Neryvia Pillay Bell and Harry Kemp on the elasticity of taxable income, new data and estimates for South Africa.

On 17 November the SA-TIED programme will host an online seminar on the elasticity of taxable income, new data and estimates for South Africa. The webinar will be given jointly by Neryvia Pillay Bell and Harry Kemp under the work stream on Public revenue mobilization for inclusive development.

Abstract: Taxpayer responsiveness to taxation: Evidence from bunching at kink points of the South African income tax schedule | Neryvia Pillay Bell

The author applies the bunching methodology to South African administrative tax data over the period from 2011 to 2017 to investigate the responsiveness of individual taxpayers to changes in marginal personal income tax rates. She finds significant evidence of bunching among the self-employed but no evidence of bunching among wage earners. Among the self-employed, bunching is greatest at the highest kink in the income tax schedule and smallest at the lowest kink. Female self-employed exhibit greater bunching behaviour than male self-employed, and responsiveness appears to decrease with age. The responsiveness of the self-employed appears to be due to tax avoidance by shim fting income into future periods through retirement fund deductions, as well as a real labour supply response. Despite the significant excess bunching observed, the implied elasticities of taxable income—under the assumption of a uniform heterogeneity distribution around the kink—are not very large. 

Abstract: The elasticity of taxable income - New data and estimates for South Africa | with Johannes Hermanus Kemp

The elasticity of taxable income is a key tax policy parameter that plays an important role in the formulation of tax and transfer policy. This paper extends work by Kemp (2019) by using a new panel of individual tax returns and the phenomenon of ‘bracket creep’ to produce updated estimates of the elasticity of taxable income for South Africa. Whereas the previous work focused on assessed taxpayers (i.e. individual assessed tax returns), the current study uses a newly constructed panel for the period 2010–16 that includes information captured on both individual income tax return (ITR12) and employee tax certificate (IRP5) forms. The elasticity of taxable income is estimated at around 0.4, somewhat higher than the estimate presented previously. The elasticity for broad income is estimated at around 0.2, similar in size to the earlier study. As in that earlier study, it was found that behavioural responses are concentrated in higher-income groups, as suggested by the higher elasticity estimates for individuals in the top two income tax brackets and the top 10 per cent of income earners. An important finding is that the ETI for high-income earners increases notably in the latter half of the sample, mainly due to increased itemized deductions in the face of a rising tax burden due to insufficient bracket adjustments and rising marginal rates.

For more details or to register contact chiweza@wider.unu.edu