Sharon Bowles MEP

Speech to the ECB Conference on Statistics

Good evening, thank you for inviting me.

Maybe I should start by saying I have a mixed relationship with statistics. I am not one of those who think statistics can be made to say anything you want, because in my engineering research days they didn't always say what I wanted! I can also tell you that as a politician they do not always say what you want either!

The financial crisis, the sovereign debt aspect and economic governance has made statistics sexy - now everyone wants them - not just sexy ones (figuratively speaking) but there is increased demand for all kinds of data. The financial crisis has had an impact on European Statistics, for improvements in knowledge of new products and services from housing markets to communication of statistical information, for quality in terms of both reliability and speed and new regulatory procedures such as the bank stress tests. Some of this begs the question what are we going to do with all the data to make sure we get useful and practical outcomes, which I will come to in due course.

Before that I will summarise the statistics issues related to the crisis that have come to the Parliament in the five years that I have been there and our responses.

First we had a national statistics crisis in 2005 where the Council shied away from granting Eurostat more powers, which had been proposed by the Commission and supported by the Parliament. Indeed, we did not just support the Commission on this we positively urged them towards it, we were disappointed with the outcome.

However, hat did follow from that episode was the creation of a code of conduct and peer review process for the EU's national statistical offices. The European Statistical Governance Advisory Board was also set up, for which I was the rapporteur and negotiated with Walter Radermacher during the German Presidency, although it was finally finishing off under the Portuguese Presidency in October 2007. I didn't get everything I wanted, especially with regard to more intrusive checking of Member States upstream data, because the Member States would not accept it - back then we delicately called it 'sensitivity'. Now he is in charge at Eurostat I think maybe Walter has a softer spot for my early endeavours towards a better statistical audit, and anyway I have not forgotten that there is a three year review clause.

Then in June this year the sovereign debt crisis brought the EU back to the issue of increasing Eurostat powers and Parliament was not backward in its statements: it was the first litmus test to see whether Member States would act on their words. We suggested that Eurostat should be given 'dawn raid' powers like the Commission's competition services. Statistics really were getting sexier - quality was no longer a boring issue but a really vital necessity.

I have checked the Economic and Monetary Affairs committee's first exchange on this June's proposals and I said "Accurate statistics and improved verification of the reliability of aggregate data provided to Eurostat are essential prerequisites for any improvements of economic governance in the Union. The Commission proposal is the minimum needed as a result of the lessons of the Greek case, and reporting obligations need to be enforced among all Member States. This reporting should include details of any off balance sheet activity." I am pretty happy to stand by that today.

This week in Strasbourg we will take the final vote on the Feio Report, which is a legislative initiative report, and is our contribution to the pool of ideas for Economic Governance that has been prepared in parallel with the Commission proposals and those of the Van Rompuy Task Force.

In it we address data in the context of economic surveillance and include the role of the European Systemic Risk Board which has been set up since the crisis for data scrutiny in order to give a clearer oversight of risk and options for management.

Among our points are that there must be the right methodological tools and transparency based on harmonised economic indicators, both real and nominal. As well as these indicator values we also suggest that the indicators be complimented by qualitative assessments by the Commission.

We specifically call for further harmonisation of the methodologies for the calculation of budget aggregates in order to facilitate comparison between Member States.

And we seek to include the European Systemic Risk Board assessments as regards financial stability, stress tests, potential outward and inward spill over effects and the accumulation of private debt into the multilateral surveillance framework.

The Feio Report also includes a section on improving the reliability of EU statistics, aimed at enhancing the role of Eurostat. Again this includes on-site inspections without notice, requiring disclosure of when there has been an independent audit and harmonising public finance data based on standardised and internationally acceptable methods of accounting. We also want consistent and open disclosure of certain off-balance sheet liabilities, in particular in respect of future payments required for public sector pensions and for long term contracts with the private sector for the leasing and provision of public facilities.

We will I am sure be carrying through our ideas into the consideration of the Economic Governance proposals from the Commission, on which we are also just starting to work, and for much of which we now have full co-decision power with the Council, not just consultation.

I have just mentioned that we see a need for standardised and internationally acceptable - or I would say consistent - methods of accounting and in that context the EU and the ECB have been at the forefront of promoting a strong international quality framework and global indicators.

From the national to the international, we must take account of the need to improve statistics from the bottom up, raising public confidence, not least when the obvious background to most of what I have said so far has been an aspect of the crisis born out of lack of transparency and accuracy.

The UK is addressing this issue by setting up an office of budget responsibility and as a generality Commissioner Rehn has indicated his support for independent approaches. The UK office of budget responsibility has to act objectively, transparently and impartially. It will make independent assessments of the economy, public finances and fiscal sustainability, examining the impact of decisions made by the Government.

Another way in which the issue can be addressed is simply by making it easier to access data, and again a UK experience here is that the freedom of information act and access to statistical data has raised confidence.

Earlier on I mentioned that one of the Parliament suggestions on economic governance data was supplementation by qualitative assessments. This leads me to a very important point that market intelligence - whatever the field but in particular for financial matters - is crucial both to a fuller understanding of what data has captured, and also for earlier spotting of trends given that statistical collection and analysis may lag events by some time. At the start of the financial crisis and with issues of sub prime mortgages and excessive lending I think I expressed this in terms of "didn't the supervisors ever take a walk down the High Street!" Of course it is more than that, but soft data is important, and of course networks to collect it will be more sophisticated than my walk down the High Street.

Following on from the financial crisis great efforts are being made to ensure that all activities in the financial sector are regulated, which for the main part means reporting and data collection will increase. This is well intentioned and there is no point fighting it, the current political climate is that for a large majority safety now ranks higher than it did before, ahead of efficiency.

At the same time we would be foolish if we thought that we could stop movement to different ways of doing business - some of them unlit, dark, shadow or whatever you want to call them - which will be a consequence of regulation. It is the regulatory equivalent of Heisenberg's uncertainty principle - the more you watch and pin down the position, the less you will know about the velocity - or the more you regulate the more you create incentive for moves to where you do not know what is happening. If it takes 18 months to plan and achieve a statistical analysis, especially for some regulatory purpose, what chance have you got of knowing where things are instantaneously? So a certain versatility and flexibility is essential to keep up, even though this presses on several comfort zone which favour the certainty of stability.

If I stick for a moment with technical analogies, wings, brakes, CD players all operate through servo-motors, the basic elements of which are a high gain amplifier and a feedback loop. It is well known that if there are long delays in the feedback loop servo systems do not work, the brakes fail, the speed or direction goes awry. The same is true of financial servo systems, there are plenty of amplifiers in derivatives, high frequency trading and leverage: data, which will soon be collected as a result of new regulation provides the feedback - the trick will be to make sure it is fast enough.

In some fields one of the standard ways of collecting data has been by survey. This has immediate downsides of time lag, subjectivity and cost. For example in order to establish purchasing price parities by tracking supermarket prices, much of the information should now be obtainable from supermarket automated systems. Turning to the financial system, payments systems similarly could be used for aggregated data instead of having to survey banks, emphasising that the right kind of anonymity procedures need to be in place.

One thing is certain in the new era of more fully regulated financial services, we will all be swimming in data, as already indicated for example from the additional clearing and repository reporting for derivatives. Of course it is not the collection of data that renders markets safer and police-able; it is the interrogation and analysis of that data. So we do always need to question who and what exactly is it for. How are we going to use it, is it for a real time analysis or for ex-post checking? In practice both will be used but I would say it is the former, real time analysis, which gives safety, which gives the sufficiently fast feedback loop; ex post the best you can hope for is rectification and punishment. Data analysis, and yes speed, are fundamental to prevention rather than cure.

It has always puzzled me why we do not use more of the techniques of engineering in finance - at least it has puzzled me ever since I learned with some alarm just how comparatively unsophisticated transaction recording and tracking is. So here too if we want better tracking and automation it would be a good idea to get agreed standards for reference data.

There are lots of things that flow around in pipelines and wires, water, gas, transport, communications, or my own speciality of electrons and holes in semiconductors, and the various systemic risk boards and the US Office of financial research are - or should be - ultimately charged with developing a much better understanding of real time flows and tracking in finance. As I have said, statistics now need to be quick as well as accurate.

We can track on the internet vast quantities of information, so surely we can do the same for financial flows with tools for real time management. My thought is these systems must be evolved as part of the response to accelerating technology.

Finally, and stepping even further out of the normal financial lines, I have personally harboured a desire to apply quantum statistics to financial markets; electrons tunnel to where they shouldn't sooner than expected, perhaps markets do the same. But maybe I have to keep that as the subject for my next thesis.

Thank you for inviting me here today, and I wish you well in the rest of your conference.

ENDS