Ladies and gentlemen,
It is a great pleasure to make the closing remarks at the end of this very stimulating conference. Let me first thank the organisers of the conference for their excellent handling of a workshop which has been extraordinary successful in giving a lot of food for thought for all of us.
Today and yesterday you have debated very intensively the economic effects of creating a single currency – the euro – for an increasingly integrated European single market. I have no doubt you all agree with me that the achievements in terms of integration and convergence have been very impressive. The introduction of the euro has not only helped consolidating the tremendous gains that have been made in the process of European economic integration since the establishment of the European Coal and Steel Community (ECSC) in 1951. These gains have been both deep (covering an increasing number of markets) and wide (covering an increasing number of countries). I also strongly believe the euro is and will continue to be a key catalyst for further integration. One prominent example is the effect the introduction of the euro has had on accelerating financial integration. Notably, the elimination of exchange rate risk among euro area countries has fostered very significantly the depth and breadth of financial markets and boosted financial innovation. There is also evidence on much closer integration of stock markets in the euro area. Another important example is its effect on trade integration: evidence suggests that the euro has stimulated trade not only across the euro area, but also with the rest of the world. The export share with countries outside the euro area rose to almost 20% of GDP in 2003 from around 17% in 1997 (a similar development is recorded for imports). It is also remarkable that, possibly as a consequence of trade and financial integration, as well as the increasing degree of convergence of macroeconomic policies, dispersion of real GDP growth rates in the euro area has remained very close to its historical average and no signs of increased divergence in growth rates have emerged so far.
Today, I will not try to comprehensively review each and every of these important achievements. I would rather like to highlight one of them: the truly impressive process of inflation convergence in the euro area. In this context, I shall address the issue of inflation differentials within the area, in particular their size and recent evolution. I shall then focus on two particular issues. First, it has been argued that monetary union may contribute to a larger specialisation in production in the euro area. In this case, the need for relative price adjustments and inflation differentials to deal with sectoral developments may increase. I will briefly discuss the current evidence on sectoral specialisation. Second, I will elaborate on the function of the inflation differentials and emphasize the importance of monitoring relative unit labour cost.
Let me remind you that in the run-up to the formation of EMU, the dispersion of national inflation rates has been lastingly reduced. In 1992 the difference between the highest and lowest inflation rate in terms of HIPC across the euro area countries was 13.5 percentage points. The dispersion steadily declined throughout the 90s, triggered by the preparations for EMU, and broadly stabilised at the beginning of Stage Three of EMU at around 3.3 percentage points. Since then, with the exception of a modest increase over the period 2000-02, the level of inflation dispersion across the euro area has changed very little, with the average standard deviation of inflation differentials being around 1. Current inflation differentials within the euro area are not unusually large. Rather, their magnitude is similar to those differentials seen in the United States among the 14 US Metropolitan Statistical Areas, namely 3.2 percentage points, with an average standard deviation of about 0.9.
It should be emphasised that in a monetary union inflation differentials across regions or countries are the natural way of adjusting relative prices in the face of asymmetric demand or supply developments. I would like to make 4 remarks in this respect.
First, changes in relative prices within the single currency area have been larger and more frequent than anticipated. The idea that the euro area is an area where the single currency prevents adjustment of relative prices is clearly refuted. For instance, the capacity of the German economy to catch up rapidly with the average of the euro area in terms of cost competitiveness after the loss of competitiveness associated with German reunification is very impressive. From that viewpoint the euro area is more flexible and adaptable than had been anticipated by many observers.
Second, the previous process is symmetric. In some cases the existence of inflation differentials is the natural consequence for an economy that has been outperforming the average in terms of cost competitiveness in order to progressively adjust towards a more normal level in line with the average. This problem had been observed, for instance, in the case of the Netherlands and is possibly still relevant for Ireland.
Third, inflation differentials are not only the necessary means of achieving relative price adjustments within the euro area, like in the case of Germany, but they are also a natural feature of the euro area in view of the gradual convergence or catching-up process regarding GDP per-capita.
Fourth, there is one case where persistent inflation differentials are not advisable in terms of national economic policies: this is when an economy having an already relatively low level of cost competitiveness continues to face further deterioration due to higher inflation. In that case additional attention must be given to increasing the flexibility and adaptability of both institutions and market structures in the national economy concerned in order for the appropriate adjustment mechanism to function.
An interesting issue that has been touched upon in the last two days is whether the European integration process will result in higher sectoral specialisation across countries. As mentioned above, this is potentially important because a higher degree of specialisation might imply a greater vulnerability to asymmetric shocks and a greater need for relative price adjustments.
As to the relevance of this argument, there are two lines of reasoning in the economic literature. The first, supported by Krugman (1991, 1993) and the new economic geography models, suggests that, as trade barriers are reduced, inter-industry trade increases and opportunities for exploiting economies of scale and specialisation in production should arise whenever countries have a comparative advantage. Consequently, countries’ production structures would become less diversified and thus increasingly vulnerable to asymmetric shocks while business cycles would become less synchronised. The second, and opposite, line of reasoning, developed by Frankel and Rose (1998), suggests that European integration would instead lead to greater intra-industry trade integration and more similar economic structures. In this case, business cycles would become more strongly synchronised through the convergence of factor endowments and technology, as well as reduced exchange rate variability, which otherwise would tend to weaken intra-industry trade.
The evidence on whether EMU has favoured or prevented specialisation in the euro area is not clear yet. According to an overview published by the European Commission, there have so far only been modest changes in the pattern of industrial concentration and geographical specialisation within the euro area. While production specialisation has gradually increased since the 1970s, export specialisation appears to have decreased, which could be partly explained by the increased importance of intra-industry trade. Similarly, a study of sectoral specialisation carried out by the European System of Central Banks in 2004 found the production structure of euro area countries to be relatively similar. Indeed, it appears more homogeneous than in the United States and relatively stable over time. In contrast, a host of studies have highlighted agglomeration or clustering effects. They suggest that the agglomeration of production factors (by generating an environment that creates positive externalities, the presence of competitors and informal links with complementary businesses, the supply of qualified labour and investment capital, and the proximity of research capacity) is increasingly important and could indeed increase specialisation and reduce cross-border trade within industries.
At the same time, some recent research (for example, Kalemli-Ozcan et al., 2003, and Giannone and Reichlin, 2005) documents that risk sharing tends to increase in a currency union, bolstering the case for higher specialisation in production. The peculiarity of this channel is that any resulting asymmetries in GDP fluctuations would not translate in income volatility because ownership is diversified. This essentially will help to smooth consumption across countries and, therefore, limit the welfare cost of GDP fluctuations.
Overall, the time span since the start of the EMU is too short for strong empirical conclusions and a view of the effects of EMU on specialisation is consequently still difficult to obtain.
Let me turn now to the determinants of inflation dispersion in the euro area. A first insight into the possible causes of inflation differentials in the euro area can be gained by performing an inflation accounting exercise, which breaks down the euro area inflation differentials, measured in terms of the GDP deflator, into their underlying determinants, namely internal factors – such as unit labour costs, profit margins and net indirect taxes – and external factors – such as import prices. This kind of analysis suggests that during the period 1999-2003 internal factors were the most important contributors to inflation differentials relative to the euro area average in the case of 9 member countries, while import costs played a major role in the case of only 3 countries: Belgium, France and Luxembourg.
As regards the internal sources of inflation differentials, the most important component explaining the dynamics of GDP inflation has been the dispersion in unit labour costs. Looking at the unit labour cost developments in more detail, the compensation-per-employee component was generally the most important in explaining inflation differentials. At the same time, productivity growth showed a relatively low level of dispersion across the euro area. This points to relatively significant differences across countries in those components of labour compensation that do not reflect productivity growth. In this regard, part of the differences in price developments may stem then from differences in wage developments, including inertial components of wage-setting rules, such as those caused by automatic indexation clauses of wages to prices, which exist in some euro area countries.
To the extent that historical inflation differentials within the euro area seem to largely reflect differences in cost developments across euro area countries, I would like to stress here that the relative evolution of cost competitiveness, as measured for instance by unit labour cost, is a key equilibrating mechanism in the single currency area.
Ladies and gentlemen, let me make three remarks in conclusion.
First, there is no doubt in my mind that the full benefits of the European common market could not be realised without the single currency. I propose to you the simple thought experiment: What would be, today, the US market if we would have had different currencies in California and Texas, Massachusetts and Wyoming? Would we call it an “achieved single market”?
In that perspective, the single currency offers us all the economic benefits associated with an “achieved single market”. In this respect, as regards interest rates and exchange -rate risk for each particular economy, there are no “winners” and “losers”. It is a “win-win” game …. All economies, without exception, are benefiting from the very low market interest rates that reflect the credibility of the currency: it is true that this is a considerable additional advantage given to those countries that had previously very high interest rates. At the same time, all economies, without exception, are benefiting from the elimination of exchange risk within the single currency area, which is a considerable improvement, including for those economies that had already very low interest rates, so that not only each economy is benefiting from a much better functioning of the single market itself, but also from a significant improvement of its own monetary situation. It is equally true for Germany, for Spain, for Italy. It is true for the 12 countries forming the euro area. The single currency area is, in my view, a solid and coherent economy with all partners having strong advantages.
Second, I have associated myself with the strategic concept of “competitive disinflation” during all the period during which a previously high inflation economy, namely the French economy, was looking for being more effective in reinforcing its competitiveness through lower production cost and lower inflation rather than through monetary devaluation. Now we are in an area where monetary stability is fully ensured. This does not mean that our preoccupation with competitiveness should disappear. In this new environment of “competitive stability” the careful examination of national competitiveness indicators, like the relative evolution of unit labour costs, reemains important. The ECB has set up a benchmark of price stability in a medium and long term perspective - according to its definition - but a large part of economic policy remains national. In this respect I said, before the setting up of the single currency, that in the new monetary regime economic policy makers would substitute competitiveness indicators to previous indicators like exchange rates and market interest rates.
Third, let me say to which extent we rely on research to analyse the structural transformation ongoing in the single currency area by the virtue of a fully achieved single market with a single currency. My intuition is that this phenomenon is extremely complex. That explains why trends are so difficult to substantiate, and conjectures not easily proved or rejected: the two effects underlined by Krugman on the one hand and by Frankel and Rose on the other hand, are probably operating simultaneously in the euro area, together with the universal phenomenon of agglomeration or clusterisation.
All my experience tells me that in such a rapidly changing world, where the progress of sciences and technology, the globalisation process and the very profound structural changes in Europe are simultaneous, we have to be humble in front of facts and figures and be ready to take on board, as soon as it is produced, the good research you are delivering
I thank you for your attention
Reproduction is permitted provided that the source is acknowledged.