Value Added Tax Structure Reforms and Revenue Performance in Kenya: Evidence From National Monthly Time-Series Data (2016–2025)
DOI:
https://doi.org/10.51317/nfxryr15Keywords:
Dynamic regression, Kenya, revenue performance, tax structure reforms, Value Added TaxAbstract
This study examined the effects of three Value Added Tax (VAT) structure reforms, namely the 2018 introduction of an 8 per cent rate on petroleum products, the 2019 imported-services reform and the 2020 temporary reduction of the standard rate from 16 per cent to 14 per cent, on VAT revenue performance in Kenya. Despite repeated VAT amendments, revenue relative to GDP has shown limited improvement, raising questions about the revenue effects of specific reforms. Anchored on Optimal Tax and Laffer Curve theories, the study used national monthly VAT collection data from the Kenya Revenue Authority covering January 2016 to December 2025. The monthly change in logged VAT revenue was analysed using an autoregressive ordinary least squares model with differenced reform indicators and monthly dummies, applying Newey–West standard errors. The model explained 59.2 per cent of the variation in monthly revenue changes, while the reforms were jointly significant (F(3, 100) = 8.71, p < 0.001). The imported-services reform increased VAT revenue (β = 0.0391, p = 0.020), whereas the petroleum rate reduction (β = −0.0554, p = 0.002) and the temporary standard-rate reduction (β = −0.0781, p = 0.006) reduced collections. Alternative specifications revealed no significant sustained growth effects, supporting the interpretation that the reforms produced level shifts in revenue. The findings indicate that base-broadening measures are associated with increased VAT revenue, while rate reductions are associated with lower collections. The study recommends prioritising VAT base-broadening measures and conducting ex-ante revenue-cost assessments before implementing rate reductions.
Downloads
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Anne Chepkemboi Rotich, Patrick Kibati, Symon Kiprop

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.