Abstract

This paper presents preliminary findings from dissertation research examining how international frameworks for governance of statistical data interact with on-the-ground realities in Latin American and Caribbean (LAC). Statistical data governance often assumes a resource abundance and institutional stability absent in LAC contexts, thus creating persistent implementation gaps between design and practice.

Using socio-technical interaction networks (STIN) analysis, this research maps governance baselines against contextual baselines; including country adherence patterns, statistical capacity indicators, and institutional arrangements; to identify where the IMF Data Standards Initiatives (E-GDDS, SDDS, and SDDS Plus) succeed and struggle across 32 countries in the region.

Surface level analysis already reveals significant implementation variation across the three tiered governance levels, from E-GDDS for countries looking to develop their governance to SDDS Plus for those with robust existing systems. For example, Paraguay required 23 years to progress from E-GDDS commitment to SDDS observance, while Uruguay and Costa Rica achieved immediate compliance. Only Brazil and Chile have reached SDDS Plus, the highest level of observance, each requiring approximately eight years to do so. Twenty-one countries remain at E-GDDS despite years of commitment.

These divergent pathways demonstrate that contextual fit- the match between governance requirements and local values, needs, skills, and resources- determines implementation success. For example, Paraguay's advancement coincided with enacting its first statistics law in 80 years, establishing a modern national institute, and cataloging 79 agencies across its statistical system, suggesting governance frameworks must sequence implementation around institutional prerequisites rather than assume they exist.

This research shares further preliminary analysis of this space. It also contributes to data interoperability by developing actionable guidance for national statistical offices and multilateral institutions working to build connected, trustworthy data ecosystems that respond to actual constraints rather than ideal conditions.