This work stream has two main aims. First, to arrange access to individual and firm-level taxpayer data at South African Revenue Service (SARS) for research use. The goal is to create well documented longitudinal datasets with key variables that can be augmented with research topic-specific variables. This serves as a basis for not only tax research but also for other research strands that require micro-level administrative data. Second, the work stream will produce several studies on the incidence of taxes, determinants of non-compliance and under-filing, and the effectiveness of tax incentives in South Africa. In addition to administrative data, survey data will be utilized in conjunction with tax/benefit microsimulation. The gains from a combined analysis of microsimulation and macro models will also be explored.
This work stream builds on the previous research programme in South Africa, which made considerable progress in the area of tax research. The most important advance was the access to high-level administrative data, which enabled a deeper and wider analysis of research on taxation. In addition to its value for academic research and the users of the new findings among policy makers, the use of administrative data also serves many of the data needs of the state. Policies that promote development need to be built on reliable facts and administrative data offers an obvious and inexpensive way to do this.