October 14, 2006
Examining Divergences in the Eurozone: Red alert for Portugal and Spain, some relief for Italy
One of the hottest topics in the debate on Eurozone economics over the past months has been the question whether dangerous divergences are building between the 12 countries which share the euro. In fact, the fear that some countries might have lost competitiveness to a dangerous degree has been behind scenarios of countries leaving EMU (there has been a lively debate in the blogosphere and on Eurozone Watch - see here and here and here).
For a conference in Berlin , Ulrich Fritsche (from the Hamburg University) and the German Insitute for Economic Research) and I have tried to evaluate how bad divergences really are.
To this end, we have constructed an index for unit labour cost in international currency relative to the rest of EMU (100=1998). Our thoughts behind chosing unit labour costs has been that consumer price inflation might easily be distorted because of changes from direct to indirect taxation, changes in the oil price or (with regard to fresh food items) fluctuations in the harvest. Moreover, especially in small countries which are to a large extent integrated into the European economy, a loss of competitiveness might not show up to its full extent in the CPI if consumer prices are mostly determined by the neighbouring countries’ price level. So, arguably, unit labour costs in euro should be the best measure available for one country’s cost competitiveness.
In order to gauge whether divergences are benign or a reason of concern, we have choosen two yardsticks: First, we have compared the current unit labour cost position with historical precedents. Second, we have compared the deviation for each country from its long-term average with fluctuations in regional unit labour costs between the 50 states of the U.S., the census regions of the U.S. and the 11 (west German) Länder of the Federal Republic of Germany. The idea behind this was that Germany and the U.S. have provided examples of working currency unions and deviations in excess of those observed in these countries might be reason for concern.
The results are quite interesting: Portugal and Greece show a competitive position today which is worse than at anytime in the past 25 years - and worse than anything ever experienced in any U.S. state or any German Land.
The development in Portugal is especially awkward as the country has experienced a very dismal growth performance since 2001 with GDP hardly growing at all and unemployment having almost doubled. Yet, unit labour costs continue to increase faster than in the rest of Europe, indicating that wages are not reacting significantly to the dire economic conditions and the apparent overvaluation of the small country.
The situation for Spain looks slightly better: Though the country’s unit labour cost position is now close to its historical high just prior to the EMS crisis in 1992 in which the peseta was forced to devalue sharply, it is still within the range which we have observed within the U.S. However, competitiveness has deteriorated more than in any German Land.
The big surprise, however, comes from Italy: According to our computations, competitiveness is still rather close to its long term average and well within the bounds experienced by the German Länder or the U.S. states. Moreover, competitiveness is still way below what it was prior to the EMS crisis in 1992.
This might indicate that Italy’s loss of competitiveness has not been as bad as generally thought, making it less of an exit candidate of EMU than previously thought. This would also be in line with the fact that Italy’s current account deficit is rather benign when compared to those of Portugal or Spain (OECD projection for Italy: 2.1 percent of GDP; Spain: 8.6 percent of GDP; Portugal: 9.6 percent of GDP).
The reason that Italy’s real appreciation is often misjudged might have to do with the fact that the EU commission often uses the base year 1995 for its data, the year in which Italy’s unit labour cost position reached its quarter-century low. At that time, the lira had devalued sharply in nominal terms. In the years following, it regained some of its value before the country entered EMU. If we thus just compare the loss of competitiveness since 1995, it looks as if Italy had lost competitiveness to a similar degree as Spain or Portugal. If we take the development in a longer perspective, we probably have to admit that some of the real appreciation after 1995 was just a partial correction of a prior overshooting of the nominal exchange rate.
Those of you who are interested in the details can access our discussion paper at the Hamburg university’s website via EconPapers.
“that some countries might have lost competitiveness to a dangerous degree has been behind scenarios of countries leaving EMU”
Just for the record this has never been the brunt of my argument about Italy. The issue I have been arguing is that Italy is unable to raise trend growth sufficiently - to say something over 2% - to be able to make the dynamics of sustaining the debt workable, in particular giving the heavy load on public finances which is looming in the none too distant future.
Thus the essential argument is that default is unthinkable inside the eurozone, default is more or less inevitable (factoring in the inability to develop a solid political consensus to change things), therefore this really only has one end point. The only question in my book is when.
This is what it means to say that public finances are ‘unsustainable’, an expression which is oft used but seldom thought about.
The data you have looked at are interesting, but again you can see that Italy crossed the long term average threshold in 2003 and unit costs have continued to rise. This is not a reflection of an economy which is in a significant transition for the better. Quite the contrary. The fact that Greece, Portugal and Spain are worse is hardly encouraging.
The Spanish situation is a complex one, and I have long thought that Spain’s future to some extent depends on Italy: ie if Italy defaults, and there is one hell of a mess, then the endebtedness of the Spanish individual (rather than the state) is going to become a very serious matter. The Spanish banking system will take the hit.
Greece’s position is certainly no better than Italy’s , but obviously it is much smaller, and the global implications of melt down in Greece would be nothing like so severe. It is precisely the capacity for producing chaos elsewhere which makes Italy so important.
And ditto for poor little Portugal, which has been struggling quietly in the background as you say since around 2000.And again as you say, getting virtually no growth. And it is again this that worries me in the Italian context. If Italy follows the Portuguese path of attempting to restore fiscal order by knocking a couple of points off trend growth, then this would put Italy at a growth rate of minus one percent, which suddenly becomes clearly unsustainable, and then bang, just watch the lights go out.
More quibbles: you are looking at individual productivity (or one measure of this), but you also need to think in terms of societal productivity, and to do this you need to take some measure of the percentage of the population not working (the dependancy ratio or carrying load) and how this evolves going forward.
Also you need to think about the value distribution of all this. Aggregate data is fine for some things, but you really need to drill down to the sector level to get a clearer picture. One of the suggestions is that - especially following the recent massive legalisation of migrants - Italy may now be extremely price competitive in low value work, but this is not where we want or need to be. One theory has always been that a relatively smaller workforce, doing relatively more valuable work could balance the books. I have never really bought this argument because of the stocks and flows issues (see below) but I certainly think there is no harm in trying.
So you need to think about the important growth sectors and relative productivity in those sectors, since to survive Italy needs to shift activity continuously up the chain.
This is where we hit the stocks and flows thing, since with a median age of 44 it is much harder to rapidly raise the average education level of the population, which is precisely what you need to do to survive.
Dear Edward,
I was not referring to you with the comment that some people are constructing their case for an Italian exit from EMU around the loss of competitiveness. Both the analysis of Deutsche Bank of an “EMU trap” as well as that of the Société Générale make this point.
However, you are completely correct that Italy DOES have significant problems and that it might be little case for encouragement than some other countries are better off.
Still, the amount of misvaluation relative to the long run average proposes that Italy’s competitiveness problems might be easier to resolve than Portugal’s and Spain’s and that the main problem in Italy really is the public debt level.
In addition, I believe that very few people actually have understood the extent of the problem Spain will face once the construction boom ends.
Best, Sebastian
Bingo Sebastian.
I think you ticked all the boxes here.
Nothing like taking another, and closer, look at the data !!
Good work.
A report worth looking at by Morgan Stanley on Italy:
LINK
To quote:
” Since the inception of EMU, Italian GDP has increased by 1.33% per year versus 1.31% for Germany. I wonder which economy was the sickest. Italy’s trade balance was in perfect equilibrium in 2005, with the non-oil balance posting a hefty 2.4% GDP surplus. In contrast, Britain’s and Spain’s trade deficits were, respectively, 5.6% and 7.6% of GDP last year. Last, Italy’s harmonised unemployment rate dropped from 11.4% just before EMU started, to 7.4% on the latest reading, only two points above the UK level. This is a much larger decline than in France or Germany and comes second only to Spain. “
Since the link doesn’t work here is the entire report:
The downgrade of Italy by two credit rating agencies is likely to re-invigorate a theory that has some popularity in the financial markets, namely that Italy could well be the first casualty of an ill-conceived monetary union. The rationale: Italy has suffered from a huge loss of competitiveness, its productivity is at a standstill and its public debt is running out of control. Real life has demonstrated that the country cannot compete with her neighbours on a level playing field, i.e., not without the repeated ‘shots in the arm’ provided by devaluations. According to this line of reasoning, investors should ultimately impose a higher risk premium and, as debt servicing becomes more costly, their pessimism would turn into a self-fulfilling prophecy: Italians would start dreaming of a 1992-style devaluation as the only way to bail out their flagging economy, at the expense, of course, of their main trading partners.
I am afraid that the script of this horror movie is based on flawed macroeconomic analysis and poor use of dubious statistics. On the contrary, I believe that Italy is on the mend and likely to attract considerably more foreign capital than it has done in the past.
First, three key hard facts do not square with this ‘Italy bashing’. Since the inception of EMU, Italian GDP has increased by 1.33% per year versus 1.31% for Germany. I wonder which economy was the sickest. Italy’s trade balance was in perfect equilibrium in 2005, with the non-oil balance posting a hefty 2.4% GDP surplus. In contrast, Britain’s and Spain’s trade deficits were, respectively, 5.6% and 7.6% of GDP last year. Last, Italy’s harmonised unemployment rate dropped from 11.4% just before EMU started, to 7.4% on the latest reading, only two points above the UK level. This is a much larger decline than in France or Germany and comes second only to Spain. Italy bashers, who are at pains to explain how a lame duck could manage to reduce unemployment without fuelling wage inflation, may question the quality of unemployment statistics. However, these measures are based on large-sample household surveys and are therefore more reliable and consistent than claimant counts, which are reliant on national unemployment insurance systems. Nevertheless, they have some idiosyncratic features on which I will comment later.
So why are so many analysts convinced that Italy has been priced out of the game by super-competitive Germany? Because they look at measures of competitiveness such as unit labour costs or export volumes. On these criteria, the Italian situation looks truly desperate. Take, for example, the relative unit labour costs in the manufacturing sector calculated by the EU Commission. On this measure, Italy’s competitive position relative to its euro area partners has lost 23.5% since 1999 while Germany’s position has improved by 17.6%. There is worse to come: on the OECD measure of real export market performance, Italy has lost 27 points of market share since 1999, twice as much as France, while Germany has gained 4.1 points. Clearly, something must be wrong: how could an economy suffering enormous losses of competitiveness boast a falling unemployment rate and a balanced trade account?
In my view, both unit labour costs and export volumes are tainted by serious statistical uncertainties. Starting with unit labour costs, the weak link is not the measure of costs, but that of productivity, a notorious headache for statisticians. On data gathered by the US Bureau of Labour Statistics, Italian hourly productivity in manufacturing was the same in 2005 as it was in 1999, while it increased by 27% in Germany and France over the same period. I find it hard to believe that Italian hourly productivity has really stagnated for six years, and suspect that a combination of large-scale regularisation of illegal immigrants and tax incentives for employers to hire employees who were working without being recorded in payrolls have distorted productivity data. In short, hundreds of thousands of workers in Italy have moved from the black to the legal economy, artificially bringing down productivity data and probably flattering unemployment numbers as well: It is likely that black economy workers answered ‘No’ when asked the question ‘Are you working?’ in the household survey, and now say ‘Yes’ if they have been hired legally in the meantime.
In fact, since the output of the underground economy is included in GDP as measured by ISTAT, it is fair to say that both past productivity and past unemployment levels were artificially inflated and that current data are closer to reality. From this angle, the fact that a four-point cut in the unemployment rate did not fuel wage inflation becomes less intriguing: in reality, unemployment has declined less than official measures suggest. Back to productivity, the rate of growth was underestimated because of these changes in the structure of the labour market. In the real world, hourly labour productivity has probably increased in Italy about as fast as in the average of mature economies, as a result of technological progress and capital deepening. The problem is that we do not know exactly at what rate.
Turning to exports, here is the conundrum: standard indicators say that Italy’s competitiveness has been seriously eroded and, yet, Italian exports are doing well, compared with its peers. On OECD data, Italy’s market share of global trade, measured in current dollars, was 3.7% in 2005 versus 4.1% in 1999, a 10% decline. The OECD bloc suffered from the same loss — its market share dropping from 74.5% to 66.9% over the same period, in favour of China and oil-exporting countries. A more relevant measure of Italy’s export performance is to compare its nominal exports with those of the euro area. On this yardstick, the only one that really makes sense in a currency union, Italy actually outperformed its peers: Italian nominal exports have increased by 4% more than the average of EMU country exports since the inception of the monetary union. Here, I believe that the statistical flaw is in the measure of prices. In reality, for lack of a direct measure, ISTAT is using unit values, i.e., export values divided by quantities. When the euro shot up, in 2002-03, a very counterintuitive fact was that Italian export ‘prices’ (in fact unit values) increased. In my view, this point, which escaped analysts’ attention, reveals the underlying and hidden truth: Italian producers have reacted to globalisation by off-shoring production centres to low-cost countries such as Romania and Tunisia and, more importantly, by upgrading their product mix towards more expensive products, in the fashion sector to take a well-known example. Here again, the problem is that there is convincing indirect evidence, such as the export performance, that corporate Italy has changed for the better, but no direct and reliable statistics.
Don’t get me wrong. I’m not saying that Italy is the new corporate Eldorado or the next stellar macro performer in Europe. The public debt inherited from the past is a permanent sword of Damocles. Labour laws, especially redundancy rules, are rigid, corporate governance is weak, the internal market is inefficient, especially for services, the commercial property market is opaque and in the hands of a small group of investors, bureaucrats and their love affair with ‘documenti’ are not particularly business-friendly, the population is ageing, pension reform is still ‘a work in progress’ and healthcare reform is still in limbo. And yet, at the margins, Italy might be the place where change is taking place at the greatest speed. In this regard, Mr. Prodi’s strategy is the right one, I believe: commitment to budget consolidation, reduction in non-wage labour costs in order to boost employment and reduce further structural unemployment, and supply-side reforms such as the liberalisation of services. In fact, I would have a wager that the credit rating agencies may sooner rather than later have to re-evaluate the Italian case in a more positive light, given faster potential growth and lower budget deficits. Investors should not wait until then.
Edward got me the link (which works):
http://www.morganstanley.com/GEFdata/digests/20061025-wed.html
[...] Mr. Regling first pointed out that the competitiveness lost both by the wage increases in the reunification boom as well as by the nominal appreciation in the EMS crisis has been regained. “German competitiveness is now back at its 1989 level”, he said. Interestingly, this is pretty much in line with my paper jointly with Ulrich Fritsche (see also my post on the paper). Uli and me also come to the conclusion that the German price competitiveness measured in unit labour cost relative to the rest of EMU is as favorable as never before since the early 1980s: [...]
Another argument in favour of Italy is it’s encouraging export trends. Apparently the economic stagnation of the last five years was related to a re-structuring process of small and medium enterprises which are today able to provide high-quality products at competitive prices. Exports growth is the basis of GDP surge and positively ending trade balance. In the long term there is also reasons to foresee a further expansion of Italian exports due to growing new economies (Asia) which are seeking for luxuries goods and new consumption patterns.