Gulnur Baidauletova

Type of the article: Research Article

Abstract
Electronic invoicing is an important component of digital tax administration that improves tax transparency and strengthens revenue administration. This study examines the association between electronic invoicing and regional tax revenues across Kazakhstan’s regions. The analysis uses regional panel data for 16 regions covering 2010–2024 and employs fixed-effects panel regression models together with k-means cluster analysis. The empirical dataset was obtained from the reports of the State Revenue Committee of the Ministry of Finance of the Republic of Kazakhstan and the Bureau of National Statistics of the Republic of Kazakhstan. The results show that electronic invoicing is not significantly associated with regional tax revenues in either the baseline model (β = 0.039; p = 0.538) or the specification with a time trend (β = –0.024; p = 0.690). However, the specification using the shifted e-Inv indicator yields a positive and statistically significant association (β = 0.302; p = 0.047). Because e-Inv is a nationwide step dummy, this result reflects temporal robustness rather than evidence of a delayed fiscal effect. Across all model specifications, GRP per capita remains the strongest determinant of regional tax revenues (β ≈ 0.825-1.102), highlighting the central role of regional economic capacity. The cluster analysis identifies three distinct groups of regions with different fiscal and digital characteristics, illustrating substantial regional heterogeneity in Kazakhstan. The findings suggest that the association between electronic invoicing and regional tax revenues depends on model specification and broader economic conditions. In contrast, regional economic capacity remains the most robust determinant of tax revenues.