The pervasiveness of measurement error

A review of findings on measurement error, typically scattered in the literature, finds that structural awareness of and interest in measurement error is haphazard and often located in peripheral areas and documents the progress since Morgenstern (1950) published his monograph. The chapter takes stock of the recent literature so that we can learn from the findings, which also concern ‘non-classical measurement errors’. Starting from Morgenstern’s original findings that cover trade, prices, natural resources, agriculture, (un)employment and national income, overviews of relevant literatures and real-world examples in textboxes are provided. The origin and nature of inaccuracies of key economic observations have changed but the size of measurement error persists even today. Many instances exist where measurement error is ‘non-classical’: measurement errors often are not randomly distributed, infrequent and arbitrary, depend on the dependent variable and/or explanatory variable and on earlier measurement errors. The chapter also discusses the Sustainable Development Goals.