Hettiarachchi, K.K.S.Jayawickrama, J.M.A.2026-09-292026-09-292023-09-20Proceedings of the Peradeniya University International Research Sessions (iPURSE) – 2023, University of Peradeniya, P 2781391-4111https://ir.lib.pdn.ac.lk/handle/20.500.14444/8110Corporate income tax (CIT) is a key income tax in Sri Lanka and it generally accounts for about 58% of total income tax revenue and about 9.8% of total tax revenue. However, the ratio of CIT to GDP has fallen from 2.5% in the early 1980s to 1.3% in recent years. It is argued that factors related to structural variables, policies, compliance and administration of taxes have been contributing to the continuous loss of potential tax revenue. The objective of this study is to find the determinants of CIT losses (gap) which have been calculated using standard methods of tax loss computation. The CIT gap is defined as the difference between the potential taxes that companies should pay and the actual taxes collected by the revenue authorities. The potential CIT revenue depends on standard tax base and tax rate variables. However, actual tax revenue depends on the altered tax bases after various exemptions and deductions, tax rates after incentive schemes, acts of tax evasion and tax fraud, inefficiency in revenue collection authorities, corruption, etc. In this specification, we decompose CIT losses into five components: gap due to tax exemptions, deduction and qualifying payments, special rate concessions, under-reporting and collection failure. From the theoretical specification, we derive a testable equation of CIT loss, which presents CIT loss/gap as a function of standard tax variables such as tax base and marginal tax rate and deviation from such standard variables due to tax incentives and concessions and tax fraud and inefficiencies in collection. We use company-specific data of 1015 companies in a year available for seven assessment years from 2010/11 to 2016/17. The use of log dependent variable eliminates the companies with zero tax losses in the estimation. Panel Least Squares method and fixed effect model are used for the estimation based on Hausman test results. We find that income elasticity of CIT loss is high and it indicates 1% increase in tax base result in 1% increase in tax loss. Further, standard tax rate reduces CIT losses. However, special tax rates, income concessions, tax fraud and abuse of tax incentives have a positive effect on the corporate income tax losses.en-USCorporate income taxDeterminants of tax gapPanel estimationTax gapArticle