'Measurement error of global production', by Peter A.G. van Bergeijk.

ISS working paper 632


This working paper discusses the need and possibility to report measurement error together with key (macroeconomic) statistics as shown by a case study of the real rate of growth of world GDP (Gross Planet Product). The IMF estimates for individual years since 1980 and continue to change when a new vintage of the World Economic Outlook data base is published (each year in October). The different vintages provide an indication of the extent of measurement error. According to two measures for measurement error the IMF data for Gross Planet product on average have an implicit minimal measurement error (IMME) of four percent and maximum ratio (MR) of eighteen percent. Even for long-term growth rates that are calculated over two decades growth rates have a substantial measurement error, namely an IMME of 1.7% and an MR of 8.0%. Measurement error of Gross Planet Product is thus economically and statistically significant and needs to be addressed in studies that analyse or use global production data. Measurement error in economics currently is significant, is not showing improvement over time and could be reported transparently without technical or budgetary problems.


Measurement error, IMF, GDP, world production, implicit minimal measurement error, maximum ratio.

Download the full text of this working paper


More information

About the author

Peter A.G. van Bergeijk is Professor of International Economics & Macroeconomics at the International Institute of Social Studies (ISS) of Erasmus University and leader of the research programme Economics of Development and Emerging Markets. His most recent book is Earth economics. An introduction to demand management, long run growth and global economic governance, published by Edward Elgar in 2013.