Pol Eco of Stats: World Bank's New $1.90 Poverty Line

New targets require new measures of poverty, right?

There has been, in the memorable characterization of Robert Wade, a "political economy of statistics" concerning poverty and inequality. As standard-bearers of neoliberal, market-led globalization, the World Bank would of course claim that the world's decades-long towards putting faith in market mechanisms has resulted in a global reduction of both. Robert Wade begs to differ. Needless to say, methods of measuring these "bads" remain contentious.

There are of course all sorts of ways to measure poverty. For quite some time now, the World Bank has kept an easily-remembered shorthand of expenditures of less than $1.25 on a PPP basis. This, of course, was preceded by the famous $1/day standard before 2008. There are thus many complaints about this standard: sure $1 is a nice round number, but can such inter-country comparisons really be made even after adjusting for purchasing power?

Well, guess what: just as people were still arguing about the $1/day standard when the $1.25/day standard came along, the World Bank is now about to introduce a $1.90/day standard. As we transition from the Millennium Development Goals (MDGs) anchored on the $1.25 standard to the Sustainable Development Goals (SDGs) which replace the MDGs this year, I suppose the development experts came to the conclusion that we need a new international poverty line as well:

The World Bank is to make the most dramatic change to its global poverty line for 25 years — raising its measure by a half to about $1.90 per day — in a move likely to swell the statistical ranks of the world’s poor by tens of millions. The move from $1.25 would be the biggest revision since the World Bank introduced its $1 a day yardstick of global poverty in 1990...

The bank is expected to follow the event by shifting its poverty line to about $1.90 ahead of its annual meetings in Lima, Peru, in early October — a move likely to result in significant shifts in the estimated size and distribution of the planet’s poor. It is difficult to predict exactly how many more people will be defined as poor. However, when researchers at the bank tested a notional poverty line of $1.92 earlier this year, it led to a surge of 148m.

Most of the difference came in east Asia where the ranks of those falling below the poverty line almost doubled from 157m at the old $1.25/day measure to 293m. In Latin America, the result was an increase of 8m, or more than 25 per cent, in the number of poor to 37m, while in south Asia the ranks of the poor grew by 7m to 407m. Under that line, sub-Saharan Africa remained steady at some 416m.

Interestingly, there are now complaints coming from the right. If you assume that globalization has reduced inequality in recent years--and most orthodox economists would say it did--then why change the standard? The new accusation is that the World Bank would not have any work otherwise "helping" the newly (re-)classified folks qualifying as poor as per its mission of helping to eradicate poverty:

Angus Deaton, the Princeton economist and persistent critic of a poverty line that he argues has been misleading for years. “You’ve got a line that no one knows where to put it, PPPs that change, and underlying data that is bad,” he said. “It is sort of a statistical problem from hell.” The World Bank’s administering of the poverty line also carried a hint of conflict of interest, he said, as the bank’s main task was fighting poverty, and its very existence depended on its own poverty measures.

Mr Deaton added: “I think they have some institutional bias towards finding more poverty rather than less.” 

Talk about the political economy of statistics: from the left you have folks like Wade saying these measurements are bogus. On the right they say these statistics and bogus and self-serving.  There's just no pleasing some people.