Tax digitization and revenue performance: Pre and post analysis

Abstract

The effect of digitization on two distinct outcomes is examined using quarterly time series data for the period 2010Q1 to 2025Q4 provided in the FIRS Annual Reports, the CBN Statistical Bulletin, the NBS, and OECD Revenue Statistics in Africa. An ex post facto design is adopted, with the tax digitization phenomenon represented by a binary dummy, which distinguishes the pre-digitization time period of 2010Q1–2015Q4 from the post-digitization time period of 2016Q1–2025Q4. This is because the unit root tests performed using ADF confirm that all the series are stationary at level (I(0)); therefore, OLS regression is performed in the estimation process. Both models give different results: Digitization has a positive and statistically significant impact on tax revenues, and the post-digitization period corresponds to a 43.6 percent increase over pre-digitization baseline levels. The opposite is true of non-tax revenue. It has only a slight and negative impact, and digital tax administration reforms have hardly had any impact on non-tax revenues at all. Interestingly, the exchange rate, rather than inflation, emerges as the most stable indicator in both models, a characteristic that is typically observed in Nigeria because it relies heavily on oil- and FX-dependent fiscal inflows. The essential variation in revenue in either category is not due to inflation. Combined, the results imply that tax digitization is effective and supplies the revenue stream that it is designed to capture. Thus, there is a need for institutional reform pertaining to non-tax revenue administration, as well as revenue diversification and the provision of additional digital tools.