Internationally, it has become widely accepted to use the zero-rating of value-added tax (VAT) to
mitigate the regressivity of the tax. From an economic perspective, it remains contestable whether
zero-rating is the most cost-effective way of targeting the poor as richer income groups receive
(in absolute terms) a bigger benefit than the poor. Hence, there is a need to reconsider VAT zerorating.

This paper addresses some topical issues on VAT zero-rating in South Africa. We first ask whether
(conceptually) zero-rating should be a consideration within the context of tax theory literature,
and then quantify the impact on the poor if zero-rating were to be removed, as well as the tax
revenue implications thereof. We mainly use micro data sourced from the Income and Expenditure
Survey (IES) 2010/11. A crucial issue we address is the effectiveness of VAT zero-rating
compared to existing social transfer programmes in aiding the poor - our findings show that VAT
zero-rating is not as cost effective in targeting the poor.