Quantitative evidence of post-crisis structural macroeconomic changes Roberto Camagni, Roberta Capello, Andrea Caragliu, Barbara Chizzolini Politecnico di Milano To be discussed at the Advisory Board Forum, Brussels, 12 June 2018
Aim of the presentation Quantitative evidence of post-crisis structural macroeconomic changes is here provided with the aim to discuss with the advisory board members about: the possible consequences of such changes for Europe and its territory; the possible evolutions of such changes; the linkage of such changes with the policy debate.
Groups of countries obtained through a custer analysis on GDP performance in the post-crisis period (2012-2016) Low growth countries Medium growth countries High growth countries Group Country 1 Cyprus 1 Finland 1 Greece 1 Italy 2 Austria 2 Belgium 2 Croatia 2 Denmark 2 France 2 Germany 2 Netherlands 2 Portugal 2 Spain 2 United Kingdom 3 Bulgaria 3 Czech Republic 3 Estonia 3 Hungary 3 Ireland 3 Latvia 3 Lithuania 3 Luxembourg 3 Malta 3 Poland 3 Romania 3 Slovakia 3 Sweden 3 Slovenia
GDP levels 2000-2017 100 110 120 130 140 2000 2005 2008 2010 2012 2015 year Cluster 1 (Low growth Countries) Cluster 3 (High growth Countries) Cluster 2 (Medium growth Countries)
Comments Clusters look less geography-dependent (East-West, North-South divide less visible than before the crisis). Low growing countries are not only the Southern ones; not all Eastern countries are fast growing; Northern countries are present in all groups; The relative performance of the clusters in the postcrisis period looks similar to the other two periods, namely: fast growing countries were also faster before the crisis and with limited effects of the crisis; the growth rate of the post-crisis period is higher than the pre-crisis; low growing countries were also growing less in the previous two periods and less after crisis than in the pre-crisis; the medium growing countries always lied in between the other two, and show a simialr performance before and after the crisis.
Possible explanations on the differentiated growth paths: pre-crisis and post-crisis investments trends Low-growing countries Medium-growing countries Fast-growing countries Gross Fixed Investments (E_IFK): trends pre and post crisis cluster 1 Gross Fixed Investments (E_IFK): trends pre and post crisis cluster 2 Gross Fixed Investments (E_IFK): trends pre and post crisis cluster 3 300,000 250,000 200,000 250,000 225,000 200,000 175,000 40,000 30,000 20,000 150,000 100,000 50,000 0 15-95 15-96 15-97 15-98 15-99 15-00 15-01 15-02 15-03 15-04 15-05 15-06 15-07 15-08 15-09 15-10 15-11 15-12 15-13 15-14 15-15 15-16 15-17 150,000 125,000 100,000 75,000 50,000 10,000 0-10,000 4-95 4-96 4-97 4-98 4-99 4-00 4-01 4-02 4-03 4-04 4-05 4-06 4-07 4-08 4-09 4-10 4-11 4-12 4-13 4-14 4-15 4-16 4-17 3-95 3-96 3-97 3-98 3-99 3-00 3-01 3-02 3-03 3-04 3-05 3-06 3-07 3-08 3-09 3-10 3-11 3-12 3-13 3-14 3-15 3-16 3-17 E_IFK TREND pre TREND post E_IFK TREND pre TREND post E_IFK TREND pre TREND post Legend: Red - pre-crisis (1995-2008) trend Green post-crisis (2012-2017) trend Blue annual investments
Possible explanation for the differentiated growth paths: investments trends Comparing post-crisis with pre-crisis investment trends: low-growing countries show a similar investment trend (but lower than the other two groups of countries); medium-growing countries have a steeper investment trend; fast-growing countries have a much steeper investment trend.
Long run explanation of investment growth: 1995-2012 vs. 1995-2015 Pre-crisis period Dependent Variable: investment growth rate 1995-2012 Coeff. Std. Error t-statistic Prob. Constant -1.29 0.55-2.35 0.02 FDI growth rate (t-1) 0.01 0.00 1.42 0.16 DGDP growth rate (t-1) 0.68 0.18 3.75 0.00 Interst rate -0.49 0.00-4.08 0.00 Unit labour cost -0.15 0.02-6.36 0.00 Dummy crisis -0.07 0.01-7.27 0.00 Investment trends -1.15 0.10-11.33 0.00 Speed of adjustment -0.41 0.05-8.64 0.00 of investment ot a long run value Pre-crisis, crisis and post-crisis period Dependent Variable: investment growth rate (1995-2015) Coeff. Std. Error t-statistic Prob. Constant 0.435 0.514 0.846 0.398 FDI growth rate (t-1) 0.189 0.087 2.172 0.030 DGDP growth rate (t-1) 0.387 0.151 2.556 0.011 Interst rate -0.018 0.002-10.142 0.000 Unit labour cost -0.001 0.001-1.212 0.226 Dummy crisis -0.076 0.011-6.995 0.000 Speed of adj of I to long run trends -0.427 0.039-10.996 0.000 Investment trends -0.807 0.102-7.924 0.000 Dummy post-crisis -0.067 0.013-5.220 0.000 DGDP growth rate (t-1) in the post-crisis period 1.920 0.359 5.342 0.000
Comments In the post-crisis period: the reactivity of investment growth to GDP growth triplicates: higher cumulative effects (I GDP I); investments become more volatile, i.e. they are less linked to their long-term trend.
Possible explanation for the differentiated growth paths: export performance (1995-2016) Variable Coefficient Std. Error t-statistic Prob. Euro/US$ exchange rate 1.402918 0.224092 6.260458 0.0000 Deflator in hi-med countries (wrt. low-growing) 0.108632 0.058504 1.856837 0.0639 Deflator in low-growing countries -0.721240 0.203870-3.537738 0.0004 Japan and US GDP growth rate 0.003889 0.001151 3.378030 0.0008 BRIC GDP growth rate 0.006684 0.001555 4.299001 0.0000 2009-0.122844 0.018533-6.628260 0.0000 Eastern countries 0.011816 0.005140 2.298914 0.0219 Constant -0.004438 0.010134-0.437980 0.6616 R-squared 0.419266 Mean dependent var 0.055669 Adjusted R-squared 0.411821 S.D. dependent var 0.073059 S.E. of regression 0.056031 Akaike info criterion -2.911490 Sum squared resid 1.714150 Schwarz criterion -2.849148 Log likelihood 814.4826 Hannan-Quinn criter. -2.887135 F-statistic 56.31280 Durbin-Watson stat 1.413576 Prob(F-statistic) 0.000000
Comments In the post-crisis period: rise in price deflator hits only low growing countries; medium and fast growing countries instead suffer less (due to likely high price competitiveness and to likely specialization in sectors with anelastic demand). These last countries perform better due to a wider structural transformation in their economies.
Possible explanation for the differentiated growth paths: R&D investments/gdp levels (2000-2016) 100 110 120 130 140 2000 2005 2008 2010 2012 2015 year Cluster 1 (Low growth Countries) Cluster 3 (High growth Countries) Cluster 2 (Medium growth Countries)
In the post-crisis period: Comments - a general slowdown of R&D/GDP growth rate is registered, and in low growing countries R&D / GDP level even decreases; - the structure does not change: medium growing countries are above the other two groups of countries in R&D investment levels and growth since 2008; - fast-growing countries increase their innovative activities, even if they remain far below the other two groups.
Possible explanation for the differentiated growth paths: export/gdp levels 2000-2016 100 110 120 130 140 2000 2005 2008 2010 2012 2015 year Cluster 1 (Low growth Countries) Cluster 3 (High growth Countries) Cluster 2 (Medium growth Countries)
Comments In post-crisis period: - export / GDP level attenuates its increase in all groups of countries; - fast-growing countries register the highest export levels, and remain those with the highest increase. Devaluation of local currencies w.r.t. the euro and lower prices of goods are at work; - medium-growing countries demonstrate the lowest competitiveness in conquering new international market shares.
GDP loss as % of border regions' GDP Border regions' GDP / Country GDP Possible explanation for the differentiated growth paths: presence of border effects 12% 100% 90% 10% 80% 8% 70% 60% 6% 50% 40% 4% 30% 2% 20% 10% 0% 0% GDP loss (left scale) Share of border regions on Country's GDP (right scale) : Cluster 1 : Cluster 2 : Cluster 3
Geographical exposure to border losses A taxonomy of legal and adm. barriers GDP losses by country 100% High exposure 90% Slovakia Slovenia Czech Republic Luxembourg Austria Hungary 80% 70% Belgium 60% Netherlands Denmark 50% Latvia 40% Lithuania Sweden Europe Spain Italy 30% Portugal Estonia Bulgaria Romania Poland France Germany Ireland 20% 10% Low exposure Finland Legend: Geographical exposure to border losses = share of border regions GDP on country GDP Degree of border regions losses = level of losses United Kingdom 0% 4% 5% 6% 7% 8% 9% 10% 11% : Cluster 1 : Cluster 3 Degree of border regions' losses Low losses : Cluster 2 High losses Greece
Total GDP loss due to border effects as share of NUTS3 GDP Losses due to border effects tend to accrue to economic powerhouses of the EU economy, where the highest endowment of growth factors are present and therefore the highest sensitivity to sub-optimal use of these resources. A large variability emerges within each land border region, with large agglomerations emerging as areas suffering the most.
Comments Geographical fragmentation has a high economic cost, since is generates suboptimal use of resources. The costs of legal and administrative barriers represent a high loss: 3% of EU28 GDP and 9% of GDP of border regions (source: Camagni et al., 2017). This cost influences all groups of countries, but especially medium-growing countries. Fast-growing ones are less exposed and have a lower loss. Losses tend to accrue to economic powerhouses of the EU economy,
Regional disparities (Theil index).05.1.15.2.25 2000 2005 2008 2010 2012 2015 Year Between countries Theil index Total Theil index Within Countries Theil Index
Trends in regional disparities The Theil index confirms previous forecasts of the MASST model (ET2050), namely: the end of inter-national reduction of disparities; the continuing increase of intra-national disparities; the increase since 2008 of overall regional disparities.
Within countries regional disparities by groups of countries.02.03.04.05.06 2000 2005 2008 2010 2012 2015 Year Within Theil, low growth Countries Within Theil, high growth Countries Within Theil, middle growth Countries
Within countries (intra-national) regional disparities by groups of countries The Theil index shows: fastest growing countries show a faster increase in internal disparities since the beginning of the crisis; all clusters show an increase in internal disparities; this increase started well before the crisis (2003-2004) in the case of fast growing and medium growing countries.
Disparities between agglomerated and rural regions.75.8.85.9.95 2000 2005 2010 2015 2008 2012 year
Disparities between agglomerated and rural regions The Theil index between agglomerated and rural regions shows: a reduction during the pre-crisis period, in which rural areas where growing; a stability during the crisis, due to the downturn which characterised agglomerated areas; an increase after the crisis.
Tentative conclusions (after crisis) A geograhically-neutral, multi-speed Europe; crucial role of investments and structural change; important role of price competitiveness in highgrowing countries, and a limited role of R&D investments; high economic costs of geographical fragmentation, spread around all European countries;
Tentative conclusions (after crisis) increase in regional disparities leading possibly to increased political fragmentation; the opening of a new dichotomy between urban and rural areas (with similar effects on political fragmentation); crucial role of both macroeconomic (national) and territorial elements multi-scalar, selective policies needed; regional policies should become stronger, more effective and, most of all, more visible.