How accurate are initial and revised government forecasts of economic fundamentals? The case of Taiwan

A government's ability to forecast key economic fundamentals accurately can affect business confidence, consumer sentiment, and foreign direct investment, among others. Government forecasts are subject to error, as can be seen by the frequent revisions that are made to initial, and even revised, official forecasts. A government forecast based on an econometric model is replicable, whereas one that is not based on an econometric model is non-replicable. Governments typically provide non-replicable forecasts of economic fundamentals, such as the inflation rate and real GDP growth rate. In this paper, we develop a model to generate one or more non-replicable government forecasts, examine the measurement errors contained in non-replicable government forecasts, compare replicable and non-replicable government forecasts using efficient estimation methods, and examine the accuracy of initial and updated (or revised) government forecasts. An empirical example to forecast economic fundamentals for Taiwan shows the relevance of the proposed methodological approach. The empirical analysis shows that replicable and non-replicable government forecasts can be distinctly different from each other, that efficient and inefficient estimation methods, as well as consistent and inconsistent covariance matrix estimates, can lead to significantly different outcomes, that government forecasts of economic fundamentals can differ markedly between initial and revised forecasts, and that alternative models and methods can lead to differences in the accuracy of government forecasts.