Reliable information on changes and the level of economic ac- tivity are central for macroeconomic analysis, but preliminary output data available when policy decisions are made are typically revised as more and better information become available. There is a large literature on devel- oped countries documenting such revisions and discussing its implications, but evidence from developing, notably low-income economies, is scarce. The objective of this paper is therefore to analyze the nature, causes and policy implications of output data revisions, i.e., the discrepancy between forecasts, nowcasts as well as `early' backcasts of real GDP growth and the level of nominal GDP in low-income countries. First, using data from the IMF's World Economic Outlook, we show that output data revisions in low-income countries are, on average, larger than in other countries, and that they are much more optimistic. Second, we analyze the determinants of output data revisions and show that economic and technical factors drive revisions, rather than factors related to the political economy of the IMF itself as commonly assumed.
Ley, E.; Misch, F. Output Data Revisions in Low-Income Countries. Presented at Joint RES-SPR Conference on &#8220;Macroeconomic Challenges FacingLow-Income Countries,&#8221; Washington, DC, January 30&#8211;31, 2014. International Monetary Fund, Washington DC, USA (2014) 25 pp.