OpenAudit is an academe-civil society initiative that brings new AI and deep learning tools to advance leading-edge governance research along with reform advocacy to strengthen accountability, effectiveness, and civic engagement in the use of public funds. Beginning with the Philippines, where the national Commission on Audit (COA) has published more than 30,000 annual audit reports since 1998—possibly the longest and most comprehensive series of annual audit reports in the developing world— OpenAudit aims to transform these documents into a searchable and analyzable public resource. By processing these previously inaccessible audit reports using agentic AI and advanced computational methods, we reveal patterns of fiscal management across all Philippine Local Government Units through normalized disallowance ratios.
This platform provides an beta version of what will be a living public-good infrastructure for evidence-based research and policy reform advocacy that can be expanded to other developing democracies with publicly-available and digitized audit reports. We present novel "red flag" indicators of fiscal anomalies based on judgements by auditors made in these reports concerning the (a) disallowances (disapproval in audit of a transaction, either in whole or in part); (b) suspensions (a temporary disallowance, entailing transactions or accounts which appear illegal/improper/irregular unless satisfactorily explained or justified); and (c) charges (irregularities concerning non-collection, undercollection, non-remittance, or underremittance of collections). We provide data for these over the period 2016-2022 in ratio format as well as in terms of per-capita levels.
We combine three authoritative government datasets:
For each LGU, we calculate several metrics both annually and in aggregate over the period 2016-2022:
Our comprehensive dataset represents the most complete public accountability platform for the Philippines, covering virtually all LGUs with available audit data. The platform updates as new COA reports are released, ensuring citizens have access to the latest information.
The aggregated data reveals broader patterns that merit investigation:
OpenAudit is built on principles of transparency and reproducibility:
We acknowledge several limitations in the current dataset:
Despite these limitations, the disallowance ratio provides a valuable starting point for understanding fiscal management patterns and identifying areas that warrant closer examination.
OpenAudit has been honored MIT Open Data Prize in recognition of its contribution to making public sector information usable for research and government accountability. This initiative represents a collaboration between academic institutions and civil society organizations working to strengthen fiscal transparency and accountability in the Philippines.