Measuring Uncertainty in PPP Aggregation: A Study of GEKS and Weighted CPD Approaches Using African Data

Abstract

This study investigates the estimation of Purchasing Power Parities (PPPs) above the basic heading (BH) level conducted across 47 African countries with Nigeria as the numeraire. Using price and expenditure data aggregated from detailed item-level structures, the study compares five prominent PPP aggregation methods: GEKS-Fisher, GEKS-Tornqvist, Weighted GEKS-Fisher, Weighted GEKS-Tornqvist, and Weighted Country-Product-Dummy (WCPD) with an emphasis on the computation of their associated standard errors. The study reveals consistency among the methods with marginal differences in country-level PPP estimates. Notably, the standard errors of the estimates are also similar but Weighted GEKS-Tornqvist has the smallest Standard error value of 1.1676. This comparative evaluation provides empirical insight into the trade-offs and robustness of PPP aggregation approaches and guides methodological choices for regional and global ICP applications.