The elasticity of taxable income in a developing economy: Evidence from Zambia’s 2017 Pay-As-You-Earn [PAYE] reform

Abstract

This study estimates the elasticity of taxable income (ETI) at the top of the income distribution in a low-income African economy and translates the estimates into welfare and optimal-rate implications for Pay-As-You-Earn (PAYE) policy. The analysis is guided by the optimal taxation sufficient-statistic framework, in which the ETI, the Pareto parameter of the upper-income tail, and the social welfare weight on top earners jointly determine welfare-consistent tax rates. A difference-in-differences design is applied to the Zambia Revenue Authority administrative panel data for 2014 –2019, exploiting the 2017 increase in the top marginal PAYE rate from 35 percent to 37.5 percent. The sample contains 3,646,850 person-year observations for 372,223 private formal sector employees. The baseline ETI is 0.27 and is robust to alternative specifications and sample restrictions. Responsiveness varies sharply by age and gender, from 0.134 for young men to 1.182 for young women. At the prevailing top rate, the baseline marginal excess burden is 0.162, the marginal welfare cost is 0.193, and the marginal cost of public funds is 1.19. With a Pareto parameter of 2.0, the full-sample revenue-maximising top rate is found to be 48.1 percent with a social welfare weight of 0.5. This exceeds the prevailing 37 percent, although the optimal rate for young women lies below the current rate. The results show that Zambia retains modest fiscal space at the top of the private formal PAYE distribution, but the aggregate masks costly subgroup responses. PAYE reform should therefore combine cautious, evidence-led rate appraisal with regular monitoring of demographic heterogeneity and targeted measures that protect formal employment among highly responsive groups.