Category Archives: OPHI Working Papers

Transitions in Poverty and Deprivations: An Analysis of Multidimensional Poverty Dynamics

This paper explores a novel way to analyse poverty dynamics that are specific to certain measures of multidimensional poverty, such as the “adjusted headcount ratio” proposed by Alkire and Foster (2011a). Assuming there is panel data available, I show that a simultaneous and comprehensive account of transitions in deprivations and poverty allows complex interdependencies between dimensions in a dynamic context to be handled and, at the same time, allows for several advanced types of analyses. These analyses include (i) a decomposition of changes in multidimensional poverty, which reveals why poverty decreases or increases; (ii) a framework to examine and understand the relationship between the dashboard approach and dimensional contributions and multidimensional poverty in a dynamic setting; (iii) a presentation of methods that illuminate the process of the accumulation of deprivations. The suggested types of analyses are illustrated using German panel data. The implications for monitoring, policy evaluation and strategies for analyses using repeated cross-sectional data are discussed.

Citation: Suppa, N. (2017). “Transitions in poverty and deprivations: An analysis of multidimensional poverty dynamics.” OPHI Working Paper 109, University of Oxford

Measuring Malnutrition and Dietary Diversity: Theory and Evidence from India

Adequate nutrition constitutes one of the most basic dimensions of human well-being. Ample evidence exists for the functional link between a diverse diet and health outcomes or economic performance. However, a concise measure to capture nutritional diversity that utilizes typical household-level data, often the only data available in developing countries, is yet to be developed. In this paper, I propose a theoretical framework for such a measure by extending the Alkire-Foster (AF) methodology. The new framework enables the calculation of both the incidence and intensity of nutritional deprivation. Applying this framework, I construct a Nutritional Deprivation Index (NDI) for Indian states using household survey data on food consumption. The NDI is unique, and, compared to existing measures, it is more effective in both identifying the inadequately nourished and revealing the extent of food deprivation.

Citation: Oldiges, C. (2017). “Measuring malnutrition and dietary diversity: Theory and evidence from India.” OPHI Working Paper 108, University of Oxford.

Measures of Human Development: Key Concepts and Properties

The measurement of human development is a rich field which has seen a veritable explosion of new and innovative international indices. With the advent of the Sustainable Development Goals’ and their emphasis on interlinkages across deprived conditions, and with the promise of a ‘data revolution’ it seems that the proliferation of new indices that track multiple inter-related phenomena will continue. This paper sets out two basic set of criteria that, we believe, would be very helpful for structuring new policy metrics, and would provide human development experts, statisticians, and others with tremendously useful concepts to bear in mind when assessing and using different societal indices. The first criteria, ably articulated in Foster Seth Lokshin and Sajaia recently, is to clarify whether a human development indicator measures well-being, inequality, or poverty. These features of any population are distinct and each are of singular importance, but it is also useful not to confuse them. The second criterion is to ensure that the structure of the index is clearly explained. In particular, we discuss the importance of clarifying four methodological features: whether an index can be broken down by indicator; whether it can be disaggregated by population subgroup; whether it reflects the joint or overlapping conditions of a person, or evaluates dimensions one by one, and what kinds of weights or values are used to construct the composite index. While these criteria may seem, on the face of it, rather dry, a clear answer to each is essential (and also rather powerful) to understanding the policy relevance of each index.

Citation: Alkire, S. (2016). “Measures of Human Development: Key concepts and properties.” OPHI Working Paper 107, University of Oxford.

Composite Indices, Alternative Weights, and Comparison Robustness

Composite indices are widely used and can be highly influential. Yet most remain controversial owing to inter alia the arbitrary selection of component weights. Several studies have proposed testing the robustness of rankings generated by composite indices with respect to alternative weights but have not provided sufficient guidance on the choice of these alternatives. This paper proposes a holistic yet theoretically novel approach for selecting sets of alternatives weights and assessing comparison robustness that is applicable to linear composite indices with any finite number of dimensions. This approach is applied to robustness testing of inter-temporal country improvements generated by arguably the world’s most influential composite development index, the UNDP Human Development Index (HDI). More than two-thirds of HDI country improvements between 1980 and 2013 were found to be not robust to the selected set of alternative weights.

Citation: Seth, S. and McGillivray, M. (2016). “Composite indices, alternative weights, and comparison robustness.” OPHI Working Paper 106, University of Oxford.

Growth and Poverty Revisited from a Multidimensional Perspective

The actual impact of economic growth on poverty reduction is of fundamental importance to the development agenda, especially in view of the Sustainable Development Goals. So far, studies have focused on income poverty. This paper offers new empirical evidence on growth and poverty measured from a multidimensional perspective using the global Multidimensional Poverty Index. Results from a first difference estimator model suggest that while economic growth reduces multidimensional poverty, this impact is well below a one-to-one relationship. We also find that economic growth has a far bigger impact on reducing income poverty than on reducing multidimensional poverty. Results from an alternative cross-section model also support this result and additionally suggest that countries with higher levels of exports, a higher share of industry and services in their GDPs, and higher control of corruption have lower multidimensional poverty. All in all, the results highlight the need for countries to grow in order to reduce poverty, but they simultaneously suggest the limited power of economic growth per se to achieve grand reductions in poverty.

Citation: Santos, M. E., Dabus, C., and Delbianco, F. (2016). “Growth and poverty revisited from a multidimensional perspective.” OPHI Working Paper 105, University of Oxford.


Chad Multidimensional Deprivation and Vulnerability Survey

This paper summarises the results of The Chad Multidimensional Deprivation and Vulnerability Survey (EPMVT), fielded from June to August 2012 in Chad. The EPMVT was designed to provide information on standard socio-economic variables and five additional ‘missing dimensions’ of poverty data, namely, dimensions that people living in poverty continue to mention as constituent parts of their experience yet for which there are no comparable international data. Specifically, the EPMVT collected information on the characteristics of the household’s dwelling unit, education, employment, disabilities, fertility levels and mortality, breastfeeding and infant feeding practices, early child marriage, and hand-washing practices. In addition, data were collected on work quality, physical safety, empowerment, dignity, and psychological wellbeing. These data have two additional advantages for poverty analysis. First, the EPMVT was designed to be merged with a sub-sample of the 2011 Consumption and the Informal Sector in Chad survey (ECOSIT III). This merger facilitates having consumption data alongside the new variables tested in the EMPVT, as well as taking advantage of other relevant socio-economic variables present in the ECOSIT survey. Second, the EPMVT collected information from the head of the household and all eligible women age fifteen and older from the same household, making intra-household comparisons possible.

Citation: Oxford Poverty and Human Development Initiative (OPHI). (2016). “Chad Multidimensional Deprivation and Vulnerability Survey.” OPHI Working Paper 104, University of Oxford.

Multidimensional Poverty in Europe 2006–2012: Illustrating a Methodology

Multidimensional approaches to poverty and deprivation have a long and distinguished history in conceptual and philosophical work (Sen 1992). This chapter explores multidimensional poverty using EU-SILC data from 2006 to 2012. We calculate a multidimensional poverty index based on the Alkire Foster (AF) methodology – a widely used flexible methodology which can accommodate different indicators, weights and cut-offs. We draw on existing Europe 2020 indicators, as well as on indicators of health, education and the living environment. Aggregated and country cross sectional results are presented. A short analysis of dynamics of multidimensional poverty is also included.

Citation: Alkire, S. and Apablaza, M. (2016). “Multidimensional poverty in Europe 2006–2012: Illustrating a methodology.” OPHI Working Paper 74, University of Oxford.

Comparing Monetary and Multidimensional Poverty in Germany

This paper compares Germany’s official income-based poverty measure with a multidimensional poverty index based on the Alkire-Foster method. For their comparative assessment, I employ the capability approach as a conceptual framework. I find both measures agree on certain aspects, such as socio-demographic risk factors. However, I also document a substantial mismatch in who is deemed poor, which seems to originate from inherent, conceptual features of the measures. More generally, the results also suggest additional individual income reduces multidimensional poverty, if only at a decreasing rate. Examining regional variations, I find that the measures do not agree about trends in poverty and that there is no clear-cut link between aggregate income and multidimensional poverty. I conclude that, despite some basic agreement, the choice of poverty measure makes a difference, with properly designed multidimensional poverty indices having the advantage of better reflecting human well-being.

Citation: Suppa, N. (2016). “Comparing monetary and multidimensional poverty in Germany.” OPHI Working Paper 103, University of Oxford.

Dimensional and Distributional Contributions to Multidimensional Poverty

The adjusted headcount ratio M0 of Alkire and Foster (2011a) is increasingly being adopted by countries and international organizations to measure poverty. Three properties are largely responsible for its growing use: Subgroup Decomposability, by which an assessment of subgroup contributions to overall poverty can be made, facilitating regional analysis and targeting; Dimensional Breakdown, by which an assessment of dimensional contributions to overall poverty can be made after the poor have been identified, facilitating coordination; and Ordinality, which ensures that the method can be used in cases where variables only have ordinal meaning. Following Sen (1976), a natural question to ask is whether sensitivity to inequality among the poor can be incorporated into this multidimensional framework. We propose a Dimensional Transfer axiom that applies to multidimensional poverty measures and specifies conditions under which poverty must fall as inequality among the poor decreases. An intuitive transformation is defined to obtain multidimensional measures with desired properties from unidimensional FGT measures having analogous properties; in particular, Dimensional Transfer follows from the standard Transfer axiom for unidimensional measures. A version of the unidimensional measures yields the M-gamma class  containing the multidimensional headcount ratio for g = 0, the adjusted headcount ratio M0 for g = 1, and a squared count measure for g = 2, satisfying Dimensional Transfer. Other examples show the ease with which measures can be constructed that satisfy Subgroup Decomposability, Ordinality, and Dimensional Transfer. However, none of these examples satisfies Dimensional Breakdown. A general impossibility theorem explains why this is so: Dimensional Breakdown is effectively inconsistent with Dimensional Transfer. Given the importance of Dimensional Breakdown for policy analysis, we suggest maintaining the adjusted headcount ratio as a central measure, augmented by the squared count measure or other indices that capture inequality among the poor. The methods are illustrated with an example from Cameroon.

Citation: Alkire, S. and Foster, J. (2016). “Dimensional and distributional contributions to multidimensional poverty.” OPHI Working Paper 100, University of Oxford.

From Multidimensional Poverty Measurement to Multisector Public Policy for Poverty Reduction: Lessons from the Colombian Case

This paper focuses on the analysis of conceptual, normative and institutional issues of the Colombian Multidimensional Poverty Index (C-MPI). The principal questions are the following: What is the decision-making process that lies behind the Colombian experience? What are the main lessons from the Colombian case in terms of institutional arrangements for the implementation of its index? Although the medium and longer-term effects of the C-MPI on poverty reduction are still to be seen and thus to be evaluated, there are some important lessons. First, a multidimensional poverty index’s utility in terms of public policy depends not only on the mathematical and statistical robustness guaranteed by the Alkire Foster methodology but also on the ability of the policy maker to represent the public policy priorities through its normative choices. Second, acknowledging the inherent trade-offs involved in conceptual, statistical and public policy concerns is key to accurately defining the purpose of the measure. Finally, if the purpose of the MPI is to stimulate coordinated action to reduce poverty, an accurate design will not be enough; it is also necessary to provide a solid institutional architecture that supports the process from the design of the index to its application.

Citation: Angulo, R. (2016). “From multidimensional poverty measurement to multisector public policy for poverty reduction: lessons from the Colombian case.” OPHI Working Paper 102, University of Oxford.