nr 004 - On the Predictability of GDP Data Revisions in the Netherlands

DNB Working Papers
Publicatiedatum 28 juli 2004
The first part of this paper is based on a study by Faust, Rogers and Wright (2004). They found some evidence of predictability of GDP revisions for the G-7 countrie s, especially for the UK, Italy and Japan. In this paper we investigate the quality of the first Dutch GDP releases by using the same technique. Our findings suggest that Dutch GDP revisions are also predictable to some extent. These results are strengthened when applying the more general state-space estimation procedure. The statespace model is used to estimate the final or unobserved data, given the preliminary or observed data. Key words: Preliminary data, final data, revision, GDP, state-space model, Kalman filter JEL codes: C12, C13, C22, C53