Researchers have developed new and more robust ways to measure global extreme poverty, based on people’s access to essential goods. This data presents a more complex – and more troubling – story of poverty than existing narratives would suggest.
In international development circles, most people are familiar with the World Bank’s data showing that extreme poverty has declined dramatically over the past several decades, from 43 per cent of the world’s population in 1981 to less than 10 per cent today. This narrative is based on the World Bank’s method of calculating the share of people who live on less than $1.90 per day (in 2011 “PPP” terms).
But a growing body of literature argues that the World Bank’s PPP-based method suffers from a major empirical limitation, in that it does not account for the cost of meeting basic needs in any given context (see here, here and here). Having more than $1.90 PPP does not guarantee that a person can afford the specific goods and services that are necessary for survival.
In recent years, scholars have developed a more accurate method for measuring extreme poverty, by comparing people’s incomes to the prices of essential goods in each country (specifically food, shelter, clothing and fuel). This approach is known as the “basic needs poverty line” (BNPL), and it more closely approximates what the original concept of “extreme poverty” was intended to measure. The OECD has published BNPL poverty figures for most countries. Robust data exists for the years during which both household surveys and direct price data are available, which generally covers the period 1980-2008 (more on this below).
A new time series of global extreme poverty
While the OECD publication is the most reliable source for BNPL time series data, the published global series does not include the original price data for one of the world’s largest economies, China. With Michail Moatsos, the author of the OECD work, we incorporated the original price data for China in a recent paper in New Political Economy, presenting a complete picture of global poverty using the BNPL approach. Here, we connect this with additional BNPL data for the year 2011, drawing on research by Robert Allen. Allen relies on different sources of price data than the OECD paper, which may affect the trend from 2008 to 2011. However, we have chosen to combine these datasets because they are the best available estimates of extreme poverty for the periods they cover.
The BNPL data indicates that the story of global poverty over the past few decades is more complex, and more troubling, than existing narratives allow. Figure 1 shows that global extreme poverty increased quite substantially during the period of liberalisation and structural adjustment in the 1990s. Progress resumed in the 2000s. In 2011, around 17% of humanity could not afford basic essentials, down by less than 6 percentage points from 1980. The number of people in extreme poverty increased from 1.01 billion to 1.20 billion over this period.
It’s important to note here that the OECD publication also includes poverty estimates for the entire period from 1820-2018. Crucially, though, these estimates are not always based on direct data. For most countries, household survey data does not exist prior to about 1980.* Instead, the OECD figures use historical GDP growth rates as a proxy for changes in household consumption during this period. This approach faces significant limitations, however, as GDP growth rates do not adequately represent changes in non-commodity forms of consumption, particularly during periods of colonisation and liberalisation. Moreover, in the OECD publication, food prices are generally not available after 2008. For the post-2008 period, figures are calculated on the assumption that food prices moved in line with CPI, which is quite often not the case. These figures must therefore be treated with caution. (For more on these issues, see footnote 2 in our New Political Economy paper, as well as Appendix I (1.1) in this recent paper in World Development.)
What the latest findings mean
The estimates that we show here have important implications for the Millennium Development Goals. MDG1 set out to halve the proportion of the global population living in extreme poverty between 1990 and 2015. Figure 1 shows that the poverty rate was 21.2% in 1990. If we accept the assumption that food prices moved in line with CPI, the global extreme poverty rate in 2015 would have been 12% (according to the OECD estimates). In other words, according to the BNPL approach, MDG1 was missed, albeit narrowly. These results stand in contrast to the World Bank’s claims that the target was achieved much earlier, in 2010, and exceeded by a wide margin in 2015. It is worth noting that the 2015 figure may not be accurate, as it is unlikely that basic food prices have moved in line with the CPI.
While robust BNPL data does not exist after 2011, we do have data from the UN Food and Agricultural Organization on food insecurity. It shows that the proportion of the world population experiencing “moderate and severe” food insecurity increased steadily during the period 2014 to 2020, from 23% to 30% (see Figure 2). As access to food is central to the BNPL method, we may assume that post-2011 poverty trends have probably not improved much, if at all.
Extreme poverty is not a natural condition, but a sign of severe dislocation. Historical data on real wages since the 15th century indicates that under normal conditions, across different societies and eras, people are generally able to meet their subsistence needs except during periods of severe social displacement, such as famines, wars, and institutionalised dispossession, particularly under European colonialism. What is more, BNPL data shows that many countries have managed to keep extreme poverty very close to zero, even with low levels of GDP per capita, by using strategies such as public provisioning and price controls for basic essentials.
In other words, extreme poverty can be prevented much more easily than most people assume. Indeed, it need not exist at all. The fact that it persists at such high levels today indicates that severe dislocation is institutionalised in the world economy – and that markets have failed to meet the basic needs of much of humanity. To address this problem, and to end extreme poverty – the first objective of the Sustainable Development Goals – will require public planning to prioritise the production of, and guarantee access to, the specific goods and services that people need to live decent lives.
*Survey data becomes available at different times for different countries – eg, 1977 for India; 1980 for Madagascar; 1981 for China and Brazil; and so on. We choose to start our analysis in 1980, when coverage begins to pick up and when extrapolated years are close enough to direct data that major inaccuracies are unlikely.
The Reader's Digest Guide to Intimate Relations 