Value Added Tax (VAT) has been in effect in Nepal for over 27 years, yet empirical assessment is constrained by the availability of only 21 years of official time-series data (FY 2002–2022). This study investigates the determinants of VAT refund approval using a quantitative approach and multiple regression modeling. The results reveal that refund claims and rejections significantly influence approval rates, whereas total VAT collection has no direct impact. Although the model shows a high adjusted R² of 0.997, multicollinearity diagnostics confirmed VIF values were within acceptable limits, ensuring model validity and minimizing concerns of overfitting. The findings suggest that procedural clarity, timely documentation, and improved enforcement of self-assessment are more critical than sheer tax volume. With growing VAT contributions to GDP and revenue, persistent refund inefficiencies continue to erode taxpayer confidence. The study recommends simplifying procedures, adopting digital systems, and instituting administrative reforms to strengthen the credibility of the VAT refund mechanism in Nepal.
Aryal et al. (Fri,) studied this question.