UTILIZATION OF WHISTLEBLOWING SYSTEM, AUDIT FINDINGS AND OPINIONS IN REVEALING CORRUPTION LEVELS TO SUPPORT SUSTAINABLE GOVERNANCE IN PROVINCES IN INDONESIA

Siti Roma Harahap (Al Azhar University)
Aria Masdiana Pasaribu (Al Azhar University)
Windy Aginta (Al Azhar University)

Multidiciplinary Output Research for Actual and International Issue (Morfai Journal)
Vol. 6 No. 6 (2026): Multidiciplinary Output Research For Actual and International Issue

This study aims to analyze the utilization of the whistleblowing system (X1), audit findings (X_2), and audit opinions (X3) in revealing corruption levels (Y) to support sustainable governance across provinces in Indonesia. Employing a quantitative panel data approach comprising 68 observations, the analytical framework incorporates descriptive statistics, model selection procedures (Chow, Hausman, and Lagrange Multiplier tests), classical assumption testing, and panel regression analysis executed via EViews. The diagnostic evaluation selected the Fixed Effect Model (FEM) as the optimal estimation model. Furthermore, classical assumption tests confirmed that the empirical model satisfies all statistical requirements, featuring normally distributed residuals (p = 0.762367), no multicollinearity issues (correlations < 0.80), and an absence of heteroscedasticity (p = 0.187250). The regression results reveal that the utilization of the whistleblowing system (X1) has a statistically significant negative impact on corruption levels (Y) (beta = -3.518266, p = 0.0181). Conversely, audit findings (X2) (beta = 25.02458, p = 0.1204) and audit opinions (X3) (beta = 67.93004, p = 0.5489) exhibit positive relationships with corruption levels, though their individual effects are not statistically significant at the 5% level. Simultaneously, all independent variables jointly exert a significant combined effect on corruption levels (F = 3.129075, p = 0.031686). The Adjusted R2 value of 0.087035 indicates that 8.70% of the variance in corruption levels is explained by the model, while 91.30% is driven by external factors beyond the scope of this study.