Determinants of the growth semi-elasticity of poverty reduction

In this paper we examine the mathematical relationship between growth and distributional change on absolute (i.e. percentage point) changes in FGT poverty measures assuming a log-normal income distribution, which we argue to be a conceptually superior and more policy-relevant measure than the much used regular growth elasticity of poverty reduction. We also test the empirical relationship of these semi-elasticities of growth and distributional change on poverty and find them to explain actual changes in poverty very well (in fact, much better than a related study by Bourguignon (2003) that studied the growth elasticity of poverty reduction). This is particularly the case when poverty depth and severity is considered. Using our results helps in interpreting past performance in poverty reduction and will allow a rapid and quite reliable prediction of the impact of growth and distributional change on (absolute) poverty reduction across countries, taking heterogeneity in country circumstances into account.