Growth redistribution decomposition
Table 1 decomposes the change in headcount poverty in terms of the effect of growth and of changes in inequality. In Mauritius, the headcount movement can be explained in large part by a growth effect. 1.9 percentage point of the 1.7 percentage point fall from 5.8% to 4.1% can indeed be attributed to a growth effect. The bottom half of Mauritius' population lost around 1% of total national consumption to the top half. This slight worsening of inequality was not, however, substantially detrimental to the effect of growth on poverty reduction. The fall in poverty between 2001 and 2006 would have been from 5.8% to 3.9% if inequality had remained unchanged. Hence, Mauritius's development in the early 2000s did roughly succeed in reducing poverty through growth and at a relatively modest poverty cost through an increase in inequality. In South Africa, the total change in poverty between 1995 and 2005 is neither numerically nor statistically significant. The effects of growth and redistribution on poverty have, however, been important (see Table 1). Growth reduced poverty by around nine percentage points. The increase in inequality increased poverty by roughly the same numerical value. Hence, both the growth and the redistribution effects have almost exactly cancelled each other. The headcount in 2005 would have been roughly 10 percentage points lower (from 42% to 33%) had it not been of the increase in inequality. Hence, poverty has changed little in South Africa between 1995 and 2005, but it is not to say that little else has changed: average consumption has increased substantially, but inequality has also risen importantly, and this has cancelled all of the positive poverty effects of growth.