A Matter of Definition

The Economist ran a column about the poverty line in its May 24th edition. People have grown accustomed to the catchphrase "a dollar a day", but a study conducted by the World Bank recently indicated that prices in some countries were much higher than previously thought, which suggests the definition of the poverty line (technically, of $1.08 at 1993 purchasing-power parity) should be revised. By the old definition, "969 million people suffered from absolute poverty in 2004, a drop of over 270 million since 1990"; much of the drop is due to a decrease in poverty in China.

Researchers have pondered various ways of updating the 15-year-old threshold. For instance, "$1.08 in 1993 was worth about $1.45 in 2005 money." Researchers apparently decided that number was too high. The whole debate comes across as a pure academic exercise, because I can't imagine anyone living on even $2 a day in the States three years ago ($60 a month) and thinking of themselves as not poor, but you need to measure absolute poverty somehow. Once the researchers decided they did not like the $1.45 threshold established by translating the 1993 value into today's terms based on inflation, they "gather[ed] 75 national poverty lines, [...] pick[ed] the 15 lowest [... and reached] a new international poverty line of $1.25 a day." The article makes interesting comments on absolute vs relative poverty, which I won't quote here to keep this post short. With this new threshold, it turns out that "about 130 million [Chinese people] more than previously thought" were poor. The researchers, however, claim that their new calculations still make the number of poor people in China decrease between 1990 and 2004, because "[China's] trajectory is much the same." (Hmm...) The Economist concludes that "it matters little where a poverty line is drawn" and that anyway a $1.25 threshold will never be adopted because it is not as catchy as "a dollar a day". Instead, "a better option might be to reset the poverty line at $1 in 2005." So much for rigorous definitions.