 |
|
28.06.2007
EMU: Divergence or Convergence? This article appeared orginally in Nouriel Roubini's Blog at RGE Monitor.
Has the process of monetary unification in Europe led to economic convergence or economic divergence in the Eurozone since 1999? This is a key issue that was discussed at a conference on EMU Divergences in Berlin last week that was co-organized by RGE. Indeed, the long-run performance and success of the EMU depends on whether economic convergence or divergence is occurring. The ideas and debates at the conference are nicely summarized by Sebastian Dullien and Daniela Schwarzer (who co-organized the conference) in a blog at Eurozone Watch.
Before the formation of EMU there were long debates on whether or not the Eurozone was an optimal currency area. The economic criteria for a successful monetary union were widely debated in principle. Now several years after the formation of the monetary union (formally started in 2002 but already effectively in place since 1999) there is enough economic data to make a preliminary assessment of the success of the monetary union.
It is useful to go back to some of the conceptual criteria for a successful monetary union and then test whether the economic performance of EMU has been consistent with such criteria.
Here is a concise overview of the economic criteria that make a currency or monetary union desirable:
a) There are little asymmetries in shocks and in macroeconomic transmission so that business cycles (per capita output levels and growth rates) are not widely and persistently divergent across countries;
b) Consumption risk is sufficiently diversified across borders. In other terms, international financial markets work efficiently, so that agents can easily smooth consumption via risk sharing and international borrowing and lending across time.
c) Fiscal policy – at the national and union level - can help stabilize national economies given asymmetric shocks;
d) Prices and wages are sufficiently flexible so that relative prices (including real exchange rates) can adjust sufficiently even in the absence of a domestic currency;
e) Factors are sufficiently mobile across regions of the union also in the short run, at low private and social costs.
In the RGE paper (co-authored by Christian Menegatti, Elisa Parisi-Capone and myself) that we presented at the EMU conference we conducted a systematic overview of the evidence on economic convergence or divergence within the EMU.
In summary we found that,
- There is only little evidence of per capita GDP convergence (in PPP terms), with Ireland being the only true example of standard of living catching up. Portugal, Greece and Spain do not show evidence of per capita income catching up. Italy shows a decline in the standard of living.
- Based on some measures there has been some decrease in growth dispersion within the EMU countries. Whether the recent pick up in growth in the Eurozone will continue this trend is a still an open issue.
- Financial channels (credit and capital markets) provide only very modest degrees of smoothing of national shocks in the EMU, especially compared to their role within the United States. Only 7% of shocks were absorbed during the full sample. However, this risk-sharing channel via the financial channels and the fiscal channel has significantly improved over time in the Eurozone and during the EMU period. 36% of the idiosyncratic shocks are now smoothed via these channels across the Eurozone. This is consistent with the evidence that the degree of financial integration has increased over time in the Eurozone.
- The role of the financial channel is larger than the one of the fiscal channel; even in the most recent EMU period only 9% of asymmetric shocks are absorbed by the fiscal channel (as opposed to 1% for the full sample period). In the most recent EMU period the financial and credit channel smooth 27% of shocks.
- The real exchange channel provides a mixed picture: there is some evidence of mean reversion, especially in Germany; but there is also evidence of persistent loss of competitiveness in countries such as Italy, Spain and Portugal.
- Finally, real interest rates have moved in way that has been partly destabilizing: based on trends in housing markets and other variables differences in real interest rates may have exacerbated financial and asset price bubbles in some economies with the risk of a boom-bust cycle. This is what happened to Portugal and to the Netherlands and what could happen to Spain.
Based on these results what can we say about the success or limitations of EMU? I would summarize the evidence from our paper and from the discussions at the conference as follows:
The process of real convergence among EMU countries has been a mixed bag. Per capita income convergence has stalled as only Ireland shows evidence of “catch-up” of per capita income. Per capita income has stalled in Italy and fallen in Portugal while the catch up of living standards in Spain and Greece has slowed down. Growth differential dispersion has not increased on average for the Eurozone but the divergence among the four largest economies (Germany, France, Italy, Spain) that represent 80% of the GDP of the Eurozone has increased.
Also growth differentials appear as persistent: until recently the laggards (Italy, Portugal and, until 2006, Germany) were persistently growing very slowly while a group of other countries – Spain, Ireland - were growing very fast.
Persistent asymmetric shocks have hit the economies of the Eurozone: given the rise of China and Asia, Eurozone countries with traditional comparative advantage in labor intensive manufacturing (textile, apparel, shoes, etc.) have sharply lost competitiveness and experienced a real appreciation that is a sign of currency misalignment. These are Italy. Portugal, Spain and Greece. The external adjustment through a real depreciation following the external shock has been perverse: instead of a real depreciation these countries have experienced a real appreciation as unit labor costs have risen sharply relative to other Eurozone members and relative to the rest of the world. So, the external adjustment channel has not worked for these economies.
The reason is that wages and prices are sluggish and wage growth has outpaced productivity growth. This failure of wages to adjust to the loss of competitiveness reflects in part labor market institutions – in particular centralized wage bargaining – that prevents an adjustment of wages to lower productivity. Thus, since there is not enough price or wage flexibility, the adjustment of relative prices to an external shock is very slow and – at times – perverse. This rigidity of prices and wages has been found – in papers presented at the conference - to be related to a variety of structural characteristics of labor market institutions that are very slow to adjust.
The only country where the external adjustment has taken place has been Germany: but even in this case the adjustment was very slow and took more than a decade: the shock of German unification and the ensuing loss of competitiveness of Germany took more of a decade to be unraveled via a painful process of corporate restructuring and significant wage moderation.
A common monetary policy set by the ECB has not been always optimal not just because different Eurozone economies have had somewhat divergent business cycles. Rather, the very beginning of EMU led to very divergent real interest rates across the Eurozone that exacerbated growth differentials: the process of nominal interest rate convergence that the monetary union implied occurred while some economies had much higher inflation rates given different growth rates and business cycles. Thus, in the process leading to EMU real interest rates sharply fell in countries such as Spain, Portugal, Ireland and the Netherlands. This reduction in real interest rates led to an economic boom and overheating that – in the countries above – took also the form of a housing boom with excessive investment in real estate and sharply rising home prices. The ensuing economic bust in the Netherlands and Portugal was particularly painful.
The housing bust that is likely to now occur in Spain - and possibly even Ireland – will be similarly painful. Spain is a particular case of the growing divergences within EMU. Since 1999 its growth has been significantly higher than the Eurozone average but this high growth is mostly driven by an unsustainable housing boom – with residential construction now close to 19% of GDP - that is now at risk of ending into a severe bust. The high growth driven by housing has hidden the significant loss of competitiveness that the real appreciation of the exchange rate has entailed. Thus, the bust may be even more painful as Spain, even more than Italy or Portugal, has lost trade market shares to Asia and other trading partners.
While the competitiveness channel and the real interest rate channel did not lead to stabilization of asymmetric shocks, other channels of smoothing of such shock have also had limited role. Shocks to national output (GDP) can have smaller effects on income (GNP) if there is enough international portfolio diversification (capital market smoothing channel). Shocks to national income can have smaller effects on disposable income if fiscal channels provide shock absorption (fiscal channel); and shocks to disposable income will have smaller effects on private consumption if consumers can do international borrowing and lending (credit channel). But in the Eurozone these three channels of smoothing of idiosyncratic shocks work in a very limited way: while in the US such channels smooth about 75% of shock in the Eurozone smoothing is only 36% in the most recent EMU period. These smoothing channels are becoming more significant since the inception of EMU but still are relatively modest from a quantitative point of view.
The lack of fiscal federalism – as well as the fiscal constraints imposed by the Stability and Growth Pact – are the reasons why fiscal policy in the Eurozone has been pro-cyclical rather than counter-cyclical and why fiscal policy has not provided shock absorption of asymmetric shock. Some of the proposals in the conference – in particular that of a Eurozone-wide unemployment insurance scheme – would go in the right direction: they would make fiscal policy more countercyclical and would provide greater risk insurance of national shocks. But whether such proposals can be implemented in practice from a political economy point of view is still open to question.
In summary, the experience with EMU has been mixed. While growth divergence has not increased since 1999 there are a number of strain in this monetary union: asymmetric shocks, asset bubbles caused by common monetary policy, labor market rigidities, lack of adjustment of real exchange rates, limited role of risk insurance through financial channels, lack of risk sharing via federal fiscal policies have exacerbated the effect of asymmetric shocks that have hit different economies. At the same time structural reforms that would make economies more flexible and more able to respond to asymmetric disturbances have been implemented only very slowly. One serious open question if whether the countries – Italy, Portugal, Spain, Greece - that have experienced a significant loss of competitiveness and real exchange rate misalignment will be able implement reforms that will increase productivity growth, reduce relative unit labor costs and allow them to regain their lost competitiveness. There are some reasons to be cautiously pessimistic. Even ECB executive board member Lorenzo Bini-Smaghi - who strongly argued that EMU has been a success – admitted that these countries may have to go through a long and painful adjustment period to regain the lost competitiveness. |  |
