Presentation - Measuring poverty over time

Launch of the special issue of the Journal of Economic Inequality

Launch of the special issue of the Journal of Economic Inequality: ‘Measuring poverty over time’

The conceptualization and measurement of poverty has been the subject of intense study for more than a century. The adoption of the Millennium Development Goals (MDGs) in 2000 gave additional stimulus to these longstanding efforts, this time at the global level, and the progress made towards reducing world poverty is encouraging. World Bank estimates, for example, suggest that the percentage of the population in the developing world living below US$1.25 a day declined from 52 per cent in 1981 to 22 per cent in 2008, bringing the number of people living below this line down from 1.94 billion in 1981 to 1.29 billion in 2008. These numbers are now based on over 850 household surveys for almost 130 developing countries, representing 90 per cent of the population of the developing world, compared with only 22 surveys for 22 countries when the first such estimates were reported in the 1990 World Development Report.

Alongside improvements to the quantity and quality of data, there have also been significant advances in conceptualizing and measuring poverty. One noticeable shift is the movement away from an exclusive focus on consumption-poverty, with attention now often also given to non-income dimensions of wellbeing, such as literacy achievements, nutritional health status, and individual empowerment. This has led to lengthy debate concerning the way in which various dimensions of wellbeing are best combined into a single multidimensional poverty measure.

But, even staying within the uni-dimensional consumption-poverty domain, there are many unresolved issues which deserve attention, especially regarding the time dimension in poverty. Consider, for instance, two individuals who are currently the same distance below the (consumption) poverty line. One person has spent his whole life in poverty; the other has just dropped below the poverty line for the first time. Labelling the two individuals equally poor would seem an inadequate characterization of their relative poverty status, unless current consumption fully captures the cumulative effects of a lifetime of poverty, which is unlikely. If nothing else, both individuals most likely face quite different poverty prospects in the future.

The conceptual limitations of the static ‘snapshot’ approach to poverty measurement have long been recognized, giving rise to the distinction between chronic and transitory poverty and the development of accompanying measures. When the chronic and transient components of poverty are distinguished, attention usually focuses on the average consumption shortfall or the proportion of time spent in poverty across a person’s lifetime, ignoring the sequencing and duration of poverty spells. From this perspective, a person having spent the first half of their life in poverty, a person having spent the last half of their (equally long) life in poverty, and a person having spent half their life in poverty in total, though with poverty spells alternating with non-poverty spells, would be equally poor and would be assigned equal degrees of chronic and transient poverty. This is questionable. Long periods in poverty may cause permanent damage (and thus affect a person’s poverty prospects) in a way that cannot be properly captured by the snapshot and components approach to poverty measurement. The timing of consumption shortfalls during a person’s lifecycle may also matter, with early childhood poverty proving to be especially detrimental in terms of cognitive development as well as the person’s future earnings profile. It is argued that both the duration and sequencing of the poverty spells in a person’s lifetime matter in assessing that person’s poverty status, and by extension, the poverty status of the population the person belongs to. The literature has only just begun to grapple with the challenges of incorporating and formalizing these more sophisticated intertemporal perspectives in measures of poverty, let alone their practical application.

The six studies in this special journal issue take an intertemporal perspective on poverty measurement. They were first presented at the UNU-WIDER conference on ‘Frontiers of Poverty Analysis’, in September 2008 in Helsinki, and have since been extensively revised and polished following feedback from the audience and academic peer review.

It is clear that a more complete understanding of poverty requires much more attention to the time dimension. As the studies illustrate, it is also clear that many theoretical and practical questions remain unresolved in measuring poverty over time, offering many fascinating opportunities for further investigation.


Welcome and Introduction to the JOEI Special Issue
• Tony Addison (Chief Economist and Deputy Director, UNU-WIDER)
• Branko Milanovic (Member of JOEI Editorial Board)

New Findings in the JOEI Special Issue
• Luc Christiaensen (Guest Co-Editor of the JOEI Special Issue) | View Presentation
• John Hoddinott (Deputy Director, PHND, IFPRI) | View Presentation

Implications of the Findings: Policy and Research Perspectives
• Jaime Saavedra (Director, PRMPR, World Bank) | View Presentation
• Peter Lanjouw (Research Manager, DECPI, World Bank) | View Presentation

Discussion with Speakers
• Branko Milanovic, Chair
Panelists: Jaime Saavedra, Peter Lanjouw, Tony Addison,
John Hoddinott, and Luc Christiaensen

Closing Remarks
• Tony Addison

This event was part of a new World Bank series on innovations in poverty measurement and analysis, organized by the Bank’s Poverty and Inequality Measurement and Analysis Practice Group, in joint collaboration with UNU-WIDER and the Journal of Economic Inequality.