Tracking long-run growth in the euro area with an atheoretical tool
Tracking long-run growth in the euro area with an atheoretical tool: The role of real, financial, monetary, and institutional factors
Euro area economies have undergone significant changes and endured diverse shocks over the past 30 years. They were bound by the process of European economic and monetary integration, have experienced nominal convergence, and shared a single currency and monetary policy (Campos and Coricelli 2019). Upon the launch of the euro, several forces narrowed differences across euro area countries. At the same time, these countries have also experienced nominal exchange rate gyrations (1992-1993), the burst of the dot-com bubble (early 2000s), the Great Moderation, the financial turmoil starting in August 2007 and followed by the Global Crisis and the sovereign euro area crisis, and lastly, starting in 2013, a period of protracted low inflation. Figure 1 provides a glimpse of this tumultuous period. Responses to the crisis that started in 2007 are also relevant to our analysis. The ECB has implemented exceptional standard and non-standard monetary policies. Moreover, institutional reforms were introduced throughout the crisis, and we saw an enhanced pace of structural reforms in several countries.
We ask which factors may have played a role in stimulating growth or reducing it over the past three decades. Are these factors real, financial, monetary, and/or institutional? How do these factors interact? For now, ours is a detective story told in broad brush strokes. This is an analysis of correlation, not causality, but it flags links that are worth looking at more closely. Our approach is a novelty in the growth literature, which normally focuses on determinants largely based on the Solow exogenous model or endogenous growth theory. Our aim is to provide an a-theoretical toolkit looking at possible factors behind fluctuations and differences in growth rates among euro area countries since 1990.
A set of possible real, financial, monetary, and institutional factors
In a recent paper (Comunale and Mongelli 2019), we assemble a large set of real, financial, monetary, and institutional variables covering the period between 1990Q1-2016Q4 for euro area countries. We also look at two sub-groups of euro area countries: the euro area ‘core’ (Belgium, Germany, Finland, France, Luxembourg, and the Netherlands) and ‘periphery’ (Spain, Italy, and Portugal). Each of these blocks contains diverse variables. A representative variable for each block is highlighted in Figure 1.
Among the real variables we include fiscal deficit, debt over GDP, global GDP, and a proxy for price competitiveness as the growth rate of the broad real effective exchange rate (REER). In the context of sovereign and systemic (mainly financial) stress, we include indicators taking these aspects into consideration, which are especially important for the last ten years of data: the country-specific Composite Indicator of Sovereign Stress (SOVCISS) and the common Composite Indicator of Systemic Stress (CISS) as computed by Holló et al. (2012).1 In the financial block, we also include several measures of the financial cycle based on credit, house prices, and equity prices from the work done by the WGEM Team on Real and Financial Cycles (ECB 2018). We also make use of a new set of within-country synchronicity indices between cycles from Comunale (2020).2 For the monetary factor, we use shadow interest rates of Wu and Xia (2016) to capture both conventional and unconventional monetary policy actions. We also look at the European Index of Regional Institutional Integration (EURII), which maps developments in European integration from Dorrucci et al. (2015).
Figure 1 Growth rates and growth factors for the euro area
Note: These are data for the aggregate EA19. For GDP growth, we take the four-quarter moving average and +/-2 standard deviations. The vertical lines represent the dot-com bubble and the end of the tech-cycle, the start of the global financial crisis, and the sovereign debt crisis and the start of the recovery/low inflation period. The areas are selected following Hartmann and Smets (2018). EURII is the European Index of Regional Institutional Integration and CISS is the Composite Indicator of Systemic Stress.
We then select the factors that may have influenced growth based on the events above by using the weighted average least squares (WALS) by Magnus et al. (2010). Diagnostic tools are implemented to choose the correct estimators and setups. These include a heterogeneous panel error correction model, which helps us to quantify the contributions of the selected variables on growth in the short and long run, a dynamic factor model with instrumental variables to correct for cross-sectional dependence and possible endogeneity, and panel and country-by-country VARs for a subset of variables in order to shed light on the transmission channels and country-specific results (Comunale and Mongelli 2019).
EU institutional integration, competitiveness, and financial cycles matter
We find a positive association of EU institutional integration with long-run growth, particularly for periphery countries. This is a robust result across specifications and setups. If we further split our institutional index into its components, we see a large and significant positive role for financial and political integration, while economic and financial integration are ineffective in boosting growth. Deeper financial integration seems to have beneficial effects on the core, but is not significant in the periphery. The opposite holds for political integration, as an increase in the latter boosts long-run growth only for the periphery (Comunale and Mongelli, 2020).
An improvement in competitiveness matters in persistently sustaining long-run growth in the euro area, as does a decline in sovereign and, especially, systemic stress. Notably, as reported in Berg and Miao (2010), the REER is not a policy instrument, but mainly a result of policy actions and externalities. Thus, funding the appropriate, more productive sectors can increase competitiveness and then long-run growth. The global linkages and spillovers are further investigated by including world GDP, which seems to be positively linked to euro area growth, but only in the short run.
The equity price cycle is positively associated with GDP growth just pre-crisis, when some countries experienced a substantial increase in the magnitude of the positive side of the cycle. This is linked to higher growth only in the very short run, and it did not have a persistent effect on the overall performance. The loans to NFCs, instead, could have had a positive role for growth in the long run, and especially for the core countries. For the periphery, we do not see any significant impact of these loans on GDP growth. This result may depend on how the funding has been used in the different economies, i.e. for more- or less-productive sectors. The only synchronicity index that passes the WALS test is the one between long-term rates and loans to households, expressing the different links between long-term interest rates and financial conditions.3
The roles for public debt and (unconventional) monetary policy
Debt over GDP for the periphery is linked to negative growth only in the short run (and this drives the same results for the entire sample). This is in line with the general empirical literature on the relationship between public debt and economic growth, which is far from conclusive on the issue (Panizza and Presbitero 2014, Mika and Zumer 2017).
The monetary policy stance is proxied by the short-term rates until the zero lower bound, and then by the shadow rates (Wu and Xia 2016). There was a strong co-movement between EONIA (in levels) and GDP growth interrupted at the crisis in 2008Q3 (Figure 1). In our outcomes, the coefficient for the rates is positive but only significant in the long run. As expected, the sign is always positive, because monetary policy is set endogenously – when GDP rises, interest rates are set to go up; if GDP declines, the rates are set to decrease. In fact, in the early part of the sample, GDP leads interest rates. Stagnation after the Great Recession accounts for the monetary policy stance taken to react to the situation. From the time of the sovereign debt crisis, the transmission mechanism broke down, and monetary policy has been accommodating (to increasing degrees). In 2013, we witness a decoupling of the shadow rate from GDP growth, and the shadow rate captures the unconventional monetary policy.
Importantly, our results must be considered as preliminary, and would require corroboration by further analysis. We cover a period of intense flux in European economic, financial, monetary, and institutional history. Some of the countries in the sample experienced switches in policy regimes. Thus, much remains to be done in future research. For the factors, the role of EU funds might also be taken into account. This possible determinant is not included here because of (still) limited availability of supportive data in its time dimension. Similarly, it’s too soon to assess the growth effects of the new euro area governance and the working of the Single Supervisory Mechanism.
Authors’ note: The views expressed in this column are those of the authors and do not necessarily represent those of the ECB, the Bank of Lithuania or the ESCB.
Berg, A, and Y Miao (2010), “The Real Exchange Rate and Growth Revisited; The Washington Consensus Strikes Back?“, IMF Working Papers 10/58, International Monetary Fund.
Campos, N and F Coricelli (2019), “Euro 20/20: Twenty papers to better understand the single currency”, VoxEU.org, 8 July.
Comunale M, and F P Mongelli (2019), “Who did it? A European Detective Story. Was it Real, Financial, Monetary and/or Institutional? Tracking growth in the Euro Area with an atheoretical tool”, Bank of Lithuania, Working Paper Series no. 70/2019.
Comunale, M, and F P Mongelli (2020), “Euro Area growth and European institutional reforms”, Forthcoming in N Campos, P De Grauwe, and Y Ji (eds), Structural reforms and economic growth in Europe, Cambridge University Press, March.
Comunale, M (2020), “New synchronicity indices between real and financial cycles: is there any link to structural characteristics and recessions in EU countries?”, International Journal of Finance and Economics, forthcoming.
Dorrucci, E, D Ioannou, F Mongelli and A Terzi (2015), “The four unions ‘PIE’ on the Monetary Union ‘CHERRY’: A new index of European Institutional Integration”, Occasional Paper Series, no. 160, ECB.
ECB, WGEM team on Real and Financial Cycles (2018), “Real and financial cycles in EU countries – Stylised facts and modelling implications“, Occasional Paper Series no. 205, ECB.
Hartmann, P, and F Smets (2018), “The first twenty years of the European Central Bank: monetary policy“, Working Paper Series no. 2219, ECB.
Holló, D, M Kremer, and M Lo Duca (2012), “CISS – a composite indicator of systemic stress in the financial system“, Working Paper Series no. 1426, ECB.
Magnus, J R, O Powell, and P Präufer (2010), “A Comparison of Two Model Averaging Techniques with an Application to Growth Empirics”, Journal of Econometrics 154: 139-153.
Mika, A, and T Zumer (2017), “Indebtedness in the EU: a drag or a catalyst for growth?“, Working Paper Series no. 2118, ECB.
Panizza, H, and A F Presbitero (2014), “Public debt and economic growth: Is there a causal effect?”, Journal of Macroeconomics, 41, 21–41.
Wu, J C, and F D Xia (2016), “Measuring the Macroeconomic Impact of Monetary Policy at the Zero Lower Bound”, Journal of Money, Credit and Banking, 48 (2-3).
 CISS is an indicator that uses information from equity, bonds, exchange rate volatilities, banks, and payments systems and weights more when the stress has been found in several markets at the same time.
 These new measures provide us with proxies of macro-financial co-movements, capturing whether positive and negative cyclical phases coincide, i.e. they can be either both positive or both negative, regardless of their amplitudes.
 This is highly heterogeneous across EU members, being higher in Finland and Portugal (80% of the times these cycles are synchronised) and much lower in Germany, the Netherlands, Spain, and Italy (less than 60%) (Comunale 2020).