When I saw this World Bank stats in the Economist I initially thought "hey, look at this, the number of extreme poor is declining in absolute numbers and relative of the world population. This is good news, we are on the right track".
Wrong! I recalled some discussion from university and tried to re-think "what really is poverty? Is it absolute, i.e. less than 1$ a day OR is it relative to other people?". Think about it... Sticking to the World Bank definition of absolute poverty of less than 1$ a day is simple and convenient but gives the wrong picture. Even if the poor people might not get poorer in absolute income, they become poorer compared to the rest of the world.
An analogy: If we would just measure how long people life, we could say "great, the people in Mali are still dying at 40 years and not earlier. So at least they are not getting poorer". BUT if you then see that people in developed countries live longer and longer (80s), it becomes obviously absurd.
That said, I believe we can't measure poverty disconnected. Poverty is relative to other people and means you can't afford basic things that your peers can (by the way it's seems similar with happiness). I love simplicity but the World Bank stats have outlived their purpose now because they paint a wrong picture. The gap between rich and poor is growing which means inequality and poverty is growing. Poverty ultimately has to be defined in relative terms, a claim supported by others in the field such as Brown/Ainley 2005 "Intl. Relations". What do you think?