Introduction: measurement error in economics

Economists should consider that their data have on balance an implicit minimal measurement error of 5-10 per cent. An optimal strictly positive level of inaccuracy exists; striving for ever-increasing accuracy is counterproductive. Economic inaccuracy for designed, administrative and opportunity statistics could easily be reported transparently without any notable technical or budgetary problems. Reporting measurement error in the underlying data is an important element of the transparency that is part of academic integrity and professional applied research and should be part of the economic curriculum. The mainstream view that data problems are concentrated in the Global South is not helpful. It is true that the statistical capacity of developing countries needs to be strengthened, but the mainstream neglects the fact that very serious measurement errors also plague the statistics of the more advanced economies. Therefore, the statistical discourse needs to be decolonized.