3 ms·
Uhm, they define poverty as living on less than one dollar per day. Then they show it dropping over the last 20 years ... but don't control for inflation. The
by wheels 16y ago
Uhm, they define poverty as living on less than one dollar per day. Then they show it dropping over the last 20 years ... but don't control for inflation. They show 10% improvement against a benchmark that (aside from being meaningless as a way of measuring abject poverty) lost more than 50% of its value and label it progress. This to me suggests basic innumeracy. It's entirely unclear from the data presented whether real poverty has been reduced at all.
- jacobolus 16y agoThese statistics are normalized across countries and over time by purchasing power parity. I don’t know what the precise calculation is, but it is relatively sophisticated. I’m sure if you search around you could find the details. They are just reported in simpler terms to make the comparison easily comprehensible to the non-technical. See http://en.wikipedia.org/wiki/Poverty_threshold http://en.wikipedia.org/wiki/Poverty_threshold and also http://en.wikipedia.org/wiki/Geary-Khamis_dollar http://en.wikipedia.org/wiki/Geary-Khamis_dollar (In other words, there are at least a few statisticians working for the UN and the World Bank, and they tend not to be “basically innumerate”.)
- wheels 16y agoHere's the actual paper, which does name its units (international poverty line): http://www.salaimartin.com/media/pdf/Africa_Poverty_NBER.pdf http://www.salaimartin.com/media/pdf/Africa_Poverty_NBER.pdf And criticism of their results from someone working on producing similar numbers for the World Bank: http://blogs.worldbank.org/africacan/is-african-poverty-falling http://blogs.worldbank.org/africacan/is-african-poverty-fall... I must say that it irks me that they would use such a easy-to-understand-and-wrong ($1/day, when they mean "below the international poverty line") unit in writing without specifically qualifying it.