0.58) between the mortality indicators with the entire sample of child popula- items selected, meaning they correlate to an assumed latent tion as a denominator instead of households, it will overcome concept of “standard of living” in the study. the limitation of excluding the households, which do not have under-five children at the time of the survey. Second, this process avoids the duplication of deprivation in terms of Estimation of MPI health and education across household members. In the case Compared to the initial period of Townsend (1979) and of the index of standard of living, we have estimated it at the Foster et al. (1984), the measure of multidimensional pov- household level and generalized the household index value erty has been advanceing in terms of its dimensional spread to all the household members living in that particular house- and the number of items used in each dimension. In particu- hold by exporting this variable to person file. lar, the introduction of the DHS, a homogeneous and reliable While estimating the composite index, the UNDP (2010) household survey over 80 developing countries, has facili- human development report used geometric mean (GM) to tated the researchers to construct the index on unit-level data obtain HDI. We believe that this method of estimating GM of sets using a wider number of indicators related to both house- each dimension of indices (education, health and standard of holds and individuals. The global MPI becomes widely used living) is a better procedure to derive MPI rather than the measure since 2010 after its regular publication in human simple mean as used by Alkire and Foster (2007) and Alkire, development reports of UNDP. Alkire, Roche, et al. (2011) Roche, et al. (2011). Because GM makes sure that a low have estimated global MPI using three dimensions: educa- achievement in one dimension is not linearly compensated tion, health, and living standard. As per this approach, a per- by high achievement in another dimension. The GM reduces son on each indicator is identified as deprived or not deprived the level of substitutability between dimensions and at the using information for any one household member. Then, it is same time ensures that 1% decline in the index, say, life aggregated across all the household members. This criterion expectancy at birth has the same impact on the MPI as 1% of identifying poor and calculating MPI in their method has decline in education or income index. serious drawbacks. Researchers have often raised questions Thus, we estimate MPI using GM method by the about measurability and aggregation process of indicators in following: the method of MPI calculation proposed by Alkire and her colleagues (Rangarajan & Dev, 2015). For instance, if one MPI = EPIH ×× PI SLPI , (1) () person in a household is undernourished that does not mean all household members are undernourished. Similarly, if one child has not attended school for 5 years or more that does where EPI is education poverty index, HPI is health poverty not mean that other children did not go to school. Moreover, index, and SLPI is the standard of living poverty index. For in the case of under-five mortality, if a household does not comparison, we also estimated MPI using simple mean by have under-five children, such households will not be the following: included in that particular dimension of MPI. According to Nepal DHS (2011), the proportion of such households was as () EPI+HPI+SLPI (2) MPI = . high as 11%. This is the main reason that Rangarajan and Dev (2015) have not estimated MPI using Alkire, Roche, et al.’s (2011) identification approach for neighboring devel- oping country “India.” Rangarajan and Dev (2015) also Thus, our methodology is different from Alkire, Roche, questioned the aggregation method of individual- and house- et al. (2011) and NHDR (2014) in terms of indicators used, hold-level indicators by Alkire, Roche, et al. (2011). Some of estimation and aggregation approach. NHDR (2014) used the recent studies by Guio et al. (2016) and Notten and traditional indicators to measure human poverty index, Mendelson (2016) have made conceptual and methodologi- which is severely criticized by the scholars now (Dhongde cal advancements in the construction of material deprivation & Haveman, 2017; Dutta, 2015; Gordon, 2000; Mitra, 2016; index like one which similar to MPI. Rogan, 2016; Rutstein, 2008; Wang & Wang, 2016). For We concur with Rangarajan and Dev (2015), Guio et al. instance, NHDR (2014) used adult literacy rate for educa- (2016), and Notten and Mendelson (2016), and therefore, tion; percentage of people not expected to survive beyond made some modifications in Alkire and others’ criterion of 40 years of age (life expectancy) for health and percentage aggregation of dimension indices to overcome the above- of people not having access to safe drinking water and child said limitations. In particular, we have estimated indicators undernourishment for the standard of living. The adult lit- at the population level for rural–urban, ecological zones and eracy rate is replaced by the mean years of schooling at subnational level instead of household level in all three UNDP HDI because it does not depict the current scenario dimensions, namely, health, education, and standard of liv- of educational attainment in a country or state. Therefore, ing. This method has merit over the previous method for two UNDP includes school enrollments, which represent current 8 SAGE Open scenario. We considered both mean years of schooling and Differences were also observed across ecological zones net school enrollment ratio. Among health indicators, (namely, mountain, hill, and Terai). A large number of people NHDR (2014) fails to take more sophisticated and sensitive across the ecological regions had less than 5 years of schooling. measures such as child mortality and nutrition. Standard of The figure varies from 48.0% in hills to 59.1% in the mountain living poverty is measured by deprivation in access to drink- areas. Surprisingly, children not enrolled (age 6-10 years) ing water, which is acceptable; but the inclusion of child reflects the reverse situation. The highest number of children undernourishment cannot be justified to be included in the not enrolled (age 6-10 years) found to be the highest in Terai standard of living poverty unless they had tested for its suit- region (10.63%) and lowest in mountain region (1.40%). ability and reliability in this dimension. It also fails to incor- By development regions, the highest proportion of indi- porate several other important variables under the dimension viduals with less than 5 years of education were reported in of the standard of living: sanitation, electricity, housing, and the mid-western development region (55.5%) followed by many others. Therefore, our indicator, especially in the case far-western region (54.4%). The lowest proportion was of the standard of living poverty (measured as wealth index), found in the eastern and western development region which incorporates multiple household assets, is a better (47.7%). Except for central region (13.50%), in other regions indicator. children not enrolled (age 6-10 years) were below 10%. The From methodology (dimensional structure and aggregation far-western region recorded the lowest (2.60%) number of process) point of view, we differ with Alkire, Roche, et al. children not enrolled (age 6-10 years). (2011) on three points: number and type of indicators, their Further disaggregation based on ecodevelopmental region estimation procedure, and process of aggregation of dimen- suggests that individuals had less than 5 years of education in sion indices used in the computation of MPI. In terms of indi- the central mountain (63.7%), western mountain (60.5%), cators, we differ only at three places. First, we have used net far-western hills (56.8%), and central Terai (61.2%) are con- enrollment ratio rather than gross enrollment ratio in the edu- siderably high. On the contrary, the proportion of children cation dimension, an additional indicator height-for-age under not enrolled in school was the highest in central Terai nutritional status of children under health dimension and (19.1%) followed by eastern Terai (6.9%) and western Terai wealth index based on 30 assets instead of Alkire and col- (5.6%). These results show that there is variation in educa- leagues’ standard of living index based on six indicators. tional poverty by rural–urban residence, development Second, we also differ with Alkire and colleagues in terms of regions, and ecodevelopmental subregions. estimation procedure of individual indicators. They have esti- mated deprivation basically at a household level even for indi- Distribution of indicators by geographic regions: Health. By place cators, which conventionally supposed to be estimated at the of residence, among health poverty indicators, as far as individual level. However, we estimated each indicator based under-five mortality is concerned, both rural (4.7%) and on the nature of indicator and considered the unit of analysis urban (4.0%) areas still face a dearth of health facilities as accordingly, that is, if it is an individual kind of variable then reflected in high under-five mortality. Gaps are noticeable in measured at the individual level or if it is household variable the case of nutritional indicators also. More than one third of then measured at the household level then the score distributed total children in rural areas are poor regarding nutritional sta- among the household members. tus whereas in urban areas around one fourth of total children are poor. However, as compared to children, nutritional sta- tus of adult females is slightly better. As per our estimates, Results 18.5% females in rural areas and 13.5% females in urban areas have poor nutritional status. Descriptive Statistics By ecological regions, we found maximum adult females Table 2 presents descriptive statistics of indicators covered with BMI < 18.5 in the Terai region (22.5%), but children under each dimension of MPI by place of residence, ecologi- with poor nutritional status were maximum in the mountain cal regions, developmental regions, and ecodevelopment region (height-for-age 52.8%; weight-for-age 35.5%). subregions. The descriptive statistics show a large difference Under-five mortality was the highest in the mountain region in each of the specific indicators of all three dimensions of (6.4%). As against this, hills were found to be the least poor MPI, namely, education, health, and standard of living. in terms of select health indicators. Among developmental regions, broadly mid- and far- Distribution of indicators by geographic regions: Education. By western region show the greatest extent of deprivations on an place of residence, about 54.4% people do not complete 5 average in all health indicators. In the case of under-five years of schooling in rural areas as compared with 32.2% in mortality and adult females with BMI < 18.5 found to be urban areas. On the contrary, children between 6 years and highest in the far-western region (mortality < 5%-6.60%; 10 years of age not enrolled are 7.7% in rural areas as com- and BMI < 18.5%-23.6%). On the contrary, height-for-age pared with 3.1% in urban areas. The result shows a huge gap and weight-for-age are the highest in the mid-western region in rural–urban educational poverty. (height for age 49.0%; weight for age 36.2%). 9 Table 2. Descriptive Statistics of Indicators, Nepal, 2011. Education Health Standard of living Children not enrolled Wealth status poorest Less than 5 year of (aged 6-10 years; n = Under-five death (n = Height for age Weight for age and poor wealth schooling (n = 36,908) 6,305) 5,306) (moderate; n = 2,392) (moderate; n = 2,392) BMI < 18.5 (n = 6,179) quintile (n = 49,791) 95% CI % 95% CI 95% CI 95% CI 95% CI 95% CI 95% CI Background characteristics % LL UL LL UL % LL UL % LL UL % LL UL % LL UL % LL UL Place of residence Urban 32.2 31.0 33.5 3.1 2.0 4.6 4.0 2.6 6.1 25.4 20.1 31.5 15.2 11.1 20.6 13.5 11.4 15.9 6.3 5.7 6.9 Rural 54.4 53.8 54.9 7.7 7.0 8.4 4.7 4.2 5.4 40.6 38.6 42.7 29.1 27.2 31.0 18.5 17.5 19.6 44.6 44.2 45.1 Ecological zone Mountain 59.1 57.2 61.1 1.4 0.6 2.9 6.4 4.5 9.2 52.8 45.6 59.8 35.5 29.0 42.6 16.0 12.7 19.9 71.8 70.3 73.3 Hill 48.0 47.2 48.8 3.7 3.0 4.5 4.6 3.8 5.6 40.6 37.5 43.7 25.4 22.8 28.3 11.8 10.6 13.2 52.5 51.8 53.2 Terai 52.8 52.1 53.5 10.6 9.6 11.7 4.5 3.8 5.3 36.2 33.6 38.9 28.4 26.0 31.0 22.5 21.1 24.0 25.5 24.9 26.0 Development region Eastern 47.7 46.6 48.7 5.1 4.1 6.4 4.7 3.6 6.0 35.9 32.1 39.9 25.0 21.6 28.7 16.2 14.4 18.1 34.6 33.8 35.5 Central 53.7 52.8 54.6 13.5 12.0 15.0 4.5 3.6 5.6 36.0 32.7 39.4 27.6 24.6 30.8 19.7 18.0 21.5 32.2 31.5 32.9 Western 47.7 46.6 48.8 3.4 2.5 4.6 4.1 3.1 5.6 37.3 33.0 42.0 22.1 18.5 26.2 13.7 11.9 15.6 35.7 34.8 36.6 Mid-western 55.5 54.0 57.0 4.2 3.1 5.8 4.2 3.0 5.8 49.0 43.9 54.2 36.2 31.5 41.3 18.4 15.7 21.4 61.2 60.0 62.4 Far-western 54.4 52.8 56.1 2.6 1.6 4.1 6.6 4.9 8.9 45.6 39.7 51.5 32.5 27.2 38.3 23.6 20.5 27.1 57.9 56.6 59.3 Ecodevelopment subregion Eastern mountain 51.8 48.0 55.5 1.3 0.3 5.4 5.4 2.3 12.0 45.3 31.6 59.8 23.2 13.2 37.6 9.4 5.3 16.4 65.4 62.3 68.4 Central mountain 63.7 60.3 67.0 1.4 0.3 5.9 3.9 1.4 10.2 47.1 32.5 62.2 35.9 22.8 51.4 14.4 9.2 21.8 60.0 57.0 62.9 Western mountain 60.5 57.4 63.5 1.4 0.5 4.3 7.9 5.1 12.2 58.5 48.7 67.7 41.0 31.8 50.8 21.8 16.0 28.8 84.3 82.4 86.1 Eastern hill 51.1 49.3 52.9 3.1 1.9 5.1 5.3 3.5 7.9 43.8 36.6 51.2 27.6 21.5 34.6 11.2 8.7 14.4 61.8 60.2 63.3 Central hill 40.6 39.2 42.0 4.7 3.3 6.6 3.6 2.3 5.7 28.2 22.8 34.4 20.3 15.5 26.0 10.9 8.9 13.4 32.8 31.6 34.0 Western hill 48.9 47.4 50.3 1.7 1.0 3.1 4.8 3.3 6.8 36.0 30.5 41.8 15.9 12.0 20.7 8.1 6.3 10.2 48.8 47.5 50.0 Mid-western hill 53.9 51.6 56.1 5.5 3.7 8.2 3.4 2.0 5.8 50.3 42.8 57.8 36.0 29.1 43.6 17.0 13.3 21.6 72.9 71.2 74.5 Far-western hill 56.8 54.0 59.6 3.8 2.1 7.0 6.7 4.2 10.6 57.3 48.1 66.1 40.3 31.7 49.6 22.6 17.3 28.8 79.4 77.5 81.3 Eastern Terai 45.3 44.0 46.7 6.9 5.4 8.9 4.2 3.0 5.9 30.7 26.2 35.7 24.0 19.8 28.7 19.7 17.2 22.4 16.1 15.2 16.9 Central Terai 61.2 60.0 62.3 19.1 17.0 21.4 5.0 3.9 6.4 38.5 34.4 42.8 30.2 26.3 34.3 25.9 23.5 28.5 28.9 28.0 29.8 Western Terai 46.0 44.3 47.7 5.6 3.9 7.9 3.2 1.8 5.4 39.7 32.4 47.4 32.6 25.8 40.1 20.9 17.8 24.4 17.5 16.4 18.6 Mid-western Terai 55.9 53.7 58.1 3.7 2.1 6.3 3.4 1.9 6.2 42.3 34.3 50.7 31.8 24.6 40.1 20.0 16.0 24.7 41.5 39.6 43.4 Far-western Terai 51.0 48.7 53.3 2.0 1.0 4.1 6.0 3.7 9.7 30.3 22.6 39.4 24.7 17.6 33.5 23.6 19.4 28.4 36.2 34.4 38.1 Nepal 51.2 50.7 51.7 7.2 6.3 8.1 4.7 4.1 5.3 39.2 37.3 41.2 27.8 26.1 29.6 17.8 16.9 18.8 39.6 39.1 40.0 Source. Author’s calculations based on Nepal DHS 2011 data. Note. BMI = body mass index; LL = lower limit; UL = upper limit; Nepal DHS = Nepal Demographic Health Survey. 10 SAGE Open In the case of ecodevelopmental subregion, western moun- followed by hills and Terai regions. Surprisingly, educational tain region has experienced the highest number of under-five poverty is the highest in the Terai region (0.237). Again, deaths (7.9%) and underweight children (41.0%). However, stark differences can be noticed in the case of wealth poverty. central Terai region had the highest number of adult females About three fourth people are poor in mountain region as with BMI < 18.5 (25.9%). Under-five mortality was the least compared with only one half and one fourth in hill and Terai in the western Terai (3.20%), whereas underweight children regions, respectively. and adult females with BMI < 18.5 were the lowest in the western hill. Thus, there is no single subregion, which has MPI across development regions. The results of MPI by the performed better in all health indicators. development region reveal that mid-western (0.262) and far- western (0.247) regions have the highest incidence of pov- Distribution of indicators by geographic regions: Standard of living. By erty, whereas, western region (0.188) has the least. place of residence, huge poverty differences are evident in rural– Educational poverty is the highest in the central region urban areas in the case of wealth status. The share of poorer and (0.269) followed by the eastern region (0.1560). Here again, poorest in terms of wealth status is more than 7 times in rural the educational poverty is found to be the lowest (0.119) in areas than urban areas. About 6.3% people are found to be in the far-western region, but in contrast to it, this region has the poorer and poorest wealth quintile in urban areas as against of highest health poverty (0.219) and second-highest standard 44.6% in rural areas. of living poverty (0.579). Similar trends for enrolment were By ecological regions, high variation was noticed in the case also observed under different survey reports (Asia-Pacific of wealth status. Around 71.8% people are poor in mountain Cultural Center for UNESCO [ACCU], 2001; Ministry of region followed by hills (52.5%) and Terai region (25.5%). Health and Population et al., 2012; NLSS, 2011). However, Wealth poverty is 3 times higher in mountain region than Terai. differences in health poverty among other regions than far- In the case of developmental regions, the incidence of western are not very significant. But pronounced divergence wealth poverty was just double in the mid- and far-western can be noticed again in the case of poverty in terms of stan- regions as compared with other developmental regions. dard of living. Poverty in terms of standard of living is nearly About 60% people were poor in terms of wealth in mid- and double in mid-western (0.612) and far-western (0.579) far-western regions as compared with around 35% in other regions as compared with other three regions. Furthermore, three regions. This indicates the unequal distribution of differences can also be seen in the overall MPI score by wealth across developmental regions. developmental regions. Things are not different in ecodevelopment subregions. The degree of variation in wealth poverty can be imagined from MPI across ecodevelopment subregions. Table 3 presents the the number of people belonging to poor and poorest wealth MPI and relative rank of subregions. Far-western hill is the quintile ranges from 16.1% (eastern Terai) to 84.3% (western poorest (0.305) among all followed by mid-western hill mountain). Far-western hills (79.4%) and mid-western hills (0.283). Trends in the individual dimensions suggest greater (72.9%) were other two subregions with huge wealth poverty. variations in the case of educational and standard of living All Terai-related subregions exhibit low wealth poverty. poverty. Poverty differences between least and most educa- tionally poor are 0.260, whereas in the standard of living poverty are 0.682. However, differences are much less in the MPI case of health poverty, which is only 0.131. Standing of sub- Table 3 provides estimates of MPI by place of residence, regions in terms of individual dimensions also differs signifi- ecological regions, development regions and subregions. cantly. For instance, central Terai (0.342) and estern mountain (0.082) in education, western mountain (0.254) and western MPI by rural–urban place of residence. Overall MPI estimate hill (0.122) in health and western mountain (0.843) and east- for rural areas (0.254) is about 2.5 times greater than in urban ern Terai (0.161) in the standard of living poverty are the areas (0.091). While 25% people in the rural areas are expe- most and least poor regions, respectively. riencing multidimensional poverty, this proportion is only 10% in the urban areas. Specifically, the results show that Income-Based Poverty Versus Multidimensional rural areas are at a disadvantageous position in comparison Poverty with urban areas in all three dimensions, namely, education, health and wealth status. The highest difference in rural– In this section, we compare the income-based poverty mea- urban poverty was found in wealth status. Poverty in terms of sure with MPI. Poverty is the lack of resources over time, wealth status in rural areas (0.446) is nearly 7 times higher whereas MPI is a consequence of deprivation in various than their urban counterparts (0.063). resources. There is a complementary and dynamic relation- ship between these two indicators. In developed countries MPI across ecological regions. Estimates by ecological region context, considerable variation were observed between reveal that MPI is the highest in the mountain region income-based poverty estimates and MPI, although the 11 Table 3. Multidimensional Poverty by Place of Residence, Nepal, 2011. Standard of living poverty Multidimensional poverty Education poverty index (EPI) Health poverty index (HPI) index (SLPI) index (MPI) 95% CI 95% CI 95% CI 95% CI Background characteristics Region Score LL UL Score LL UL Score LL UL Score LL UL Rank Place of residence Urban 0.10 0.08 0.12 0.12 0.09 0.16 0.06 0.06 0.07 0.09 0.07 0.11 2 Rural 0.20 0.19 0.21 0.18 0.17 0.19 0.45 0.44 0.45 0.25 0.24 0.27 1 Ecological region Mountain 0.09 0.06 0.13 0.21 0.17 0.26 0.72 0.70 0.73 0.24 0.19 0.29 1 Hill 0.13 0.12 0.15 0.15 0.14 0.17 0.53 0.52 0.53 0.22 0.20 0.24 2 Terai 0.24 0.22 0.25 0.18 0.16 0.20 0.26 0.25 0.26 0.22 0.21 0.23 2 Development Eastern 0.16 0.14 0.18 0.16 0.14 0.19 0.35 0.34 0.36 0.21 0.19 0.23 4 region Central 0.27 0.25 0.29 0.17 0.15 0.20 0.32 0.32 0.33 0.25 0.23 0.26 2 Western 0.13 0.11 0.15 0.15 0.12 0.18 0.36 0.35 0.37 0.19 0.17 0.21 5 Mid-western 0.15 0.13 0.18 0.19 0.16 0.23 0.61 0.60 0.62 0.26 0.23 0.30 1 Far-western 0.12 0.09 0.15 0.22 0.18 0.26 0.58 0.57 0.59 0.25 0.21 0.29 2 Ecodevelopment Eastern mountain 0.08 0.04 0.17 0.15 0.08 0.26 0.65 0.62 0.68 0.20 0.13 0.31 8 subregion Central mountain 0.09 0.04 0.20 0.18 0.10 0.29 0.60 0.57 0.63 0.22 0.13 0.33 6 Western mountain 0.09 0.05 0.17 0.25 0.19 0.33 0.84 0.82 0.86 0.27 0.20 0.36 3 Eastern hill 0.13 0.10 0.16 0.16 0.12 0.21 0.62 0.60 0.63 0.23 0.19 0.28 5 Central hill 0.14 0.11 0.17 0.12 0.09 0.16 0.33 0.32 0.34 0.18 0.15 0.21 10 Western hill 0.09 0.07 0.12 0.12 0.09 0.16 0.49 0.48 0.50 0.18 0.15 0.21 10 Mid-western hill 0.17 0.14 0.21 0.18 0.13 0.24 0.73 0.71 0.75 0.28 0.24 0.34 2 Far-western hill 0.15 0.11 0.20 0.24 0.18 0.32 0.79 0.78 0.81 0.30 0.25 0.37 1 Eastern Terai 0.18 0.15 0.20 0.16 0.13 0.19 0.16 0.15 0.17 0.16 0.14 0.19 13 Central Terai 0.34 0.32 0.37 0.20 0.17 0.23 0.29 0.28 0.30 0.27 0.25 0.29 3 Western Terai 0.16 0.13 0.19 0.17 0.13 0.22 0.18 0.16 0.19 0.17 0.14 0.20 12 Mid-western Terai 0.14 0.11 0.19 0.17 0.13 0.24 0.42 0.40 0.43 0.22 0.17 0.27 6 Far-western Terai 0.10 0.07 0.15 0.18 0.13 0.25 0.36 0.34 0.38 0.19 0.15 0.24 9 Total 0.19 0.18 0.20 0.17 0.16 0.19 0.40 0.39 0.40 0.24 0.22 0.25 Source. Author’s calculations based on Nepal DHS 2011 data. Note. LL = lower limit; UL = upper limit; Nepal DHS = Nepal Demographic Health Survey. 12 SAGE Open Table 4. Income-Based Poverty Versus Multidimensional Poverty in Nepal, 2011. Poverty (%) Background Multidimensional Multidimensional Income-based Human poverty index, 2010- a a b c characteristics Region poverty (AM) 2011 poverty (GM) 2011 poverty (2010-2011) 2011, by NHDR (2014) Place of residence Urban 12.83 9.11 15.5 18.5 Rural 32.96 25.38 27.4 34.0 Ecological region Mountain 43.24 23.91 42.3 38.5 Hill 32.98 22.08 24.3 29.2 Terai 26.70 22.13 23.4 33.0 Development Eastern 27.15 20.59 22.3 29.2 region Central 29.25 24.63 21.7 31.5 Western 26.85 18.82 21.4 27.2 Mid-western 39.33 26.20 31.7 36.6 Far-western 37.83 24.71 45.7 34.8 Subregion Eastern mountain 37.59 20.13 — 30.7 Central mountain 39.29 21.51 — 37.5 Western mountain 49.18 26.99 — 29.3 Eastern hill 36.96 23.35 — 30.2 Central hill 23.73 17.70 — 24.7 Western hill 30.10 17.58 — 25.6 Mid-western hill 43.09 28.26 — 38.2 Far-western hill 47.14 30.50 — 42.1 Eastern Terai 20.62 16.48 — 29.5 Central Terai 31.32 26.90 — 39.4 Western Terai 22.47 16.89 — 29.7 Mid-western Terai 31.89 21.81 — 32.5 Far-western Terai 27.95 18.75 — 28.4 Total 30.39 23.64 25.2 31.1 a b c Source. Author’s estimates based on Nepal DHS 2011 data. CBS (2010-2011), Nepal. NHDR, 2014. Note. AM = arithmetic mean; GM = geometric mean; NHDR = Nepal Human Development Report; Nepal DHS = Nepal Demographic Health Survey; CBS = Central Bureau of Statistics. relationship is strictly not linear. Both Peter Townsend (1979) poverty by CBS (2010-2011). Our poverty estimates for the and Mack and Lansley (1985) used the relationship between urban area (9.11%) show lesser poverty as compared with income and deprivation to choose their MPI items (cited in CBS 2010-2011 (15.5%) and NHDR (2014; 18.5%) but in Guio et al., 2016). But existing literature in developing coun- the case of rural poverty, the differences are not very signifi- tries such as Nepal hardly gives any evidence on the nature cant between our estimates (25.38%) and CBS 2010-2011 of the relationship between income poverty and multidimen- (27.4%). However, rural poverty shown by NHDR (2014) is sional poverty. This study fills this decisive gap. Table 4 about 10% higher than our MPI (GM) estimates. In the case compares estimates of multidimensional poverty estimates of classification based on ecological regions, mountain from our study (by the arithmetic mean method and GM region showed the highest poverty by all measures, but high method) with income-based poverty estimates by CBS variability in estimates across the methods is visible. Only (2011) and Nepal human poverty estimates by NHDR (2014). 23.91% people are experiencing multidimensional poverty Our estimates of overall MPI (GM) are similar to income- as per MPI (GM) whereas the corresponding figure by MPI based poverty. For instance, 23.64% people are poor as per (arithmetic mean) is 43.24%, by CBS 2010-2011 is 42.3% our estimates, and the corresponding figure by income-based and 38.5% by NHDR (2014). The same is the case with hill measure is 25.2%. This provides a kind of validation to our and Terai regions. A similar contrast in terms of variability of results. estimates by different methods is also visible in development Meanwhile, MPI estimates by the arithmetic mean are regions. similar to NHDR (2014) estimates. As per NHDR (2014), In the case of subregions, there is a change in relative posi- poverty is 31.1%, whereas as per our estimates, it is 30.39%. tion of subregions in terms of least and most multidimen- Rural–urban poverty differences by MPI (GM) are about sional poor. As per MPI (GM), far-western hill subregion 15%, which is around 12% in the case of income-based (30.50%) is the region with the highest multidimensional Goli et al. 13 44% in Alkire et al. (2013) in case of Nepal. However, we have no doubt in the fact that there is scope to improve the measure further given the availability of data, especially, on the new forms of assets and direct income information in the same data set. The estimates at subnational level suggest that geographi- cal location still works as a major determining factor in pov- erty as mountain region has the highest multidimensional poverty. The disparity between rural and urban poverty is significant. Despite substantial progress in the reduction of income poverty in recent years, multidimensional poverty in rural areas of Nepal remains slightly higher. It may have seri- ous implications because more than 80% of the population in Nepal still lives in rural areas. However, a careful observa- Figure 2. Relationship between MPI and income-based poverty. tion of different dimensions of multidimensional poverty Source. Figure generated based on the estimates showed in Table 4. shows that it is the rural deprivation in the standard of living, Note. MPI = multidimensional poverty index. which is a major contributor to rural–urban differences in the total MPI. These results are in tune with some of the previous poor and eastern Terai (16.48%) is the subregion with least studies (Bhurtel, 2013; Jerve, 2001; NHDR, 2009; Wagle, multidimensional poor. On the contrary, far-western hill 2007), which also noted the poor standard of living in rural (42.1%) and central hill subregions (24.7%) were the most areas, which contributes majorly to multidimensional rural and least multidimensional poor, respectively, as per NHDR poverty in the country. (2014). However, the education status seems to be not related to Overall, compared to developed countries, in Nepal, we the geographical location as our study indicates the highest found less variation in the levels of income-based poverty net enrollment in the mountain region followed by hills and and multidimensional poverty (Figure 2). This may be Terai regions, whereas the NHDR (2014) shows that hills because with an increase in absolute income levels with eco- have the lowest educational poverty. The educational pov- nomic growth, the relative disparities in material access erty seems to be indicator sensitive where NHDR (2014) increases. With the introduction of every new technology poverty index takes illiteracy rate into consideration; the and resources, richer tend to access faster than their counter- present study has taken years of schooling and net enroll- part, thus, leads to rising disparities, which also give rise to ment ratio, which are better indicators than adult literacy and increase in relative deprivation or poverty (Deaton, 2013). gross enrollment ratio. Similar contrasting results in educa- However, in low-income countries still, the overall material tional poverty are also observed by development regions and access relative to developed countries is less. Therefore, subdevelopment regions. These results assume greater there is less gap between incomes-based poverty and multi- importance in the context where Nepal government fixed dimensional deprivation. their targets of educational attainment under development plans in terms of enrollment ratios, not literacy; our results provide more robust basis for effective policy planning and Discussion and Conclusion implementation. The results are also robust because regions Unlike several recent studies on multidimensional poverty, with a high number of children not enrolled (6-10 years of which used global MPI procedure of estimation, we have age) coincide with the high number of people attending less proposed an alternative approach for measuring MPI. In this than 5 years of schooling. approach, poverty in terms of each indicator has been mea- It is also noticeable that educationally poor regions (Terai) sured at the individual level rather than at household level. are the regions, which border with the most backward states Moreover, the aggregation procedure used in this study is (Uttar Pradesh and Bihar) of the neighboring country “India” similar to the procedure used in HDI calculation. As pointed regarding education and economic progress. There may be out in the methods section, this procedure overcomes the some common explanations and linkages behind the low limitations of global MPI used by UNDP-HDR reports. Our development process in the regions, which require deeper results are methodologically robust in terms of both numbers empirical investigation (Samuels, Nino-Zarazua, Wagle, of indicators taken for each dimension and method of aggre- Sultana, & Sultana, 2011; Srinivasan, 2012). One possible gation. Our methodological improvements show that current explanation is that migration is very high in the Terai region methodology of estimating multidimensional poverty par- (49% highest among all regions), especially, the male migra- ticularly in the case of Nepal tends to deflate poverty num- tion (49.8%), which is also the case with neighboring Indian bers by a significant margin. For instance, multidimensional states (Nepal DHS, 2011). Nepalese leave their homes at an poor by GM aggregation method in this study is 24% against early age in expectation of earning a livelihood and, thus, do 14 SAGE Open not give due attention to educational attainment (Samuels features still are important determinant of multidimensional et al., 2011). poverty. This also shows that the government policies in Although interregion differences can be noticed in health Nepal mainly focus on the economic growth and employ- poverty, it is comparatively lesser than educational and stan- ment generation and ignore to bridge the gap between rich dard of living poverty. The heavy presence of international and poor. At the same time, the government also failed to agencies, that is, International Center for Integrated remove income inequalities as stated in Tenth Plan document Mountain Development (ICMOD), World Health (NHDR, 2014; Wagle, 2007). Organization (WHO), and United Nations Children’s Fund, In terms of comparability across the methods, it is impor- and so on and the introduction of many health-related pro- tant to note that CBS (2011) estimates of poverty are essen- gressive schemes by the Nepal government in the last two tially income-based poverty, which is comparable with our decades have brought the much-needed improvement in standard of living poverty estimates rather than overall MPI health status across all the regions, especially in the disad- estimates. Income is the source as well as the outcome of the vantageous regions such as mountain (Bentley, 1995; Engel wealth of a household. Past wealth helps in income genera- et al., 2013). Engel et al. (2013) noted, tion and thus generated income helps in wealth accumula- tion. Nonetheless, this feeding loop is based on the condition A consistent policy focus and sustained financial commitment that income level must be above their consumption expendi- by the government and donors throughout the past two decades, ture level. The income level of two households or individu- including substantial increases in funding for maternal health als may be same, but their wealth level may differ since the early 1990s, has allowed for widespread improvements significantly. Income level often does count for generational in access to medical services, particularly in remote areas. accumulation and transfer of wealth. Therefore, income may show the lesser incidence of poverty than wealth. This Thus, Nepal’s experience in the health sector can provide is exactly the case with Nepal. Poverty regarding wealth important lessons for other developing countries, especially, status showed the relatively higher incidence of poverty the South Asian Association of Regional Cooperation among different geographical areas (Figure 3). Even, coun- (SAARC) nations struggling with high levels of maternal try-level poverty in terms of wealth status differs signifi- mortality and poor health facilities, mainly within a circum- cantly than income poverty. Our measure of standard of stance of difficult terrain and high-income poverty. living poverty is comprehensive as it is based on wealth Regarding wealth status, poverty levels are higher across all index, which has been prepared by including multiple the regions. The high poverty in terms of wealth status indi- household assets. The estimates from this study show the cates poor conditions of housing, sanitation, electricity, drink- greater extent of wealth poverty compared with OPHI ing water, and other basic amenities. Wealth poverty contributes (2013). However, multidimensional poverty levels are much to significant differences in the overall multidimensional pov- less than wealth index–based poverty levels, which suggest erty of different place of residence and regions. Wealth poverty that Government of Nepal is operating a number of social in rural areas is 7 times higher than urban areas, whereas it is 3 security measures in the form of provision of basic educa- times higher in mountain than Terai region. Similarly, it is tion and health services to all the populations, which might almost double in the far-western and mid-western region as work as counteracting to wealth-based deprivation. compared with other developmental regions. Stark differences In conclusion, this study had examined the extent of multi- were also observed across the subregional classification. These dimensional poverty in Nepal disaggregated by geographic findings from the study are corroborated by other studies, regions. It adopted the more robust method of MPI compared which also noted the significant rise in income and wealth with global MPI of UNDP, particularly regarding indicators, inequalities during last three decades across the rural–urban their definitions, dimensional structure, and aggregation pro- regions (NHDR, 2014; Wagle, 2007; World Institute for cedure than that of the previous studies. It also took into con- Development Economics Research [WIDER], 2005; World sideration the latest methodological improvements in Bank, 2006). Bhurtel (2013) argued that it implies two things: calculating deprivation index measures by Guio et al. (2012), Guio et al. (2016), and Notten and Mendelson (2016). To con- First, the labour share of the national income has declined over clude, the findings of the study suggest that although Nepal time while the share of capital has rapidly increased. The dull has experienced a decent economic progress and a consider- growth of agriculture and stagnant manufacturing has mainly able reduction in education and health poverty with a consid- contributed to the growing economic inequalities. Secondly, the erable increase in wealth inequalities across the regions, government (Nepal) has failed to take fiscal measures to reduce overall MPI remains high. A far less has been achieved in the income inequality. Monetary measures such as providing cheap and easy credit to the poor have been largely ineffective. case of reducing the standard of living poverty, that is, wealth poverty and inequalities across the regions. Thus, the article High incidence of the standard of living poverty in the suggests that development policies and poverty reduction pro- mountain areas is also a reflection of poor provision of basic grams in Nepal must aim to reduce multidimensional poverty needs of life as well as a confirmation of the fact that natural by geographic regions, of which deprivation in education, Goli et al. 15 Initiative, Oxford Department of International Development, Queen Elizabeth House, University of Oxford. Alkire, S., Roche, J. M., & Vaz, A. (2014). Multidimensional poverty dynamics: Methodology and results for 34 countries (OPHI research in progress). Retrieved from https://ophi.org. uk/ophi-research-in-progress-41a-2/ Alkire, S., & Santos, M. E. (2010, July). Acute multidimensional poverty: A new index for developing countries (OPHI Working Papers 38). Oxford, UK: University of Oxford. Alkire, S., & Santos, M. E. (2013). A multidimensional approach: Poverty measurement & beyond. Social Indicators Research, 112, 239-257. Anand, S., & Sen, A. (1997). Concepts of human development and poverty! A multidimensional perspective. In Poverty and human development: Human development papers (pp. 1-20). New York, Figure 3. Relationship between wealth and income poverty. NY: United Nations Development Programme. Retrieved from Note. Figure is constructed based on the estimates of wealth poverty from https://scholar.harvard.edu/sen/publications/concepts-human- Table 2 and income poverty from Table 4. development-and-poverty-multidimensional-perspective Angulo, R., Díaz, Y., & Pardo, R. (2016). The Colombian multidi- mensional poverty index: Measuring poverty in a public policy health, and basic amenities must be an integral component, context. Social Indicators Research, 127, 1-38. along with their efforts to improve economic growth and Asia-Pacific Cultural Centre for UNESCO. (2001). Regional gen- reduce income and wealth-based poverty. der and ethnic disparity in education. Tokyo, Japan: Author. Retrieved from http://www.accu.or.jp/litdbase/literacy/nrc_ Declaration of Conflicting Interests nfe/eng_bul/BUL18.pdf Atkinson, A. B. (1987). On the measurement of poverty. The author(s) declared no potential conflicts of interest with respect Econometrica: Journal of the Econometric Society, 55, 749-764. to the research, authorship, and/or publication of this article. Atkinson, A. B. (1970). On the measurement of inequality. Journal of Economic Theory, 2, 244-263. Funding Bader, C., Bieri, S., Wiesmann, U., & Heinimann, A. (2016). A dif- The author(s) received no financial support for the research, author- ferent perspective on poverty in Lao PDR: Multidimensional ship, and/or publication of this article. poverty in Lao PDR for the years 2002/2003 and 2007/2008. Social Indicators Research, 126, 483-502. ORCID iD Bardhan, P. K. (1970). On the minimum level of living and the rural poor. Indian Economic Review, 5, 129-136. Moradhvaj https://orcid.org/0000-0002-8798-385X Bardhan, P. K. (1971). On the minimum level of living and the rural poor: A further note. Indian Economic Review, 6, 78-87. References Bentley, H. (1995). The organisation of health care in Nepal. Abel-Smith, B., & Townsend, P. (1965). The poor and the poorest. International Journal of Nursing Studies, 32, 260-270. London, England: G. Bell. Betti, G., Gagliardi, F., & Verma, V. (2018). Simplified jackknife Alkire, S., Adriana, C., & Roche, J. M. (2013, February). variance estimates for fuzzy measures of multidimensional Multidimensional poverty index 2013: Brief methodologi- poverty. International Statistical Review, 86, 68-86. cal note and results. Oxford, UK: Oxford Poverty & Human Bhurtel, B. P. (2013, October 17). Rich man’s world as the wealth Development Initiative, Oxford Department of International gap grows in Nepal. The Nation, The Kathmandu Post, Asia Development, Queen Elizabeth House, University of Oxford. News Network. Retrieved from http://nationmultimedia.com/ Alkire, S., & Foster, J. (2007). Counting and multidimensional pov- opinion/RICH-mans-world-as-the-wealth-gap-grows-in- erty measures (OPHI Working Paper Series, No. 7). Oxford, Nepal-30217256.html UK: Oxford Poverty & Human Development Initiative, Bosanquet, H. (1903). The “poverty line.” Charity Organisation University of Oxford. Revised in January 2008. Review, 8, 321-325. Alkire, S., & Foster, J. (2011a). Counting and multidimensional Bourguignon, F., & Chakravarty, S. R. (2003). The measurement poverty measurement. Journal of Public Economics, 95, of multidimensional poverty. Journal of Economic Inequality, 476-487. 1, 25-49. Alkire, S., & Foster, J. (2011b). Understandings and misunder- Central Bureau of Statistics. (2011). Poverty in Nepal. Kathmandu, standings of multidimensional poverty measurement. Journal Nepal: National Planning Commission Secretariat, Government of Economic Inequality, 9, 289-314. of Nepal. Available from www.cbs.gov.np Alkire, S., Foster, J., & Santos, M. E. (2011). Where did identifica- Chakraborty, N. M., Fry, K., Behl, R., & Longfield, K. (2016). tion go? Journal of Economic Inequality, 9, 501-505. Simplified asset indices to measure wealth and equity in health Alkire, S., Roche, J. M., Santos, M. E., & Seth, S. (2011). programs: A reliability and validity analysis using survey data Multidimensional poverty index 2011: Brief methodological from 16 countries. Global Health: Science and Practice, 4, note. Oxford, UK: Oxford Poverty & Human Development 141-154. 16 SAGE Open Clark, S., Hemming, R., & Ulph, D. (1981). On indices for the mea- national poverty strategies (pp. 89-120). New York, NY: surement of poverty. The Economic Journal, 91, 515-526. United Nations Development Programme. Dasgupta, P., Sen, A., & Starrett, D. (1973). Notes on the measure- Kakwani, N. (1980). On a class of poverty measures. Econometrica: ment of inequality. Journal of Economic Theory, 6, 180-187. Journal of the Econometric Society, 48, 437-446. Deaton, A. (2013). The great escape: Health, wealth, and the ori- Mack, J., & Lansley, S. (1985). Poor Britain. London, England: gins of inequality. Princeton, NJ: Princeton University Press. George Allen & Unwin. Dhongde, S., & Haveman, R. (2017). Multi-dimensional depriva- Ministry of Health and Population, New ERA, & ICF International. tion in the US. Social Indicators Research, 133, 477-500. (2012). Nepal Demographic and Health Survey 2011. Dickes, P. (1989). Pauvreté et conditions d’existence: théories, Kathmandu, Nepal: Ministry of Health and Population. modèles et measures [Poverty and living conditions: Theories, Mitra, S. (2016). Synergies among monetary, multidimensional and models and measures]. Walferdange: Centre d’études de popu- subjective poverty: Evidence from Nepal. Social Indicators lations de pauvreté et de politiques socio-économiques. Research, 125, 103-125. Drèze, J., & Sen, A. (2013). An uncertain glory: India and its con- Mohanty, S. K., Agrawal, N. K., Mahapatra, B., Choudhury, D., tradictions. Princeton, NJ: Princeton University Press. Tuladhar, S., & Holmgren, E. V. (2017). Multidimensional Dutta, S. (2015). Identifying single or multiple poverty trap: An poverty and catastrophic health spending in the mountainous application to Indian household panel data. Social Indicators regions of Myanmar, Nepal and India. International Journal Research, 120, 157-179. for Equity in Health, 16(1), 21. Engel, J., Glennie, J., Adhikari, S. R., Bhattarai, S. W., Prasai, D. P., & Montoya, Á. J. A., & Teixeira, K. M. D. (2017). Multidimensional Samuels, F. (2013). Nepal’s story understanding improvements poverty in Nicaragua: Are female-headed households better in maternal health. London, England: Overseas Development off? Social Indicators Research, 132, 1037-1063. Institute. Retrieved from https://www.odi.org/sites/odi.org.uk/ Nepal Human Development Report. (2009). Nepal Human files/odi-assets/publications-opinion-files/8624.pdf Development Report 2009: State transformation and human Eurostat. (2002). Income, poverty and social exclusion: Second development. United Nations Development Programme. report. Luxembourg: Office for Official Publications of the Retrieved from http://www.refworld.org/docid/4a8a72542. European Communities. html Foster, J., Greer, J., & Thorbecke, E. (1984). A class of decom- Nepal Human Development Report. (2014). National Human posable poverty measures. Econometrica: Journal of the Development Report 2014: Beyond geography unlock- Econometric Society, 52, 761-766. ing human potential. Kathmandu, Nepal: National Planning Gordon, D. (2000). The scientific measurement of poverty: Recent Commission. Retrieved from http://hdr.undp.org/sites/default/ theoretical advances. In J. Bradshaw & R. Sainsbury (Eds.), files/nepal_nhdr_2014-final.pdf Researching Poverty (pp. 37-58). Aldershot, UK: Ashgate. Nepal Living Standard Survey. (2011). Nepal-Living Standards Gordon, D., & Townsend, P. (1990). Measuring the poverty line. Survey 2010-2011, third round. Kathmandu, Nepal: Central Radical Statistics, 47, 5-12. Bureau of Statistics, National Planning Commission Secretariat, Government of Nepal. (2011, November). Nepal Living Standards Government of Nepal. Survey 2010/2011 (Statistical Report Vol. I). Kathmandu, Notten, G., & Mendelson, M. (2016). Using low income and mate- Nepal: Central Bureau of Statistics, National Planning rial deprivation to monitor poverty reduction. Ottawa, Ontario, Commission Secretariat, Government of Nepal. Canada: Caledon Institute of Social Policy. Guio, A.-C. (2009, February 10). What can be learned from deprivation Nowak, D., & Scheicher, C. (2017). Considering the extremely indicators in Europe. Paper presented at the Indicator Subgroup poor: Multidimensional poverty measurement for Germany. of the Social Protection Committee. Retrieved from https:// Social Indicators Research, 133, 139-162. www.bristol.ac.uk/poverty/downloads/keyofficialdocuments/ Oxford Poverty & Human Development Initiative. (2013). Nepal Deprivation%20indicators%20in%20Europe.pdf country briefing—Multidimensional poverty index data bank. Guio, A.-C., Gordon, D., & Marlier, E. (2012). Measuring material Oxford, UK: Oxford Poverty & Human Development Initiative, deprivation in the EU: Indicators for the whole population and University of Oxford. Retrieved from www.ophi.org.uk/multi- child-specific indicators (Eurostat methodologies and working dimensional-poverty-index/mpi-country-briefings/ papers). Luxembourg: Office for Official Publications of the Perry, B. (2002). The mismatch between income measures and European Communities. direct outcome measures of poverty. Social Policy Journal of Guio, A.-C., Marlier, E., Gordon, D., Fahmy, E., Nandy, S., & New Zealand, 19, 101-127. Pomati, M. (2016). Improving the measurement of material Rangarajan, C., & Dev, M. S. (2015). Counting the poor: deprivation at the European Union level. Journal of European Measurement and other issues. Economic and Political Weekly, Social Policy, 26, 219-333. 50(2), 70-74. Gwatkin, D. R., Rutstein, S., Johnson, K., Suliman, E., Wagstaff, A., Ravallion, M. (1998). Poverty lines in theory and practice (Vol. & Amouzou, A. (2007). Socio-economic differences in health, 133). Washington, DC: World Bank Publications. nutrition, and population. Washington, DC: World Bank. Ravallion, M., & Huppi, M. (1991). Measuring changes in poverty: Hagenaars, A. (1987). A class of poverty indices. International A methodological case study of Indonesia during an adjust- Economic Review, 28, 583-607. ment period. The World Bank Economic Review, 5, 57-82. Hanandita, W., & Tampubolon, G. (2016). Multidimensional pov- Reyles, D. Z. (2010). The human development index (HDI). Oxford, erty in Indonesia: Trend over the last decade (2003–2013). UK: Oxford Poverty & Human Development Initiative. Social Indicators Research, 128, 559-587. Rogan, M. (2016). Gender and multidimensional poverty in South Jerve, A. M. (2001). Rural-urban linkages and poverty analysis. Africa: Applying the global multidimensional poverty index In A. Grinspun (Eds.), Choices for the poor: Lessons from (MPI). Social Indicators Research, 126, 987-1006. Goli et al. 17 Rutstein, S. O. (1999). Wealth versus expenditure: Comparison Townsend, P. (Ed.). (1971). The concept of poverty. London, between the DHS wealth index and household expenditures England: Heinemann. in four departments of Guatemala. Calverton, MD: ORC Townsend, P. (1979). Poverty in the United Kingdom. Macro. Harmondsworth, UK: Penguin Books. Rutstein, S. O. (2008, October). The DHS wealth index: Approaches Townsend, P. (2010). The meaning of poverty. The British Journal for rural and urban areas (The DHS Working Papers, No. 60). of Sociology, 61(Suppl. 1), 85-102. Demographic and Health Research, United States Agency for Tsui, K. Y. (2002). Multidimensional poverty indices. Social International Development. Retrieved from https://dhsprogram.com/ Choice and Welfare, 19, 69-93. pubs/pdf/WP60/WP60.pdf Uematsu, H., Shidiq, A. R., & Tiwari, S. (2016, September). Trends Rutstein, S. O., & Johnson, K. (2004). The DHS wealth index (DHS and drivers of poverty reduction in Nepal: A historical perspec- Comparative Reports No. 6). Calverton, MD: ORC Macro. tive (Policy Research Working Paper No. 7830). Washington, Retrieved from https://dhsprogram.com/pubs/pdf/cr6/cr6.pdf DC. World Bank. Retrieved from https://openknowledge. Rutstein, S. O., Johnson, K., & Gwatkin, D. (2000, March). Poverty, worldbank.org/handle/10986/25146 health inequality, and its health and demographic effect. United Nations Development Programme. (1993). Human develop- Presented at the annual meeting of the Population Association ment report 1993. New York, NY: United Nations Development of America, Los Angeles, CA. Programme, Oxford University Press. Samuels, F., Nino-Zarazua, M., Wagle, S., Sultana, T., & Sultana, M. United Nations Development Programme. (2010). Human develop- M. (2011, November). Vulnerabilities of movement: Cross-border ment report 2010–20th anniversary edition. The real wealth mobility between India, Nepal and Bangladesh (Background of nations: Pathways to human development. New York, NY: note). London, England: Overseas Development Institute. Author. Sen, A. (1970). Interpersonal aggregation and partial comparabil- United Nations Development Programme. (2015). Human devel- ity. Econometrica: Journal of the Econometric Society, 38, opment report 2015: Work for human development. New 393-409. York, NY: United Nations Development Programme, Oxford Sen, A. (1972). Interpersonal comparison and partial comparabil- University Press. ity: A correction. Econometrica: Journal of the Econometric United Nations Nepal. (2014). 12-point understanding reached Society, 40(5), 959. between the Seven Political Parties and Nepal Communist Sen, A. (1973). Poverty, inequality and unemployment: Some Party (Maoists). Available from http://www.un.org.np/ conceptual issues in measurement. Economic and Political Wagle, U. R. (2007, October 11). Economic inequality in the Weekly, 8, 1457-1464. “democratic” Nepal: Dimensions and implications. Presented Sen, A. (1976). Poverty: An ordinal approach to measurement. at the Himalayan Policy Research Conference, Madison, WI. Econometrica: Journal of the Econometric Society, 44, Retrieved from https://ejournals.unm.edu/index.php/nsc/ 219-231. article/view/642/739 Sen, A. (1979). Issues in the measurement of poverty. The Wang, Y., & Wang, B. (2016). Multidimensional poverty measure Scandinavian Journal of Economics, 81, 285-307. and analysis: A case study from Hechi City, China. Springer Sen, A. (1981). Ingredients of famine analysis: Availability and Plus, 5, Article 642. entitlements. The Quarterly Journal of Economics, 96, 433-464. Whelan, B. (1993). Non-monetary indicators of poverty. In Sen, A. (1987). Gender and cooperative conflicts (No. 1342). J. Berghman & B. Cantillon (Eds.), The European face of social Helsinki, Finland: World Institute for Development Economics security: Essays in honour of Herman Deleeck (pp. 24-42). Research. Aldershot, UK: Avebury. Sen, A. (1989). Food and freedom. World Development, 17, 769-781. World Bank. (2006). World development indicators 2006. Sen, A. (1992). Poverty reexamined. Cambridge, MA: Harvard Washington, DC: Author. University Press. World Bank. (2014). Poverty and inequality measures in practice: Sen, A. (2000a). A decade of human development. Journal of A basic reference guide with Stata examples. Washington, DC: Human Development, 1, 17-23. Author. Sen, A. (2000b). Social exclusion: Concept, application, and scru- World Bank. (2016). Moving up the ladder: Poverty reduction and tiny (Social Development Papers). Office of Environment and social mobility in Nepal. Retrieved from https:// openknowledge. Social Development, Asian Development Bank. Retrieved worldbank.org/handle/10986/25173 from http://hdl.handle.net/11540/2339 World Institute for Development Economics Research. (2005). Smith, P. (1992). Measuring human development. Southampton, World income inequality database (Version 2.0a). Helsinki, UK: University of Southampton. Finland: World Institute for Development Economics Research, Srinivasan, P. V. (2012, November). Regional cooperation and United Nations University. integration through cross-border infrastructure development in Yu, J. (2013). Multidimensional poverty in China: Findings based South Asia: Impact on poverty (South Asia Working Paper Series, on the CHNS. Social Indicators Research, 112, 315-336. 14). Asian Development Bank. Retrieved from https://www.adb. org/publications/regional-cooperation-and- integration-through- Author Biographies cross-border-infrastructure-development Thon, D. (1979). On measuring poverty. Review of Income and Srinivas Goli is an assistant professor of population studies at JNU, Wealth, 25, 429-439. New Delhi. He did his master’s and PhD in population studies from Townsend, P. (1954). Measuring poverty. The British Journal of IIPS, Mumbai, India. He is teaching fertility studies, family plan- Sociology, 5, 130-137. ning, family demography, gender and quantitative methods in social 18 SAGE Open science research at JNU. His research areas are formal demography, New Delhi and IIPS, Mumbai respectively. His research interests family demography, ageing, inequalities in health, nutrition and gen- are mortality, health care, health financing and gender & develop- der status and its social determinants. Currently, he is involved in ment. He has published four journal articles in highly reputed inter- number of collaborative projects in India and abroad. national journals. Nagendra Kumar Maurya is an assistant professor at the Prem Bhandari is social researcher at the Population Research Department of Applied Economics, University of Lucknow, Center, University of Michigan, Ann Arbor, Michigan. He did his Lucknow, India. He did his master’s and PhD in Applied Economics PhD in rural sociology and demography from the Pennsylvania from University of Lucknow, India. His research interest includes State University. His research interest includes social research state finances, public economics, human development, macroeco- methods, socio-economic and cultural determinants of migration, nomics, poverty and inequality. fertility, and population health, rural social change; population and Moradhvaj is a doctoral student of population studies at JNU, New environment, sociology of agriculture, and rural development Delhi. He did master and MPhil in population studies from JNU, planning.

Journal

SAGE Open – SAGE

Published: Mar 13, 2019

Keywords: multidimensional poverty; Nepal; MPI; DHS; regional differentials

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

Regional Differentials in Multidimensional Poverty in Nepal: Rethinking Dimensions and Method of Computation:

References (60)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies