Macroeconomic data have been shown to vary substantially between sources, especially so for low-income countries. While the impact of data revisions on inference is well documented for cross-country studies, there is no systematic analysis of the robustness of results obtained from time series...