Information, fiscal capacity, and tax enforcement: An experimental evaluation

With the recent decline in revenues from trade tariffs and official aid, developing countries have to rely increasingly on domestic taxes for revenue collection. One of the most important potential sources of domestic tax revenue is the value-added tax (VAT), a tax instrument that has been increasingly adopted by developing countries over the past decades. In theory, the VAT has two desirable properties: it is more efficient than alternative tax instruments, and it is self-enforcing because the two sides of each transaction (seller and buyer) have incentives to report truthfully. However, when enforcement capacity is limited, these two properties can break down, leading to economic distortions and tax revenue losses. Lack of enforcement capacity is common in Sub-Saharan African countries, such as Uganda.  The extent to which the lack of enforcement capacity reduces the desirability of VAT systems in such countries remains unclear, because of limited micro-level empirical evidence.

This project aims to assess the scope for the introduction of stronger compliance along the VAT chain, by implementing a tax enforcement experiment developed in close collaboration with the Uganda Revenue Authority.  As such, it utilises academic insights to improve the effectiveness of the URA’s enforcement capacity, and potentially reduce the VAT compliance gap. High-level URA officials have continually supported collaboration with the research team and have provided important inputs into the research process. This experiment builds on findings from a previous research report, which is co-authored with one member of the URA's Research and Policy Development Department. Furthermore, the URA has included this tax enforcement experiment as part of their National Audit Plan for the 2017-18 fiscal year, demonstrating the high degree of involvement and ownership of the project by the URA.

The intervention will involve sending out a series of notifications to randomly-selected firms and analysing the effects of these letters on subsequent compliance behaviour. Discrepancies in firms’ VAT declarations are detected using a cross-check methodology developed by the research team, made possible by comparing what the seller and the buyer report for a given transaction. The universe of firms subject to the experiment is selected based on past compliance behaviour identified through these cross-checks. The letters will inform treated firms that the URA is closely monitoring their VAT returns and can do systematic comparisons with their trading partners’ returns. The experiment will include several treatment arms, allowing to test different contents of the letters. For instance, some letters will list specific instances of discrepancies found between the firm and some of its clients or suppliers, to test whether this increases the perceived capacity of the URA and thus the impact of the notification.