The challenges of joining EMU under price level convergence

By: John Lewis, De Nederlandsche Bank

 

Currently there are nine EU members from central and Eastern Europe who are obliged to adopt the Euro at some point in the future.   Most already meet the fiscal and interest rate criteria, or are within striking distance of doing so.  In general the far bigger challenge is meeting the exchange rate and price stability criteria simultaneously.  Inflation performance must be sustainable and no more than 1.5 percentage points above the average of the lowest three (positive) inflation rates in the EU; and the currency must have participated in ERM-II for two years without devaluation or severe tensions.

 

Countries in Central and Eastern Europe have price levels which are well below that of the Eurozone, but which are steadily rising as their economies converge in nominal and real terms to those of Western Europe. 

 

 

Table 1: Relative Price of Household Final Consumption Expenditure

Country	Price Level 2000 (Eurozone12=100)	Price Level 2005 (Eurozone12=100)
Bulgaria	30.4	41.5
Czech Republic	38.3	56.2
Estonia	39.4	61.9
Hungary	35.6	54.0
Latvia	28.5	52.1
Lithuania	40.9	60.7
Poland	43.2	58.1
Romania	N/A	51.3
Slovakia	37.0	54.6
Greece	78.3	84.6
Portugal	70.3	83.7
Slovenia	71.1	72.5

Source: Eurostat, Author’s own calculations

 

This process implies- for many years to come- either a positive inflation differential vis a vis the eurozone, a nominal exchange rate appreciation (or both); either of which could jeopardise compliance with the inflation and exchange rate criteria. 

 

Slovenia, the only one of the New Member States (NMS) which has managed to meet the entry criteria is an outlier from the other new member states both in terms of its much higher starting price level, and the levelling off of its price level at around 70% of the eurozone12.  Slovenia’s success has occurred in the absence of, rather than despite, nominal convergence.

 

Much has been written about the suitability of different NMS for EMU membership and on the appropriateness of the criteria.  The focus of this article is rather different.  Taking the criteria as given, I consider how the process of nominal convergence affects a country’s ability to meet the Maastricht criteria required for EMU entry should they wish to.  In particular, I consider whether convergence effects may pose a problem for compliance and if so, for how long they will be problematic, and how they interact with the choice of entry strategy.

 

So far we have seen two types of entry strategy adopted by the four ERM-II members from Central and Eastern Europe.  The Baltic states entered ERM-II under very tight exchange rate regimes:  Estonia and Lithuania have relatively orthodox euro-backed currency boards, Latvia has a very tight peg to the euro.  Surrendering monetary autonomy ensures the exchange rate goal is met, but the at the cost of being unable to use monetary policy to target domestic inflation. Deprived of the standard instruments of monetary policy, the monetary authorities must essentially sit back and hope that inflation is low enough. 

 

On the other hand Slovakia has joined ERM-II under a more flexible exchange rate regime combined with an inflation target.  Monetary policy is deployed to meet the inflation criterion, but policymakers must hope that the flexible exchange rate does not climb enough to breach the exchange rate criterion.

 

Whether or not these convergence effects are big enough to pose a problem depends on the size of these effects relative to the constraints posed by the Maastricht Criteria.  Although many authors have noted the problem posed by convergence effects, there has been much less attention devoted to computing the likely size of these effects and their consequences.

 

To answer these questions requires some model of the convergence process.  Predicting convergence is notoriously difficult, since there is substantial disagreement about how long convergence will take (empirical estimates vary between 25 and 50 years), and to what price level new member states will converge to (the process of convergence may stop short of full price equalisation across all countries). In a recent working paper (Hitting and Hoping? Meeting the Exchange Rate and Inflation Criteria During a Period of Nominal Convergence, DNB Working Paper 130), I consider four very different convergence scenarios.  In the first two I assume that countries converge to 100% of the EU price level in 25 and 50 years respectively.  In the second two, I estimate the convergence time based on the previous path of the relative price level. This is done for the case where the price level reaches 100% of the Eurozone, and for the case where the price level reaches only 80% of the eurozone (just below the level currently seen in Portugal, the least converged of the 12 original EMU members).  For each scenario the outcomes under both entry strategies- fixed exchange rate and inflation targeting- are considered.  What stands out is that the key results are strikingly robust across all four convergence scenarios.

 

In the fixed exchange rate case, the level of domestic inflation implied by the price convergence effect is substantially above any plausible value of the inflation criterion for all countries and will remain so for at least 15 years.  Of course in reality inflation is influenced by many other factors- and thus a country may for a few months experience inflation below the reference value.  But even in this case, such countries cannot expect to meet the inflation criterion consistently, because the underlying trend inflation implied by nominal convergence will be substantially higher than the reference value for a decade or two.  That means that such countries may have difficulties meeting the part of the criterion that requires sustainable price performance. 

 

If a country chooses the Slovak route of inflation targeting with the exchange rate allowed to float, the picture is rather different.  Assuming a 2% inflation target, price convergence effects imply an appreciation of no more than 10% over the course of two year stay in ERM-II.  An appreciation of this magnitude is not incompatible with a stay in ERM-II with an upper bound of 15% above the central parity.  Whether such a strong appreciation would be deemed to constitute “exchange rate stability” is a moot point, but the experience of Greece and Ireland (who were permitted to join at 3% and 3.5% above their initial central parity) suggests they might.  At the very least, it’s clear that the Maastricht Treaty is much stricter in forbidding devaluations than revaluations.

 

Although the Slovak koruna has appreciated strongly since joining ERM-II, this appreciation is more than could be explained by any of the convergence scenarios considered here.  That would suggest that in general, convergence effects are not big enough alone to generate exchange rate shifts outside the upper limit, or to explain all of Slovakia’s appreciation.  It also suggests that other inflation targetter’s currencies may appreciate by less than the Koruna has should they join ERM-II in the future.

 

Taken together, the key result is that the convergence criteria are not neutral with regard to the choice of exchange rate regime.  Exchange rate fixers have very little scope to accommodate price convergence via inflation, and hence find life much harder than inflation targetters who have some room to appreciate their nominal exchange rate in ERM-II. 

 

This can be seen through some simple “back of the envelope” calculations: consider an exchange rate fixer, seeking to get inflation below a reference value of 3% (the figure in the last convergence report).  Such a constraint implies that their real exchange rate appreciates by no more than 1% per year.  For an inflation targetter (whose central bank delivers the same inflation rate as the ECB), their real exchange rate appreciation would be equal to their nominal exchange rate appreciation.  An appreciation of say 10% over a two years stay in ERM-II, implies a real exchange rate appreciation of 5% per annum.  Thus, such a country could stay within the bounds of ERM-II, meet the inflation criterion and have around five times the real exchange rate appreciation than is permitted to the exchange rate fixer.  If the country used the full 15% upper margin, that would imply even greater room to manoeuvre.

 

Lithuania and Estonia had initially hoped to join in 2007, and have been steadily putting back their putative entry dates.  On the basis of any of the convergence scenarios considered here, the wait may be more than a decade.  The framework used here assumes that inflation converges linearly to the Eurozone level.  That implies that the current trend rate of inflation, arising from price convergence considerations should be around 5-6% per annum.  It will be at least 15 years before the convergence effect generates inflation of less than 3%.  Until such time, the Baltics will have to hope for a large downward deviation in inflation from its trend (and a favourable assessment of sustainability) in order to meet the Maastricht criteria.

 

One scenario would be that price levels halt at around 70% of the Eurozone price level, as they did in the case of Slovenia.  But in general Slovenia appears to be an exceptional case in that its price level process ground to a complete halt around five years ago.  A more likely scenario is arguably that Baltic price levels converge to those seen in nearby Scandinavia- which are well above the Eurozone average.  Alternatively, the Baltics could opt to deploy their remaining policy instrument- fiscal policy- to bring down inflation.  Essentially this would require either inflicting a temporary contraction in demand by tightening fiscal stance, or by reducing indirect taxes to effect a fall in inflation.  Neither strategy would deliver a long-lived reduction in inflation however, casting doubt on whether the sustainability element of the criterion could be met by such a strategy.

 

Looking across other member states yet to join ERM-II, most other central European states have some kind of inflation targeting setup.  Assuming they persist with this and adopt a Slovak style entry strategy, this yields an interesting conclusion about the sequencing of Eurozone expansion.  Although the Baltic states are often seen as being at the front of the queue to join, it is possible that they could be leapfrogged by Central European countries yet to enter ERM-II, who can use the flexibility provided by nominal exchange rate appreciations to meet the criteria sooner, should they wish to.

 

 

The views expressed are the personal views of the author and are not necessarily those of De Nederlandsche Bank.

 

 

 

---