Re-engineering EMU

Re-engineering EMU

Professor Avinash Persaud

Talking about European Economic and Monetary Union is the surest way to lose friends and make enemies. It is a deeply religious issue that resonates less with the cold calculus of optimal currency analysis, and more with that intangible, emotional essence, that is identity and patriotism.

There are believers and non-believers in EMU and never the twain meet. In my experience one of the few ways to progress a discussion about EMU is to try, as far as possible, to sidestep the more religious issues and focus on the seemingly technical, in the hope that while this may not convert anyone, it will clarify what’s at stake. Opacity also disarms the zealots.

So the question I pose today is not whether EMU is good or bad, not whether it will last or not, nor whether the UK should join or not, but whether there are institutional changes that would make it function better. Of course these issues are not entirely separate. The right institutional framework would support EMU’s economic success, its longevity, and the prospect of British membership. The wrong institutional framework would do the opposite.

Time for change

This is an apt issue today. We have the first real opportunity for institutional change in the Eurozone, since the single currency arrived in January 1999. There is now sufficient distance from that start for changes to be seen as adjustments along the way, rather than an attempt to abort the journey. Perhaps too, the prospect of British entry to the single currency, could act as a catalyst for change. Europe will be favourably disposed to UK membership, as it will take EMU past the tipping point. In terms of size and international trade, Europe would go from being the second largest economic bloc, to the first, which would serve its broader aspirations.

Some changes are likely. There will be a new ECB President by the end of this year and this change of leadership provides an easy opportunity for some changes in operation. The Bank has already initiated a study on how it may improve its impact – beyond its constitutional mandate to pursue price stability, however defined, nothing is ruled out. Earlier, the European Commission proposed changes to the Stability and Growth Pact, in part to diffuse the imbroglio surrounding Commission President Prodi’s now famous remark, that the Pact is “stupido”. And change is due. It is inconceivable that a project, so enormous and ambitious in scope, would be planned with such precise foresight, that no improvements could be thought of several years after the plan was conceived.

More subversively, the institutional arrangements required for EMU to be launched in the first place, to be politically acceptable, would only be the same arrangements required for EMU to function smoothly, by good fortune or lucky coincidence. Let me dwell on this point further. The benefits from a single currency for most European countries were clear. Removing exchange rate volatility for relatively small economies, in most cases more heavily exposed to exchange rate changes than the British economy, was an end in itself, but this would also remove the risk premium that held real interest rates so high, holding back productive investment.

Chart 1: Interest rate convergence in Europe, 1991-1999.

This picture, which will bring warm feelings to any bond market veteran, shows the wide spread of interest rates in Europe and their later convergence as prospects for EMU rose. When coupled with a relatively cheap conversion exchange rate, this reduction in interest rates delivered a long investment boom in Europe’s previously risky periphery of Spain, Portugal, Finland and Ireland.

Chart 2: Investment growth in Spain, Portugal, Finland, Ireland and Germany

Here we compare the annual growth in business investment in the four countries, which joined the single currency in 1999 with competitive exchange rates and also the German record, the black and shortest bars. Our fixation with Europe’s largest economy, Germany, shapes our perceptions of the Eurozone too much.

What was always less clear was what Germany would gain from EMU, at least in economic terms if not politically. Indeed, being the country with the greatest anti-inflation credibility, it invariably had the lowest interest rates and would therefore have the most to lose if interest rates converged to the average level in Europe. If Germany was to accept EMU, convergence had to be based on Germany’s anti-inflation credibility. It seemed logical at the time that the best way to achieve this was to model Europe’s new institutions on German institutions, on an image of Teutonic discipline, and in case anyone missed the point, to locate the ECB in Frankfurt. This helped to win German public support for EMU, though this support was never tested.

It should be pointed out that a further attractive feature for German industry at the time was that EMU would knock the habit of other countries countering Germany’s superior competitiveness with frequent and disruptive devaluations. Be careful what you wish for, today it is German industry that could do with a devaluation.

But the critical point is that modelling the ECB on the Bundesbank and the Stability & Growth Pact was principally about pinning Europe’s anti-inflation credibility to Germany’s, and winning German support for the project. It was less about how it would all work, a couple of economic cycles and rounds of enlargement later. This is perhaps most apparent with the Pact.

The Pact was designed to deal with the free-rider problem that arises when countries share a currency. Before the arrival of the common currency, the market made a clear differentiation between countries based on many factors.

Chart 3: 10 year European government bond yields, 1993-1995

In the pre-convergence days, the main drivers of this differentiation were inflation, deficits and debt as this picture of different long-term interest rates in Europe before 1995 and different levels of government deficit on the right hand sideshows. This is market discipline at work.

EMU is a whole new ball game. The inflation and credit risks caused by the fiscal largesse of any single Member State would be shared by every Member State; this creates an incentive for Member States to increase their deficits. They receive the benefits of higher spending, but share the interest rate costs. To some extent this is what happened in Argentina and Brazil where the currency was brought down by the profligacy of regional governments.

Chart 4: Text Chart – The free-rider problem

To deal with the free-rider problem, a number of rules were put in place. Price stability is a stated policy objective of the ECB, and it is not permitted to monetise the debts of any Member State or bail out them out. This is designed to raise the interest rate penalty for spendthrift governments, but it would only do so if the markets were convinced that a Member State in trouble would not be bailed out by the others. The markets are unlikely to be convinced of this. Would Member States really stand by and see Italy default with all the widespread and costly disruption to the financial system that this would cause. The economic activity of national governments is far more pervasive than regional governments. And once the market began to consider these economic costs of default, would it not threaten a contagion of defaults? These concerns are legitimate. Similar arguments were put forward by Alexander Lamfalussy back in the mid-1990s. To limit the possibility that a country would need to be bailed out, it was agreed in the Stability and Growth Pact that budget deficits would be held below 3.0% GDP. The Pact is enforced through a range of mutual surveillance measures, public finger pointing and ultimately, hefty fines.

Many commentators however share the view of Commission President Romano Prodi that it is a stupid rule that, on the threat of an expensive fine, forces European governments to cut expenditure into the down cycle of the economy in order to keep budget deficits below a fairly arbitrary level. It is like trying to limit bankruptcies by fining bankrupts.

Additionally, the Pact in conjunction with the ECB’s remit has given the impression that nobody owns the objective of output stability in the Eurozone. This powerlessness in the face of an economic slowdown stands in sharp contrast to the policy activism on the other side of the pond. Some believe that the euro’s weakness in 1999 and 2000 was due to the market’s negative judgement on this policy framework. If that were so it would be a good reason for change. The litmus test of any fiscal arrangement is whether it is credible with the markets and will reduce investors’ sense of risk and uncertainty and the risk premia in the structure of interest rates. Lower more stable interest rates should facilitate investment and growth.

Measuring credibility and risk premia is not straightforward. It is seldom revealed directly. In general, however, we would expect to see four things when risk premia fall.

Chart 5: Text chart – Proxies of risk premia

First, we would expect the yields a government has to pay to attract funding, over and above the inflation rate, to fall and to fall more than in other countries. We would expect lower relative real yields. One source of differences in real yields is the supply of government bonds and so one measure of institutional credibility is whether real yields are falling at a faster pace than a decline in budget deficits or are holding still as deficits rise markedly. Third, we would expect that the premia a government has to pay when borrowing over a long period, compared to borrowing over a short period, to fall. The yield curve would flatten. Fourth, we would expect that for given economic risks, falling risk premia would lead investors to demand more bonds, currency and equities leading to a rare positive co-movement of these three instruments.

While each of these measures has faults, using them in conjunction should be reliable. Doing so yields a surprise. The ratification of the Stability Pact lowered risk premia in the Eurozone. Between the decade prior to the ratification of the Stability Pact in 1997 and the five years afterwards, the average level of real yields in the Eurozone fell significantly.

Chart 6: Real bond yields and government deficits/1

If we compare the average level of real yields before 1997 and after in the US and Eurozone we will see that real yields are falling generally but that they fell more in the Eurozone and more relative to changes in budgets. That is indicated by the steeper line.

My colleague Michael Metcalfe carried out most of the empirical work I am showing you today and another colleague Chris McCoy prepared these charts.

Chart 7: Real bond yields and government deficits/2

Here we have plotted the annual government budget deficit and the level of real long term interest rates for the US and Eurozone over the past 17 years. There is a broad relationship between the two with the points drifting from top left to bottom right and higher budget deficits corresponding with higher real yields. Points above the average line are likely to represent years when the risk premia is higher than average and points below the line where the risk premium is lower. Since 1997 when the Pact was signed, all of the US points lie above the line and all of the Eurozone points lie below the line.

Reinforcing this picture of falling risk premia, the Eurozone yield curve also flattened significantly between the five years before 1997 and the five years after.

Chart 8: Eurozone, real yield curves, before and after the adoption of the SGP

This suggests that the Stability Pact holds more credibility that imagined. That the market recognises the free-rider problem. That abandonment is likely to be greeted negatively by the market with higher real yields and a steeper yield curve. But this does not exclude alternatives or adjustment to the Pact. Let us try and be more precise about what the market wants.

Over the past two years there have been over 400 stories on the newswires and in newspapers relating to the Stability & Growth Pact being breached.

Chart 9: Volume of stories referencing a breach of the SGP across time, 1999 – 2002.

This chart shows the spread of these stories over time. Following other studies of the market reaction to news including Charles’ 1998 paper on foreign exchange markets with Richard Payne, it may be possible to reveal the nuances of the market view from its reaction to news. If the market strongly believed that the current Stability Pact is what is needed and that there is no alternative then we should find that estimates of the risk premium rise with stories about a breach of the Pact. If the markets wanted more pro-active policies then news of a breach caused by a tax cut say, may push down the risk premia or at least push it up less than a breach caused by a slowing of growth.

To conduct this study we had to streamline the data. Mr. Prodi’s famous remark was reported in over a dozen stories over several days, but it is really one story. We removed duplicate stories and used the stories from one major news service, Reuters, to help in the consistency of dating stories. We attributed a day to each story and measured the market reaction from the previous day’s New York closing prices to the current day’s close. We went from 400 stories over two years to 81 separate stories. Running a regression analysis we found that there was no statistically significant relationship between a day in which there has been a news story on a breach of the Pact and our four measures of risk premia. But this did not mean that these stories did not have impact. In just under half of the stories, at least three out of four measures of risk premium reacted in the same direction.

On examining these stories individually we found an unexpected and surprising pattern. In the stories on a breach of a stability pact where the risk premium fell, invariably, a senior minister is quoted reaffirming their commitment to the Pact as in the case of German Chancellor Schroeder on 10 February last year. Sometimes verbal commitment is reinforced with talk of a tax hike as in the case of a rise in VAT in Portugal. Most of the negative stories concern speculation of a loosening of the Pact. As an extreme example, the euro weakened by over 1.0% on the day of Commission President Prodi’s “Stupid” quote.

There were exceptions to this rule. There is the odd positive story where measures to boost growth will threaten the deficit limits, as in the stories on 9 August 2002 surrounding the German labour reform plan. There are also a few stories which point to how the Pact is reducing policy flexibility or “cornering” policy-makers especially in the run-up to EcoFin meetings.

I am mindful of the fact that markets reaction to these stories are not consistent, that there is an element of interpretation of what these stories are saying and that we have ended up analysing a subset of the 400 stories, yet these results have significance in supporting the earlier analysis and being close to the opposite of the consensus view.

Two separate studies have reached similar conclusions. The market is concerned about the free-rider problem. The Stability Pact has served its purpose in reassuring the markets on this point. In the second study it would appear that the market is concerned about attempts to make the pact looser when the limits are being threatened. The Commission’s recent proposals have an element of this about them. It smacks of moving the goal posts to where the ball is and the market does not like that. It may be that greater discretion and more use of market discipline would make for better policies, but the evidence is not there.

At the same time there is market concern about a reduction in policy freedom especially at times of economic or financial distress. An amendment that increased policy flexibility, but at the same time tightened fiscal discipline and its enforcement may be greeted positively. This would not be easy to craft, but nor would it be impossible.

My preferred solution would be to add to the existing Pact, that in assessing whether a country had an excessive deficit, consideration would be given to whether it is on track to achieve a balanced budget across the economic cycle. An independent committee of experts could date the cycle and countries would be expected to have a balanced budget from the end of the last recession through to the end of the current recession. An assessment of whether a country had an excessive deficit would then be carried out when the recession is over and growth is picking up and when it would be easier for national governments to tighten their fiscal belts and when, if necessary a fine on those that didn’t tighten their fiscal belts would be more credible. The principle here is that making sure that countries tighten their fiscal belts at the top of the cycle is as effective at keeping deficits in check at the bottom of a cycle, but is far more sympathetic to the economic and political cycle and more enforceable. This would be a modest, yet significant change to the current arrangements.

Chart 10: Text chart – A modest proposal

Let us finally turn to monetary policy. Our analysis suggests that the ECB has had some bad press. After all inflation is on target. Real yields have fallen. Francesco Giavazzi and others have shown that ECB interest rates have not been far from those suggested by a standard Taylor Rule. Growth is weak but then monetary policy has been easy if the exchange rate is taken into account and the inflation rate. Yet there is a palpable sense of incoherency between the ECB’s stated goals, targets and what it ends up doing which is not helping its credibility. For the world’s most independent central bank there is also a surprising degree of political pressure.

The problem is that the Bundesbank succeeded not because of what was on the statute books. It succeeded because of how it operated within Germany’s particular historical, economic and political context. Shifting these legal statutes to a different context may have won initial support for EMU from Germany, but it was no guarantee of a repeat of the Bundesbank’s record and nor was it particularly appropriate to the very different context of Europe.

Enlargement, especially eastwards to countries that are catching up economically, brings with it transition problems that will require more flexibility than the Bundesbank required. More flexibility requires more independence and more independence as Europe gets larger and the physical, cultural, national distances between central bankers and citizens grows, requires more legitimacy. The Bundesbank’s legitimacy was strong, but it came from Germany’s history of hyperinflation between the wars, not from its statutes. To achieve more flexibility, credibility and legitimacy the ECB could do worse than to adopt certain features of the new monetary arrangements in the UK.

The Bank of England has less independence than the ECB on paper, but is freer from political pressure. Under the British system the government chooses a symmetrical inflation target and the Bank has operational independence on how to achieve it. The government’s choice of target brings democratic legitimacy and government support for the Bank’s operational independence. Operational independence is made credible by there being a single final target, the appointment of outside experts on fixed terms and a strong degree of transparency. Flexibility comes from not just from the symmetrical nature of the target, but from the open letter system which allows the Bank to set out, in a clear and transparent way, its strategy for getting inflation back towards its target when it has deviated significantly. It is conceivable that faced with a substantial supply-side shock the bank sets out a gradual adjustment, but by doing so transparently and according to a clear framework, this would not undermine the Bank’s credibility.

Using our same methodology of four measures of risk premia we find unambiguous evidence that the move to operational independence of the Bank of England raised the credibility of monetary policy and reduced sterling’s risk premia.

Chart 11: Budget deficits and real yields cross-section, including the UK.

The decline in real yields in the UK is greater than occurred in the United States – the angle of the line is sharper – though it is parallel to the improvement in Euroland.

Chart 12: Budget deficits and real yields through time, including the UK.

If we were to add UK data to our previous chart of deficits and real long-term interest rates we would find that since 1997, the UK points are almost all below the line suggesting that since then, the UK has had to pay less to fund its self per unit of deficit

Chart 13: UK yield curve before and after 1997

And the UK yield curve has flattened.

The features of the UK arrangements could be easily given a European dimension. Within the overall objective of achieving price stability, allowing the European parliament to set the ECB a symmetrical target will bring legitimacy to both bodies and may better take into account the changing nature of the Eurozone as it expands. A small interest rate setting committee made up of independent experts, ECB executives and a rotation of national central bank governors could reduce voting along national lines and keep the committee to a manageable size.

There is a wonderful little paper by the Belgian economist Paul De Grauwe called Fundamentals following market prices. It is, remember, supposed to be the other way around, with prices following the fundamentals. But the reality of too much research is that we find the fundamentals to fit the price action afterwards. Which leads to much importance played to a host of coincidences. The decline in the euro in 1999 and 2000 dredged up many a fundamental reason why the euro was going down and would always do so. If the Eurozone’s policies were not right, then the policy framework was at fault or worse. Our analysis of risk premia in Europe does not substantiate this view. The Eurozone has anti-inflation credibility as evidenced by historically and relatively low real interest rates. Perhaps the reality was that the economy was weak and as activist policy was constrained, it was entirely appropriate for the euro to weaken. Given the economic conditions a strong euro would have been bizarre, a sure sign of something wrong with the framework, and it would have worsened the economic downturn.

But that is not to say that everything is right in the rose garden, especially as it was designed to lure people in, without much attention to the experience when they got there. It is far from ideal to have the exchange rate do all the adjustment to the economic cycle. The Stability and Growth Pact may not need dismantling, but it needs greater consideration for the economic cycle, without providing too much opportunity for shifting the goal posts or cheating. It seems to me that this is possible to achieve by concentrating on the deficit across the economic cycle and critically making the assessment during the boom, not the recession. In monetary policy the Eurozone is a very different context to post-war Germany. What is required is more flexibility and legitimacy without loss of credibility. A dose perhaps of constrained discretion, this is the phrase used to describe the UK monetary arrangements by one of its chief architects, Ed Balls.

It seems slightly odd to me that we now have in the UK, institutions in the economic sphere that deserve some emulation internationally; though it seems appropriate, that this should be so because our institutions are neither too rigid nor allow too much discretion, but reflect a pragmatic balance.

© Avinash Persaud, 3 October 2002