Unwarranted or Warranted? A Jurimetric Analysis of Custodial-Sentence Disparity in Indonesian Corruption Cases after the 2020 Supreme Court Sentencing Guideline

Abstract

Sentencing disparity has long been diagnosed as a structural weakness of Indonesian anti-corruption enforcement, yet most scholarship remains doctrinal or confined to single-court case studies. This article applies a jurimetric (quantitative-empirical) approach to measure the magnitude, structure, and predictors of custodial-sentence disparity in corruption (tindak pidana korupsi) cases four years after the Supreme Court issued sentencing guidelines through Regulation No. 1 of 2020. An original dataset of 47 first-instance convictions for state-financial-loss corruption (Articles 2 and 3) decided in 2024 was hand-coded from the Supreme Court Decision Directory across eight Special Criminal Courts spanning Sumatra, Java, Kalimantan, Sulawesi, and Nusa Tenggara. Custodial term, statutory fine, and court-ordered restitution were analysed using dispersion statistics (coefficient of variation, Gini coefficient), the Kruskal–Wallis and Brown–Forsythe tests, rank correlation, and ordinary-least-squares regression. Custodial sentences ranged from 5 to 180 months (mean 34.7; median 24) with a coefficient of variation of 0.91 and a Gini coefficient of 0.42; court means differed by a factor of 3.5. The magnitude of state financial loss was significantly but only moderately associated with sentence length (Spearman rho = 0.53; ordinary-least-squares R-squared = 0.26), leaving roughly three-quarters of variance unexplained. Among defendants receiving the identical modal fine, sentences still varied 7.2-fold. On the balance of evidence, a substantial share of the disparity is plausibly unwarranted, indicating that the 2020 guideline has not yet delivered structured proportionality. The study supplies a replicable coding protocol and open dataset for large-N replication and grounds a concrete de lege ferenda reform agenda.