Home Political science A New Model for Balanced Growth and Convergence: Achieving Economic Sustainability in CESEE Countries
Cyclical Heterogeneity within the CESEE Region
First of all, we examine the development of business cycle synchronization within the CESEE region and the euro area, respectively. The top panel in Figure 10.2 shows the standard deviations of cyclical components for various country samples. The increase in the dispersion of cyclical components across all country aggregates since the end of 2006 clearly stands out.7 Interestingly, this divergence starts approximately two years before the crisis in the corresponding boom period, while it somehow declines slightly in mid-2008 (when the cycles seem to move downward simultaneously). During the recovery phase in 2009, the standard deviation peaks the second time, before it starts to decline again in the light of the sovereign debt crisis that emerged in 2010, where economic activity weakens all over Europe. While this pattern is also observable for the EA-12 countries, it is much more pronounced for the CESEE-8 countries. Obviously, developments in CESEE economies during the crisis were much more heterogeneous than in the euro area. A large fraction of the observed divergence is attributable to the Baltic countries, which experienced unusually high growth in the mid-2000s at the price of increasing macroeconomic and financial imbalances leading to an unprecedented boom-bust cycle (see Figure 10.3).
For the group of more recent euro area member countries (New EA), it is theoretically ambiguous in which country aggregate they actually fit best. On the one hand, they belong to the group of new member countries which joined the EU in 2004, and therefore, should be classified as emerging market economies belonging to the CESEE aggregate. On the other hand, they fulfilled the convergence criteria very quickly and joined the currency union in recent years. Therefore, it is an empirical question whether the patterns point to an inclusion in the euro area or CESEE subsample, respectively. The two dashed lines in Figure 10.2 show the dispersion of cycles in the euro area and CESEE countries, in each case including the New EA-5 economies. While the standard deviation of the enlarged CESEE sample (13 countries) hardly deviates from its original values for the CESEE-8 countries, the dispersion of the euro area sample
Figure 10.2 Cyclical heterogeneity in the CESEE region and in the euro area
significantly increases when including the New EA-5 countries, at least for the period from 2006 to 2010. Thus, we conclude that the more recent euro area countries rather tend to follow the patterns of other (non-EMU) CESEE countries than the 12 (original) euro area countries.
Similar conclusions can be drawn from the bottom panel in Figure 10.2, where we report average correlations of the cyclical components with the EA-12 cycle and the CESEE-11 cycle in two-year rolling windows. For all country groups, the (comparatively small) recession in 2003-2004 led to a decrease in cyclical correlations, although the synchronization appears generally smaller in CESEE-8 than in the EA-12. At the beginning of the recovery, starting in Q1 2005, we observe a distinctive increase in cyclical correlations for all countries, which shows that the convergence process within the CESEE region is not only triggered by their accession to the EU, but also by other factors. At the cyclical peak in Q1 2008, there is a decline in synchronicity for a short period of time for all country groups, but to a much higher extent in CESEE economies than in euro area countries. In particular the New EA-5 countries seem to deviate substantially, which is likely caused by differing cyclical peaks among these countries (i.e. their cyclical peaks take place at different points in time).
A similar pattern can be observed in early 2010, during the most recent decrease in synchronization. However, contrary to 2008, the New EA countries closely mirror the behaviour of the other euro area countries, while the CESEE-8 countries record a slightly stronger decline in cyclical synchronization.
In brief, two main insights stand out: firstly, that the New EA-5 countries show a similar pattern as the CESEE-8 region, although they seem to converge towards the 12 initial euro area members at the end of the sample period; and secondly, while the dispersion of output gaps shows a significant increase in cyclical divergence since 2006, the pattern of cyclical correlations reflects an increase in synchronicity after the recession in 2004, although the end of the boom (Q1 2008) and the current crisis (since Q4 2009) are marked by declines, particularly in the CESEE-8 countries. Overall, we observe a higher heterogeneity among CESEE countries than within the euro area with respect to both the dispersion and correlation of business cycles, although the homogeneity seems to increase substantially towards the end of the sample period.
Figure 10.3 compares individual country cycles in the CESEE region and in the euro area, giving some insights into the economic adjustment of small and large economies. While the cyclical deviations are generally larger in the CESEE region as compared to the euro area (as indicated by the larger cyclical swings), the cyclical components are also smaller
Figure 10.3 Individual country cycles in the CESEE region and the euro area
Figure 10.3 (continued)
in larger countries across the board, both in the euro area and in the CESEE region.8 The largest cyclical swings (i.e. output gaps) during the crisis are reported for small open economies in CESEE, such as the Baltic countries, where the output gaps exceed 10 per cent of potential GDP in certain time periods. Poland, on the other hand, seems to have the most stable cycle, and thus, the lowest average output gap during the crisis in the CESEE region. Similarly, the remaining three large economies (Czech Republic, Romania, Hungary) also exhibit substantially lower cyclical swings than their smaller counterparts. However, as explained below, there is no obvious difference between large and small economies in terms of correlations.
The smaller swings of larger countries support the hypothesis that larger economies are less reliant on external markets, and thus, achieve more stable cycles. One obvious explanation is the relationship between the size and openness of an economy, as shown in Figure 10.4. Total trade - that is, the sum of exports and services (of goods and services) relative to GDP - tends to be lower in larger economies, which might imply smaller (more stable) cyclical swings and/or less dependence on external developments.9 For the following analysis, where we examine the decoupling of CESEE from the euro area economy, we therefore divide our CESEE country sample into large (the Czech Republic, Hungary, Poland, Romania) and small economies (Bulgaria, Croatia, Estonia, Latvia, Lithuania, Slovenia, Slovakia). While the threshold between large and small economies is somehow arbitrarily set, the four larger countries also differ significantly in their economic policy. While the small economies either have already adopted the euro or exhibit a fixed exchange rate regime (currency board), the four large countries in our sample still feature floating exchange rates, which might enhance economic adjustment during the crisis significantly.
The following section focuses on a comparison of the euro area and the CESEE region and raises the question whether individual CESEE countries have decoupled or rather converged to the euro area cycle in recent years. As the New EA-5 countries show similar patterns to the CESEE region (as explained above), we include Slovenia, Slovakia and Estonia in the CESEE aggregate, summing up to a CESEE-11 sample. Next, due to their small size and geographical location, we subsequently exclude Malta and Cyprus from our samples. Furthermore, we will focus on the distinction between small and large countries in the CESEE region.
|< Prev||CONTENTS||Next >|