Conceptual framework

• Author: Giulio Caperna - Eleni Papadimitriou
• Pages: 2-4

2

1 Introduction

The Individual Deprivation M easure (IDM) framework defines a gender-sensitive, multidimensional me asure of

poverty. The measure assesses deprivation at the individual level, in relation to 15 key dimensions of life, namely

Food, Water, Shelter, Health, Education, Ene rgy/fuel, Sanitation, Relationship s, Clothing, Violen ce, Family planning ,

Environment, Voice, Time- Use and Work.

It offers information additional to other national surve ys, providing a high level summary of deprivation through

an index while enabling users to gain further understanding through the decomposition and disaggregation of the

scalar, gende r-sensitive, individual-le vel data on which it is base d.

The IDM was deve loped through an initial 4-year cross-disciplinary inter national research coll aboration, led by the

Australian National University ( ANU), in partnership with International Women’s Development Agency (IWDA) and

the Philippine Health and Soc ial Science Associat ion, University of Color ado at Boulder, and Oxfam Gre at Britain

(Southern Africa), w ith additional support from Oxfam America and Oslo University. The research sought to answer

the question what is a just and justifiable measure of poverty that is gender sensitive and capable of revealing

gender disparities. The first IDM study beyond the proof-of-concept trial was carried out in F iji (2014–17) by IWDA

working with Fiji Bureau of S tatistics (FBoS) (Mcine rney, Fisk, & Cr awford, 2019).

European Commission’s Competence Centre on Composite Indicators and Scoreboards (COIN) at the Joint Research

Centre (JRC) was invited by the International Women’s Development Agency (IWDA) to audit the IDM study

concerning the Fiji (20 14-17) dataset. The J RC statistical audit is based o n the recom mendations of t he Handbook

on Composite Indicators (OECD and JRC, 200 8) as well as on more recent research from the JRC. The Handbook

offers a 10-step guidance on the process of buil ding composite indicat ors. The statistical audit presented herein

aims to contribute to ensuring the transparency of the IDM methodology and the reliability of the results, with the

additional purpose of contr ibuting to the improvements of the methodology and scoring that are under

development by the current IDM program.

The audit of t he IDM focuses on two main issues: the statistical coherence of the conceptual framework (Section

3) and the impact of key modelling assumptions on the IDM ranking (Section 4, Sec tion 5). The audit foll ows three

main steps: the first focuses on the main descriptive statistics and on the detection of missing values and potential

outliers; the second on the analysis of the statistical coherence through an analysis of t he correlations of the

indicator, themes and dimensions; and the third, on the sensitivity analysis of the index and the testing of the

impact of key modelling assumptions. The results are supported by a spreadsheet in Excel format shared with the

IDM team.

The results of the audit presented herein aim at shedding light on the transparency and reliability of the IDM. It

expects to contribute to enable policymakers and advocates to derive more accurate and meaningful conclusions

as well as to guide choices on priority setting and policy formulation.

2 Conceptual framework

The Individual Deprivation Measure (IDM) is a tool meant to assess deprivation at the individual level and

complement current approaches, which measure poverty at the household level. It takes into account 15 dimensions

of individual deprivation , comprised of 23 themes, base d on 34 individu al indicators that are de fined using

responses to one or more survey questions (Tab le 1). The dimensions are categorised into three groups of

importance and weighted accordingly; the 5 most important dime nsions (dimensions 1-5, orange color) receive a

total weighting of 50%, the 5 next-most important dimensions (dim ensions 6-10, blue color) receive a total

weighting of 33% and the remaining 5 (dimensions 11-15, light blue color) receive a total weighting of 17%.