<![CDATA[This study aims to analyze the influence of population, agricultural sector, trade, Information and Communication, real estate, and education on regional tax revenue using the Partial Least Squares (PLS) method. This research employs a quantitative approach with a causal associative method, focusing on cause-and-effect relationships between variables. The data used are secondary time series data collected from official documents of relevant agencies, such as the Central Statistics Agency (BPS). Data collection techniques were carried out through documentation. The analysis results show that the population (X1) variable has a positive but insignificant effect on regional tax revenue (β = 19903.651; p = 0.400). The agricultural sector (X2) also shows a positive yet insignificant effect (β = 535.792; p = 0.858). In contrast, the trade sector (X3) has a negative effect, nearing significance (β = -13,683,591; p = 0.076). The transportation Information and Communication sector (X4) (β = 73,528,699; p = 0.251) and real estate sector (X5) (β = 61,295,163; p = 0.264) have positive but insignificant effects. Meanwhile, the education sector (X6) has a very weak and insignificant effect (β = 2,282,092; p = 0.956). The multicollinearity test results using Variance Inflation Factor (VIF) show that the education (VIF = 9.831) and real estate (VIF = 7.262) variables have relatively high multicollinearity. Nevertheless, the Breusch-Pagan test did not indicate significant heteroscedasticity (p = 0.155), meaning that the model is free from heteroscedasticity issues. The Durbin-Watson Test result of 2.107 indicates no autocorrelation in the model, ensuring the validity of the regression estimates. With an R-square value of 0.588, this model explains 58.8% of the variation in regional tax revenue. Although most variables are not significant, the findings suggest that the trade, transportation, and real estate sectors may potentially influence tax revenue. The use of the PLS method proves effective in addressing multicollinearity issues and provides more stable estimates compared to classical linear regression.]]>
