Monetary policy and data uncertainty: a case study of distribution, hotels and catering growth

This paper is a case study of the real world monetary policy data uncertainty problem. The initial and the latest release for growth rates of the distribution, hotels and catering sector are combined with official data on household income and two surveys in a state-space model. Though important to the UK economy, the distribution, hotels and catering sector is apparently difficult to measure. One finding is that the initial release data is not important in predicting the latest release. It could be that the statistical office develop the initial release as a building block towards the final release rather than an estimate of it. Indeed, there is multicollinearity between the initial release and the retail sales survey, which would then contain the same early available information. A second finding is that the estimate of the later release is sensitive to the estimate of the average historical growth rate. This means that establishing priors for this parameter and testing for shift structural breaks should be very important. -- Data Uncertainty ; Distribution Sector ; Kalman Filter ; Monetary Policy