Political Economy of Statistics: Manipulating Data

Abstract

This paper focuses on the incentives by governments to manipulate statistics, not on techniques how to do so. I wish to contribute to explaining when, where and to what extent the governments tend to falsify official data. This issue is important for three reasons: (1) Manipulating statistics undermines the very essence of data collection; (2) Falsified statistics corrode policy action; (3) Econometric estimates based on manipulated data systematically distort the results. Most of the debate in economics journals on identification and potential bias in econometric models assume that the data has not been falsified.