The National Bank of Denmark has published a working paper setting out a method to estimate and correct occupational miscoding in US survey data, with the aim of reducing measurement error in observed occupational mobility and improving inference on how mobility relates to “task distance” between occupations. Using survey redesigns in the Current Population Survey and the Survey of Income and Program Participation, the authors infer occupation-pair-specific miscoding probabilities and use them to “de-garble” observed occupational flow matrices. The correction materially lowers estimated true mobility rates (for employer changers from over 60% to around 50%, and for year-to-year mobility among all employed from 45% to around 20%) and yields stronger empirical links between mobility and O*NET-based task distance across age groups, the business cycle and wage changes. The paper also proposes a miscoding-based distance measure, arguing that miscoding itself reflects task similarity and can help distinguish very close occupations and predict occupational transitions and wage losses following unemployment.