DATA AND METHODS
Most previous studies for the euro area or CESEE countries focused on either GDP or industrial production to measure business cycle synchronization (de Haan et al., 2008). As GDP is clearly the most comprehensive output variable, we will use real GDP data (seasonally adjusted, 2005 price levels) from Q1 1999 to Q1 2012. The country sample includes the initial euro area countries (EA-12), the countries which joined the euro area between 2007 and 2011 (New EA-5), and the remaining non-euro area countries in the CESEE region, including Croatia (CESEE-8).5 All data are extracted from Eurostat and are therefore comparable both in the cross-section and time series dimension.
Calculation of Business Cycles
The output gap is an important determinant for monetary policy, as it is usually a good measure of future inflation pressures. Therefore, the synchronization of business cycles (i.e. output gaps) across countries is commonly referred to as the most important criterion of an optimum currency area (OCA). To extract the trend from the cyclical component of the GDP time series, various filter techniques have been proposed in the literature (see Gachter et al., 2012, for a comprehensive discussion).
To increase comparability with other studies, we apply the widely used Hodrick-Prescott (Hodrick and Prescott, 1997) filter to calculate cyclical components.6 The trend component is estimated by minimizing the deviation of the actual data points from this trend, with the smoothing parameter being determined ex ante. While the trend can be interpreted as potential output, the cyclical component corresponds to the output gap (i.e. fluctuation around the long-run trend).