The $2/day and $3/day thresholds are exactly twice and three times the $1/day line.

That's my favorite sentence from a new academic paper by economists Maxim Pinkovskiy and Xavier Sala‐i‐Martin.  They are extremely smart economists, but if they write this badly... should we have much confidence in the content of their paper?  The content is this:  Poverty in Africa fell steadily for a decade (1995-2006) and the continent is on track to achieve the Millennium Development Goal of halving poverty by 2015.  This may not be surprising; numerous commentators have noted that the last decade has seen Africa's best growth since the 1960s.  The surprise is the authors' evidence that the growth is broad-based and general.  That is, instead of the growth all going to corrupt leaders and their ethnic minions (the conventional wisdom amongst academics and ordinary citizens of African countries) or to oil and gold-rich countries at the expense of landlocked, resource poor countries like Burkina Faso, the reduction in poverty seems to be very general.  Here is the picture of the distribution of income for the continent... notice the distribution shifts slightly to the left from 1970-1990, but then experiences a strong shift to the right.