Evaluating growth cycle synchronisation in the EU

This paper provides an insight into the level of economic and monetary integration in Europe by analysing the degree of growth cycle synchronisation between seven European countries over the past thirty years. Two univariate trend-cycle decomposition methodologies, the Beveridge-Nelson (BN) decomposition and Harvey and Trimbur [Harvey, A.C. & Trimbur, T.M. (2003) "General Model-Based Filters for Extracting Cycles and Trends in Economic Time Series", The Review of Economics and Statistics, 85(2), 244-255.]'s unobserved component model, together with a multivariate extension of the BN decomposition incorporating trend and cycle restrictions, are used to identify the trend and cyclical components from real GDP for each of the seven countries. The cycles extracted from the two univariate approaches vary significantly in both cycle period and amplitude. The average correlation calculated from the BN cycles are also smaller than the corresponding correlation estimated using cycles extracted from the unobserved component model. This confirms the argument in Canova [Canova, F. (1998) "Detrending and Business Cycle Facts", Journal of Monetary Economics, 41(3), 475-512.] that the use of different trend-cycle decomposition methodologies may influence the results obtained. The results produced from the multivariate model indicate the presence of common features in the data. This may reflect the coordinated and common monetary and fiscal policies that these countries have shared over the sample period. However, the finding of codependent and heterogeneous growth cycles raises concerns about the operation of the European Monetary Union (EMU), as it implies that members may face significant stabilisation costs.