Monetary Policy, Data Uncertainty and the Supply Side : Living with the Statistical Fog

Economic statistics are often only an approximation of the underlying reality they are attempting to measure. Official statistics, especially in timely first releases, can be subject to significant uncertainty and revision. In this speech, Marian Bell, a member of the Monetary Policy Committee, looks at the various ways in which data uncertainty might arise and discusses how policymakers should respond. A particular difficulty for statisticians can be the allocation of movements in nominal variables to changes in prices and volumes. Such uncertainty is high when the structure of the economy and relative prices change significantly. In such circumstances, looking at a money value measure of activity may be informative. Moreover, estimated levels of real activity appear more prone to revision than growth rates, so we should be wary of conventional, mechanical estimates of the output gap, which rely heavily on information on the level of activity. Ms Bell concludes that revisions to data stemming either from more information or methodological improvements will inevitably lead us to reassess our view of the underlying truth. Improvements to statistics are to be welcomed, and understanding the nature of data uncertainty and revisions should inform our judgements about the state of the world