International Monetary Fund. Independent Evaluation Office
This report examines whether the IMF has effectively leveraged an important asset: data. It finds that in general, the IMF has been able to rely on a large amount of data of acceptable quality, and that data provision from member countries has improved markedly over time. Nonetheless, problems with data or data practices have, at times, adversely affected the IMF’s surveillance and lending activities. The roots of data problems are diverse, ranging from problems due to member countries’ capacity constraints or reluctance to share sensitive data to internal issues such as lack of appropriate staff incentives, institutional rigidities, and long-standing work practices. Efforts to tackle these problems are piecemeal, the report finds, without a clear comprehensive strategy that recognizes data as an institutional strategic asset, not just a consumption good for economists. The report makes a number of recommendations that could promote greater progress in this regard.
Since the IMF launched the data standards initiatives a decade ago, 145 of its 184 member countries have participated. This 80 percent participation rate reaffirms the importance countries place on data transparency in the globalized economy, which the initiatives promote. The wide participation can be attributed to the consultative process that has allowed for the development of a coherent program that takes account of countries' capabilities, delineates clear responsibilities between the IMF and participating countries, and establishes effective monitoring procedures to ensure the credibility of the standards for policymakers, capital markets, and the general public. The approach has also provided checks and balances and fostered accountability. The initiatives may provide insights for the promotion of similar international standards.
This paper finds that reforms introduced by the IMF to promote transparency have created more informed markets and reduced borrowing costs for those emerging market countries that volunteered for them. Using a quarterly panel estimation with fixed country effects, we find that sovereign spreads fall following the adoption of three different transparency reforms. The effects are economically important, especially for those countries with low initial transparency. We use two-stage least squares to address any endogeneity in the timing of reforms exploiting internal IMF timetables that are unrelated to country events. Next, using a panel GARCH specification, we show that spreads move more than normal in the days immediately following publication of IMF country documents.