We evaluate the effects of different equivalence scale parameter estimates on the
distribution of catastrophic health payments in South Africa. Our analysis makes
use of Xu et al.’s (2003) initial estimate, which underscores the World Health Organization’s
methodology (Xu, 2005). We also update it using more recent data
for one of the original countries included in Xu et al. (2003), South Africa. South
Africa is considered, because the data used in Xu et al. (2003) was collected before
the end of Apartheid, and the end of Aparthied has led to extensive social and
economic changes, which could have influenced equivalence. We extend the empirical
exercise by estimating a base-independent equivalence scale via semiparametric
methods. Using these equivalence scales, we examine their effect on the distribution
of catastrophic health payments using thresholds of 5%, 10% and 15%. The revised
estimates suggest that the initial equivalence estimates were overstated by as much
as 35%, such that poverty lines in the country were understated by as as much
as 17%. However, despite these large differences, the distribution of catastrophic
health expenditures were unaffected.