Most sub-Saharan African countries are characterized by low tax compliance and low tax productivity. This study tests the effects of a tax lottery under alternative reward designs on compliance as an alternative policy option for addressing the problem of low tax receipts in Tanzania. The lab experiment involved the purchase of goods with a sample of 313 undergraduate students recruited from courses with and without tax specialization. The experiment participants were randomly assigned in control and treatment groups and thereafter assigned random endowment incomes. In the treatment groups two treatments were administered: a lottery of high probability and low rewards, and a lottery of low probability and high rewards, where eligibility for the lottery was restricted to those who paid VAT on the purchase (which would be cheaper otherwise). The results of the experiment revealed a lottery of high reward has a higher impact on compliance and revenue. Our estimates show that the net revenue effects of these lotteries differ by 27 percent. Hence, the design of a tax lottery is important. Further, tax lotteries have the potential to improve taxpayer compliance and increasing revenue collection.
Download the paper in PDF format
Authors
Cyril Chimilila, Remidius Ruhinduka and Vincent Leyaro
View all CREDIT discussion papers
Sir Clive Granger BuildingUniversity of Nottingham University Park Nottingham, NG7 2RD
Enquiries: hilary.hughes@nottingham.ac.uk