Chapter 6 transitioned from considerations about selecting a measurement methodology (Chapters 3-5) to issues met in implementing real-world measures that undergird and reinforce policies to fight poverty.
It mentioned in passing the desiderata criteria that indicators be ‘technically solid'—a criterion that entails rigorous consideration of properties and also of empirical techniques.
To take this forward, we now switch focus to emphasize the practice of empirical poverty measurement. In particular, this chapter introduces empirical issues that are distinctive to counting-based multidimensional poverty methodologies. Novel issues include the requirement that indicators accurately reflect deprivations at the individual level—not just on average—and that all indicators be transformed to reflect deprivations in the chosen unit of identification (person, household).This chapter is not exhaustive. It presumes readers have a sound understanding of household survey data and their quantitative analysis, and also of various assessments of indicator validity. It supplements a presumed solid foundation with new elements that pertain to multidimensional poverty measurement design and analysis in particular.[204] A more extensive and detailed version is available at www.multidimensionalpoverty.org.
Section 7.1 introduces very briefly the different types of data sources available, namely, censuses, administrative records, and household surveys. Section 7.2 discusses issues to be considered when selecting the indicators to include in a multidimensional poverty measure. Finally, Section 7.3 presents some basic descriptive analytical tools that can prove helpful in exploring the relationships between different indicators and informing the process of the measure design. Box 7.1 discusses the different fronts on which data collection should be improved in the near future in order to permit the design of better poverty measures.
7.1