An Analysis of the Reliability of First (Flash) Quarterly National Account Data Releases for Greece

Abstract

In this article, the reliability of first (flash) quarterly national account data estimates compiled and released by the National Statistical Service of Greece is examined. The more reliable the first estimates, the better equipped are policy-makers to assess the economic situation and outlook. Against this background, this article attempts to quantify revisions using a set of standard statistics to assess the size, the direction and the volatility of revisions with a view to evaluating the reliability of first quarterly national account estimates. This article not only examines total quarterly GDP revisions, but also evaluates revisions to all components of demand. The results provide evidence that the first quarterly estimate of total GDP in both y-o-y and q-o-q terms is barely revised in prospective NSSG releases, while higher revisions are observed for GDP components. This finding stems from the fact that revisions at a disaggregated level tend to cancel out at the aggregate level.

Keywords: Quarterly national account data estimates, Revisions, Bias test

JEL Classification: B22, E01, E02