By Eduardo Ortiz-Juarez
What is next in terms of poverty reduction? This is a central question in the current debate on what objectives, targets, and indicators should be part of the new millennium development agenda. In a previous post, I noted that the design of the new global agenda for poverty reduction should emphasize the ownership of its objectives into national policies, as well as a greater flexibility to adapt global goals to local contexts. Recently, Sabina Alkire (OPHI, University of Oxford) and Andy Sumner (King's College London) have aimed at the development of a new acute multidimensional poverty index (MPI 2.0) to this new agenda, in addition to the extreme monetary poverty indicator of $1.25 a day.
Using a multidimensional index is relevant to public policy for at least two reasons. On the one hand, allows the decomposition in dimensions and subgroups. The first shows which dimensions lead to a lower level of deprivation: access to clean water, increased levels of education, etc. The second allows the analysis of where such accomplishments are being achieved: rural or urban zones, Afro-descendant child population, etc. On the other hand, people facing nonmonetary shortages not necessarily have incomes below the poverty line (and vice versa). Thus, focusing the objectives of the new agenda exclusively on monetary poverty reduction may fail in reducing or eradicating other forms of poverty, which can adversely affect the potential development of these people.
In this context, I consider that empirical efforts should be focused not only in creating an acute MPI which complements the extreme monetary indicator of $1.25 a day, but also which adapts to local contexts in order to improve the efficiency of the use of these indicators in countries’ public policy. In Latin America and the Caribbean (LAC), for example, this means targeting higher thresholds that complement, for example, the monetary indicators of $4 a day (the average of poverty lines of LAC) or in Central and Eastern Europe of $5 a day. Along this line, Ravallion (2013) argues that the indicator of $1.25 is becoming less relevant to a growing number of countries whose national lines are well above this value. Analogous to this, you might consider building a relative MPI that adjusts to the extent that welfare in a country or region shifts.
Increasing the thresholds in LAC means recognizing that the relevance of poverty is tending to disappear when focusing on very low indicators, forgetting that poverty is relative to the local context and that the consequences go beyond the lack of money. In the case of monetary poverty, according to the international indicator of extreme poverty of $1.25, just 6.5% of the regional population is in this situation. Instead, two higher indicators are currently being employed: $2.5 a day for extreme poverty and $4 a day for total poverty. According to these,17 and 30 percent of people in the region are in these situations, respectively. Figure 1 illustrates a similar example for the multidimensional case.
By using the current MPI which includes acute deprivation indicators in the dimensions of education, health, and standard of living, the incidence reaches 2.7%, 28% and 1.7% in Brazil, Nicaragua and Uruguay, respectively. However, it rises to 27%, 64% and 6%, respectively, if for the same dimensions, more demanding indicators are employed. For example, in the dimension of standard of living, the current acute MPI considers as household deprivation: no access to electricity, using manure, wood or coal as fuel for cooking, having dirt floors, not having improved or own toilet, not having clean drinking water or having sources to get it more than 30 minutes away on foot, not having more than one basic commodity (radio, TV, telephone, bike or motorcycle), and not having a car or tractor. On the other hand, the 'Specific MPI for LAC' included deprivation such as: lack of potable water connection inside the house, not having toilets with direct discharge of water, having precarious materials in floors, ceilings and walls of the house, and not having a car or truck.
Source: HDR (2011) and own estimates based on home survey.
Indicators as those used in the so-called 'Specific MPI for LAC' may or may not be representative of the standard of living in other regions (or even LAC itself). For this reason, additional empirical efforts are needed to determine the type of relevant indicators in each country or region, and its deprivation thresholds. In this sense, the examples of Colombia, Mexico, and at a sub-national level Brazil and El Salvador, or Sumner and Alkire’s proposal for promoting participatory processes where 'the voices of the poor and marginalized' guide decisions, are very relevant.
The development of an MPI with higher thresholds to respond to specific contexts, hand in hand with the existing monetary indicators, could provide us with a different view on how poverty and its dimensions evolve over time. This will help us to expand the story of 'less poverty due to growth and redistribution' to one of 'lower social deprivation due to comprehensive social policy'.