Latest Issue
Vol 8, Issue 3, September 2017 Global Policy’s September 2017 issue contains, among others, research articles on the BRICS’ new development bank, the G20 and climate change, the human right to health, and digital diplomacy. It has a special section on ‘Policy Instruments for Innovation, Investment and Global Trade’

Should Poverty be defined by a Single International Poverty Line, or Country by Country? (and what difference does it make?)

Andy Sumner - 17th December 2012
Should Poverty be defined by a Single International Poverty Line, or Country by

International poverty lines are calculated by the World Bank: $1.25 per day per person is said to represent the ‘absolute poverty line’, below which a person can hardly survive.

This is calculated from the mean of the national poverty lines for the poorest 15 countries. A slightly higher line, set at $2 per day per person, is the average of the national poverty lines for all developing countries.

To date, these lines have been accepted as the universal poverty metric, underpinning global goals such as the MDGs – at least MDG 1a – and discussions on how the world is doing in reducing poverty. But that is increasingly coming into question. One problem is that poor people increasingly live in middle-income countries, and so are not represented as long as the extreme poverty calculation sticks with the poorest 15 countries (which are actually only home to about 10% of the world’s absolute poor).

Another problem is that while international poverty lines allow us to compare like with like in monetary terms, at national level, all countries define poverty for themselves, often using different approaches.

Poverty levels in OECD countries, for instance, are often defined relative to median income (e.g. below 60 percent of the median for the EU). By contrast poverty lines in other countries are “absolute” – that is, the income required to meet basic needs (e.g. Afghanistan).

Unsurprisingly, given these issues, the recent update of global poverty estimates by $1.25 and $2 poverty measures has had a few critiques, while alternative multidimensional poverty measures are also – rightly – attracting attention as they gather speed and are taken up.

So what difference does it make if instead, we look at how many poor people there are in the world, based on how poverty is defined in the countries where those people live (rather than by international poverty lines)?

To answer this, we added up all the country-level poverty data based on national poverty definitions to produce a new and different perspective on global poverty, based on national measures from 160 countries.

What did we find? Three things:

First, there are 1.5 billion people living in nationally-defined poverty, a billion of whom are in middle-income countries (MICs). This corroborates the view that global poverty has increasingly become a middle-income country phenomenon, although much of this is down to just five countries – the PINCIs – Pakistan, India, Nigeria, China, and Indonesia (see here and here and here).

Second, when poverty is defined nationally, one in ten (170m) of the world’s poor live in high-income countries. (Of course one could question comparability- absolute and relative poverty – so when we present the global poverty data we do so with and without high –income countries). By its own definition, the US has 45m poor people.

Third, while global totals are the same, the overall number conceals some big national variations in poverty numbers depending on whether national or international lines are used (see bar chart). In Mexico and Bolivia, for example, poverty rates according to national lines are more than 40 percentage points higher than those based on the international $1.25/day measure. In Africa, by contrast, for various countries (e.g. Uganda, Tanzania, Liberia, Burundi, Nigeria and Malawi) poverty rates resulting from international lines are much higher than from national measures (e.g. about 35 percentage points higher in Tanzania and 20% in Malawi). And in India 45m people are missing in national poverty estimates that would be counted by international poverty measures.

Why might a focus on nationally-defined poverty be useful? Three reasons:

First, such a focus might fit better with the domestic task of forging national social contracts, as poverty increasingly becomes about national inequality. National poverty measures tend to be what matters most to policy makers in-country. However, those domestic measures are moving closer to the international $1.25 line in some countries – for example in China.

Second, expressing poverty in national terms implies a greater degree of involvement of national actors in defining “what is poverty?” in a given context.That has practical consequences: in a number of cases, countries are testing how to better connect national measures with eligibility for domestic social protection programs.

Third, the recognition that poverty (relative or absolute) exists everywhere entails a shift in poverty thinking – framing poverty as a universal issue relevant to all countries, rather than a ‘them and us’ question.

As the discussion on poverty measurement and classification evolves, it might be interesting to broaden the range of countries, so that highly-populated MICs, or at least the 10-20 countries where most of the poor live, become the basis for the calculation of international standards for absolute poverty.

This could entail establishing global lines (e.g. current $1.25/day) not on the average of lines of the poorest countries, but on the average of the countries with the highest numbers of poor people – 80% of the world’s poor live in just 10 countries and 90% in twenty populous countries (most of which are not currently part of the $1.25 calculation).

Such a process would perhaps better synchronize global measurement to the shifts in global poverty.

This post is co-authored with Ugo Gentilini (World Food Programme).The post also appears on Duncan Green’s blog on Oxfam’s From Poverty to Power.