Closed but Not Recovered: Size, Time, and the Limits of Audit Follow-Up in Indonesia

Follow-up of audit recommendations is usually measured by how many recommendations are closed. That measure ignores whether the money at stake is actually recovered. This study separates the two. Data come from the Supreme Audit Board's recommendation follow-up monitoring appendix for local governments, status as of Semester I 2024, covering 518 entities that reconcile on both counts and values. Across the 2018 to 2022 cohorts, local governments closed 72.13 percent of recommendations on average but recovered only 53.1 percent of the 11.34 trillion rupiah at stake. Recovery falls monotonically with the size of the recommendation, from 99.06 percent in the smallest decile to 55.71 percent in the largest, and to 39.65 percent on an aggregate basis. Closure falls by only 9.0 points across the same deciles. With cohort and entity fixed effects, a tenfold larger recommendation portfolio is associated with recovery lower by 18.7 percentage points. Resolution also stalls: closure rises by 5.12 points a year up to age eight and by 0.09 points a year thereafter, leaving 49,639 recommendations issued between 2005 and 2017 still open. The follow-up system settles small and recent matters and leaves large and old ones unresolved.