Effectiveness of System Automation on Tax Governance in Uasin Gishu County, Kenya

Abstract

In Kenya, county governments have been making substantial investments in digital revenue management systems to help them increase the efforts of own-source revenue collection, but persistent variation in revenue performance indicates that technological adoption alone might not automatically translate into improved tax governance. Even though the Integrated Revenue Management Information System (IRMIS) was implemented in Uasin Gishu County, the issues with the integrity of reconciliation, compliance behavior, and the efficiency of the collection remain obvious. This paper therefore examined how automation of systems affects tax governance in Uasin Gishu County, Kenya and the moderating role of tax education in this relationship. In particular, the researchers evaluated the effects of online process automation, transaction automation, and reconciliation and control automation to the outcomes of tax governance, determined by the efficiency of revenue collection and tax compliance. The research was anchored on Digital Transformation Theory, which argues that standardized system execution promotes transparency and less discretion, and Process Automation Theory, which asserts that standardized system execution ensures transparency and less discretion. A cross-sectional survey design was used that was explanatory in nature and was conducted within the last twelve months. A stratified random sample of 200 out of a target population of 415 respondents was chosen and 181 out of 200 respondents were able to complete the questionnaires and submit them back, which is a 90 percent response rate. The analysis of data was based on the descriptive statistics, Pearson correlation analysis, and multiple linear regression estimated using Ordinary Least Squares with diagnostic tests to ensure robustness. The results of correlation indicated that there were positive and statistically significant relationships between online process automation (r = 0.405, p < 0.01), transaction automation (r = 0.354, p < 0.01), reconciliation and control automation (r = 0.649, p < 0.01), tax education (r = 0.658, p < 0.01), and tax governance. Automation and education were joint predictors of tax governance, with 54.1 percent of the variance in tax governance explained (R 2 = 0.541, F = 51.915, p < 0.001). Diagnostic reports confirmed that the data were normal, no multicollinearity (VIF range = 1.2951.484), and homoscedastic residues. The researchers have concluded that the main determinants of tax governance in devolved revenue systems are reconciliation and control automation and tax education. It suggests that internal automated controls should be increased, system integration and reconciliation features should be enhanced, and continuous taxpayer education programs should be institutionalized in order to maximize governance results. Future studies need to embrace longitudinal and comparative multi-county design, mix-methodological studies, and the emergent technologies including artificial intelligence-based revenue analytics to enhance the current understanding of the digital tax governance within the devolved system of Kenya.