Abstract

Authors: Anggraini Widjanarti, Akhmad Zacky Nugraha, Cendani Tri Meilani, Mohammad Khoyrul Hidayat, Farhan Hafizh, Insan Istafada, Sri Pujilestari

The provision of high-quality statistical data and information is a fundamental prerequisite for central banks to support credible and accountable policy formulation. As demands for consistency, accuracy, and cross-publication coherence intensify, strengthening statistical governance, particularly in the dissemination function, has become a strategic agenda inseparable from ongoing digitalization efforts. In this context, digital transformation in statistical management extends beyond technology adoption, it requires standardized process, harmonised and integrated metadata, and robust quality assurance (QA) mechanisms to ensure reliability and consistency across publications.
External Debt Statistics are a core external-sector indicator used by policymakers, international institutions, and market participants to assess Indonesia’s external position and related risks. Inconsistencies in external debt figures across dissemination outputs can undermine credibility and adversely affect perceptions of country risk. At Bank Indonesia, pre-dissemination QA for external debt publications has traditionally relied on spreadsheet-based, macro-supported manual workflows. While adequate for basic internal controls, this approach becomes increasingly constrained as data structures grow more complex, dissemination formats diverge, and expectations for traceability, reproducibility, and auditability rise, thereby increasing operational and weakening dissemination governance.
In response, this study proposes metadata-driven QA tools as an integral component of Bank Indonesia’s digital transformation for external debt dissemination. The approach adopts the Statistical Data and Metadata Exchange (SDMX) standard as the foundation for data and metadata structures, and leverages the General Statistical Business Process Model (GSBPM) to position QA as a governed “final quality gate” within the statistical business processes. Automated validation is implemented through a Python-based rules engine within the Omni Intelligence Platform, enabling systematic checks across multiple dissemination outputs, while strengthening traceability, reusability, and efficiency of QA processes. The findings indicate that SDMX-based QA tools reduces reliance on manual processes, improves the detection and resolution of inconsistencies prior to release, and strengthen transparency and accountability in statistical governance. The proposed framework provides a scalable and adaptive foundation, aligned with international standards and best practices, to support the digitalization agenda of central bank statistics.