Trade-offs in the Design of Simplified Tax Regimes: Evidence from Kenya

Governments across Africa rely on simplified tax regimes (STRs) to tax millions of small businesses, yet these regimes raise very little revenue and vary widely in design. We combine more than ten years of administrative data from Kenya’s STR, two original surveys, and two field experiments, both covering over 10,000 taxpayers, to examine the de facto performance of three core design features: exemption thresholds, set fees versus turnover-based liabilities, and the level of the tax rate. We find that around 70 per cent of taxpayers in Kenya’s STR declare turnover below the exemption threshold, meaning that they legally owe nothing, yet the filing system applies the tax rate to every declared amount, and payments from this group account for around a third of the regime’s revenue. Field experiments that aimed to increase simplicity by making the exemption threshold salient at the time of filing or suggesting customised set fee amounts raised revenue, but almost entirely from taxpayers with the smallest baseline payments, making the de facto distribution of payments more regressive. Event studies show declared turnover fell by roughly a third after a tax rate increase and rose by roughly a quarter after a subsequent tax cut, partially offsetting the revenue implications of both reforms. We illustrate that these findings are likely to be relevant across countries by showing the high prevalence of these STR design features in Africa. We conclude that taxing small businesses through STRs involves unavoidable trade-offs between revenue, simplicity, and equity, and this raises the question of whether these taxpayers should be a revenue priority at all.