Identifying Regional Heterogeneity and Priority Factors for Infrastructure in Indonesia Using FIMIX-PLS and Importance-Performance Map Analysis

Infrastructure development is important for reducing regional disparities, but Indonesian provinces differ in economic conditions, human resources, and regional characteristics. This study examines the relationships between the Economy, Human Resources, and Regional Characteristics and Infrastructure in the 38 provinces of Indonesia, based on secondary data from Statistics Indonesia (BPS) for 2024 and 2025. To examine the structural relationships, Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed, and to investigate unobserved heterogeneity, Finite Mixture PLS (FIMIX-PLS) was used together with ImportancePerformance Map Analysis (IPMA) to determine the priorities within the segments. The model accounted for 82.3% of the variance in Infrastructure (R2 = 0.823). Yet none of the structural paths were found to be statistically significant at the 5% level: for Economy (t = 1.195, p = 0.232), for Human Resources (t = 0.895, p = 0.371), and for Regional Characteristics (t = 1.928, p = 0.054). FIMIX-PLS identified a two-segment solution, with the final partition assigning 27 provinces to Segment 1 and 11 provinces to Segment 2. IPMA showed different patterns across segments: Regional Characteristics had the highest importance in Segment 1, while Economy ranked highest in Segment 2. Regional Characteristics also had the lowest performance in both segments. As this construct represents disaster exposure rather than development quality, the results highlight the need to consider disaster risks and economic differences when setting infrastructure priorities. However, the findings remain exploratory because none of the overall structural paths were statistically significant, and the segmentation is exploratory.