Bayesian Inference for Parametric Growth Incidence Curves

The growth incidence curve of Ravallion and Chen (2003) is based on the quantile function. Its distribution-free estimator behaves erratically with usual sample sizes leading to problems in tails. authors propose a series parametric models Bayesian framework. A first solution consists modeling underlying income distribution using simple densities for which function has closed analytical form. This extended by considering mixture model distribution. However, this case, semi-explicit be evaluated numerically. last adjusting directly functional form Lorenz deriving its first-order derivative find corresponding compare these Monte Carlo simulations UK data from Family Expenditure Survey. devote particular attention analysis subgroups.