Ela Elliyana Abdul Rahman
Indonesia has increased access to financial products at a rapid pace. However, there is still some question about whether households understand their use of such products which can have both direct consumer protection and financial stability implications. This paper utilizes the 2025 National Survey of Financial Literacy and Inclusion (SNLIK) that was conducted by the Financial Services Authority (OJK), with the assistance of Statistics Indonesia (BPS). Using this data we will construct a segment-level analysis of the financial inclusion/literacy gap (the number of percentage points the composite inclusion index is greater than the composite literacy index) across 26 segements defined by socio-demographic characteristics such as domicile, gender, age range, level of education, gross income, and occupation. The results show an overall gap of 14.05 percentage points at the national level (80.51% inclusion rate versus 66.46% literacy rate). The gap is not homogenous (ranging between 8.47 and 25.89 percentage points among the segments analyzed), and it is significantly negatively correlated with the level of literacy (Pearson r = -0.45, p = 0.022; OLS slope = -0.14, 95% CI [-0.26, -0.02]; R2 = 0.20): thus, those households with lower levels of financial literacy tend to be found in those households with the largest gaps. The lowest income (< Rp500k: 22.50 pp), the youngest (15-17 years old: 22.32 pp), and the non-employed (15.46 pp) segments all have below median levels of literacy and above median gaps indicating they would be good candidates for financial education. The national literacy shortfall is 2.89 percentage points below the targeted rate of 69.35 percent in the RPJMN, and there is a gap of 12.49 percentage points with respect to the target rate of 93.0 percent for inclusion. Based on our findings, we assert that considering the gap between inclusion and literacy rather than the two indexes separately provides a more relevant framework for developing common financial capability interventions. The study is purely descriptive and uses published aggregate data; therefore, we present the limitations associated with inferential statistics in this regard.