The prevailing literature studies whether typically the type of crises change as time passes and/or across countries. Tests by Gupta et al. (2007), Rose and Spiegel (2010, 2011, 2012), and Frankel and Saravelos (2012) are among many that test for such differences taking a ‘early warning system’ framework. While this framework permits comparisons of crises typically, it does not permit the identification of case-specific information. The info about whether crises will vary from each other typically hides possible divergences from average. Typically, crisis behaviour could possibly be sufficiently different across regions (or across time), yet two crises occurring in two different regions (or at two different schedules) could share sufficient similarities.
As a way to identify such case-specific similarities, an instrument that will not rely solely on the info regarding the partnership between averages but also considers individual specific information is necessary. One such tool may be the matching technique, which aims to statistically match similar observations and offer a means of clustering observations according to a couple of pre-determined dimensions. In Sayek and Taskin (2014), we utilize the propensity score matching methodology to supply an alternative solution anatomy of the ongoing European financial meltdown in light of the globally accumulated banking crisis experience.
In this propensity score matching exercise, the treated units are thought as the crisis episodes that occurred on or after 2007 – labelled as ‘new’ crises – and the control group as all of the crisis episodes that occurred ahead of 2007 – labelled as ‘old’ crises. Depending on whether a country has been around a crisis at some time, the brand new crises are matched with the old crises. As such, the dataset used to handle the matching exercise includes the 165 episodes of crisis (non-crisis episodes aren’t contained in the dataset), where ‘being in an emergency currently’ is interpreted as cure.
The existing crisis episode is marked by its global nature and by occurring predominantly in high-income countries. The matching results claim that both of these characteristics of the recent crisis episode are very distinctive – the existing crisis shares no commonalities with past crises in either the dimensions defining the global economic environment or their national income or institutional qualities. However, despite these differences, the GIIPS economies share extensive commonalities within their respective pre-crisis domestic vulnerabilities with several past banking crisis experiences. Specifically, the euro periphery crises match mainly with crises of the 1990s, which are either among the big-five crises or the East Asian crises (see Figure 1).
Figure 1 . Radius matching
The current-account imbalances and the evolution of private sector credit show significant similarities between your respective pre-crisis periods of the recent Portuguese crisis and the 1996-1997 Malaysia and Thailand crises. However, the build-up of the recent Irish and Spanish crises tell the 1992 Japanese crisis a house bubble that was fuelled by an exuberant domestic credit expansion. The unlikely match of the recent Italian case and the 1991 Finnish crisis appears to be mainly due to the pre-crisis low degrees of economic activity and the transmission of global slowdowns through international trade linkages. However, the match between your recent Greek crisis and the 1996-1997 crises of Indonesia and the Philippines bear significant similarities within their public sector debt patterns.
These matches provide important information on what similar the GIIPS crises are among themselves. Findings claim that the GIIPS crises encompass some very dissimilar crises and also very similar ones. For instance, the Spanish and Irish crises share a substantial amount of similarities within their pre-crisis conditions, whereas the Greek crisis is quite distinct from all the GIIPS crises. This finding is evidence against a one-size-fits-all policy prescription for the GIIPS countries. Therefore, the policy design of every country’s recovery should look at the particularities of every crisis. The duration and severity of the matched past crisis experiences suggest the street ahead will still be very rough. The precise matched crises of Indonesia, Malaysia, and Japan took 6, 7, and a decade to disappear, respectively. That is suggestive that policy design shouldn’t only be case-specific, but also needs to be designed very proactively. Unless a shift in the policy structure is implemented in Europe, the dismal growth conditions in today’s crisis countries will probably continue for quite some time more.
1. In the rest of the column, the euro crisis will refer mainly to the crisis of the periphery Eurozone countries, namely Greece, Ireland, Italy, Portugal, and Spain (GIIPS).
2. The terminology owes to Eichengreen (2008).
CEPR Business Cycle Dating Committee (2014), “Eurozone mired in recession pause”, VoxEU.org, 17 June.
Eichengreen, B (2008), “Sui generis EMU”, NBER Working Paper 13740.
Frankel, J and G Saravelos (2012), “Can leading indicators assess country vulnerability? Evidence from the 2008-09 global financial meltdown”, Journal of International Economics, 87: 216-231.
Gupta, P, D Mishra, and R Sahay (2007), “Behavior of output during currency crises”, Journal of International Economics, 72: 428-450.
Rose, A K and M M Spiegel (2010), “Cross-country causes and consequences of the 2008 crisis: International linkages and American exposure”, Pacific Economic Review, 15: 340-363.
Rose, A K and M M Spiegel (2011), “Cross-country causes and consequences of the 2008 crisis: An update”, European Economic Review, 55: 309-324.
Rose, A and M M Spiegel (2012) “Cross-country causes and consequences of the 2008 crisis: Early warning”, Japan and the World Economy, 24, 1-16.
Sayek, S and F Taskin (2014), “Financial crises: lessons from history for today”, Economic Policy, forthcoming.