Euro area statistics and their use for ECB policy-making
Speech by Jean-Claude Trichet, President of the ECBdelivered at the Second ECB Conference on Statistics,Frankfurt am Main, 22-23 April 2004.
Colleagues, fellow central bankers, ladies and gentlemen,
I would like to welcome you most warmly to this the Second ECB Conference on Statistics.
I am particularly happy to see so many participants from the countries that will join the European Union (EU) in just a few days time. This milestone in European history marks the successful completion of intensive preparations in many fields, including statistics.
The goal of this conference is to review the statistical requirements for central bank policy-making. In addition, it provides the statistical community with an opportunity to bid farewell to Eugenio Domingo Solans, the Executive Board member responsible for statistics, who will leave the ECB at the end of May 2004.
Although the emphasis of this conference will obviously be the future of statistics, I would also like to take this opportunity to briefly review the current state of euro area statistics that are produced by the European System of Central Banks (ESCB).
Let me first, however, underline the importance of euro area statistics for ECB policy-making.
The importance of euro area statistics
We should not underestimate the importance of official monetary, financial and other economic statistics for the ECB’s monetary policy. As statistics provide the foundation for economic analysis and policy-making, the availability of trustworthy and timely statistical time-series is a prerequisite for an effective and correct assessment of the monetary and economic situation and future prospects. Moreover, economic agents also use these statistics for their own decisions, which then drive the developments of financial and other economic markets.
In this regard, I would like to stress the importance of a single set of statistics that is used by all economic and financial policy-makers, market participants and the public at large. The familiar credo of the media, “facts are sacred, opinions are free”, should also apply to statistics and the use of statistics. An effective policy preparation process requires that public and private decision-makers need not debate about the numbers, whatever the outcomes are. Instead they can then focus their deliberations on the best possible policy assessment based on these statistics and the forecasts derived from them.
Of course, this also places a heavy responsibility on the shoulders of those who are responsible for the development, compilation and dissemination of the statistics. In this regard, I fully agree with the content of the statement in the joint report on euro area statistics and indicators adopted by the ECOFIN Council and the Commission and forwarded to the European Council in spring 2003. The report states: “For the smooth functioning of the euro zone as a monetary union working under a single monetary policy and decentralised but coordinated fiscal policies, macroeconomic statistics of the highest quality are required. The principles of impartiality, reliability, relevance, cost-effectiveness, statistical confidentiality and transparency must be fully applied”.
The increasing importance of euro area statistics means that appropriate priorities should be set and the corresponding resources made available for compiling the contributions to these statistics at the national level. This may need to coincide with a continued shift in emphasis from simply compiling national statistics to delivering a pertinent contribution to European statistics. This focus on the euro area has already been largely achieved for statistics that are compiled and disseminated by the ESCB, such as monetary and financial statistics, the balance of payments and international investment position statistics, and quarterly financial accounts. However, for other economic statistics too, reliable euro area aggregates should be published as soon as possible and preferably at the same time as the data releases for at least the largest countries. This applies, in particular, to the short-term, high-frequency statistics on prices and costs, national accounts, labour markets, business developments and external trade, which are needed for business cycle analysis and for monetary policy preparation.
This practice of publishing simultaneously area-wide and national data follows the “First for Europe” principle. This principle implies that the statistics, at least the Principal European Economic Indicators, are published on the same day for the euro area as for the fastest countries. These countries should cover a sufficient proportion of the euro area to enable the computation of reliable estimates for the area as a whole. In addition, the release calendar of these indicators should take account of European policy needs, and revisions and seasonal adjustment practices of these indicators should be coordinated. At present, this situation has not yet been reached. And as a consequence, a number of European and national institutions, market participants, forecasters, etc. are all second-guessing, on their own, at what the euro area aggregate will be as soon as the data for one or more of the larger Member States have been released. Moreover, even when the euro area aggregate has been published, based on sufficient country coverage, it can still be revised every time another “larger” Member State publishes this statistic for the same reference period. The volatility of the euro area aggregate is clearly a second-best solution both from a monetary policy perspective and in terms of the effectiveness and efficiency of the statistics concerned.
Within the statistical community of the ESCB, a shift in emphasis towards a truly European team spirit is already taking place. For instance, when the ECB releases its monthly statistics on monetary developments in the euro area, the national central banks (NCBs) publish the results for their respective country shortly afterwards. I am convinced that the European Statistical System – comprising of Eurostat, the Statistical Office of the European Communities, and the National Statistical Institutes (NSIs) – will want to follow suit.
The harmonisation of the methodologies used by the various countries for the collection and production of statistics lies at the heart of a monetary union: it enables us, for example, to compile meaningful area-wide aggregates. At the same time, the fruitful and fairly detailed statistical harmonisation achieved in the euro area may inspire the further development of worldwide statistical standards. Given that cross-country comparability of official statistics is key to their usefulness and credibility, all countries across the globe should then want to implement such worldwide standards in their official economic and financial statistics. Of course, this applies a fortiori to external statistics, such as the trade balance, the balance of payments and the international investment position, where we still have a lot of progress to make. Partner countries will only be able to share a common analysis of their bilateral economic and financial relationships if their mutual external statistics mirror each other. In turn, this is only feasible if these statistics are compiled using the same concepts and definitions, and similar methodologies. In view of their importance for policy-making, it may be worthwhile embarking on more intensive coordination among the statisticians involved, particularly those within the euro area and in the other major economic blocs of the world, to minimise the present “asymmetries” in external statistics.
Besides monetary policy, which is certainly the main user of our statistics, the central bank’s statistical department should provide services to other central bank functions (financial stability, payment systems, operations, etc.) and, of course, to users outside the central bank (market players, researchers and the general public). The different tasks of a central bank are clearly related, and statistical information required for policy actions in these different areas has related or even shared sources, it requires similar methodological treatments, and it benefits from being embedded in an integrated statistical framework.
Going beyond the use of ECB statistics by the ECB itself, these statistics may also be able to play a role as a neutral and trustworthy benchmark for market participants and the public at large. In this context, I would like to give three examples.
First, the European Financial Markets Association (ACI) recently published the final report on the “Short-term Paper (STEP) Market in Europe – Recommendations for the development of a pan-European market”. This report recommends, among other things, that the ECB, as a neutral party, should publish statistics on volumes and prices covering the STEP market.
Second, since last November, the ECB and the NCBs have released every month harmonised euro area and national statistics on interest rates on deposits and loans by households and corporations. These and other ECB statistics are also suitable for assessing progress in financial integration in the euro area.
Third, the ECB, together with colleagues from NCBs and other experts, is developing a European standard classification of financial instruments which may eventually be of use to market participants as well.
As a last example of the importance of euro area statistics, I would like to mention the indicator that lies at the heart of the ECB’s quantitative definition of price stability, namely the Harmonised Index of Consumer Prices (HICP) for the euro area. This index is compiled and disseminated every month by Eurostat, in close cooperation with the national statistical institutes. It is made available with a reasonable degree of detail and timeliness. In addition, a very timely “flash estimate” is published at the end of the reference month.
However, for some time now, survey results have indicated that a number of euro area citizens have the impression that the inflation is higher than the moderate levels that are evidenced by the HICP development.
A first explanation of this phenomenon may be the lasting memory of some (or a few) price increases, mainly for personal services, that occurred when the euro was introduced. These increases had no lasting impact on inflation because fortunately there were not many of them. But, in certain cases, they may have left a psychological mark.
A second reason may be the need for some consumers to adapt to the euro as the new yardstick for price tags. This effort may be associated with some suspicion about actual price developments.
A third cause for the discrepancy between the inflation “feeling” of some people and the actual inflation is the fact that the prices of products that are bought “out of pocket” and at a high frequency have tended to rise faster than the average. By contrast, the prices of several products that are bought only occasionally, such as household appliances or package holidays, or the prices of services that are paid by giro transfers, for example telephone subscriptions, have typically increased much less or have even decreased. Understandably, if you are confronted daily by somewhat higher prices for relatively small purchases and you only benefit occasionally from more substantial price advantages for some expensive products, you may well perceive inflation to be higher than it actually is.
Taking all this into account, I can assure you of the quality of the HICP estimate and I have invited the consumers in the euro area to fully trust the figures which show that inflation is moderate. Households and all other economic agents can be assured that the ECB will preserve the purchasing power of the euro and that future price stability will continue to be ensured in line with our definition.
Some achievements of euro area statistics
When looking at the current state of euro area statistics, it is worth recalling that less than a decade ago, we had not even established the statistical requirements for Monetary Union, let alone that euro area statistics existed. As the ECB’s former Director-General of Statistics, Peter Bull, describes in his record of the euro area’s statistical history, the approval of the statistical requirements in July 1996 was a landmark in the statistical preparations for Economic and Monetary Union (EMU). A lot has been achieved since then, which may perhaps be best illustrated by the fact that in 2003 the ECB issued 48 statistical press releases and Eurostat many more. The ECB now publishes every month, among other things:
the most recent monetary aggregates, their components and counterparts;
the balance sheets of all banks and similar institutions in the euro area, and a list of all the Monetary Financial Institutions;
a full balance of payments, including the Eurosystem’s international reserves;
the effective exchange rates of the euro (which are updated daily);
new issues, redemptions and outstanding amounts of almost all debt securities and quoted shares in the euro area;
average interest rates paid and received by households and corporations on all kinds of loans and deposits (including loans for house purchases and for consumption).
In addition, quarterly press releases are published on investment fund statistics, specified by the investment policy of the fund concerned, and on the financing and financial investment of households, non-financial corporations, insurance corporations and pension funds, and the government. Finally, the ECB regularly disseminates the international investment position of the euro area, various financial soundness indicators, a range of government finance statistics, a variety of financial market statistics, a residential property price index, statistics on payments and statistics on the international role of the euro. Only five years ago, most of these statistics did not even exist at the euro area level.
Of course, all these statistics, the methodologies applied, the release calendars, etc. are available on the ECB’s website. In addition, they are published in the euro area statistics section of our Monthly Bulletin and in our monthly Statistics Pocketbook. The Pocketbook was launched in August 2003 and aims to provide euro area citizens, the media and other institutions with relevant, up-to-date euro area economic and financial statistics and publishable indicators in a user-friendly format. It also contains several tables comparing the euro area with other main economic regions, such as the United States and Japan, and it contains various national statistics.
These achievements have only been possible because of intensive and fruitful cooperation between the statistical departments of the NCBs and the ECB, as coordinated by the Statistics Committee. In addition, the division of tasks between Eurostat and the ECB’s Directorate-General Statistics has been laid down in an update of their Memorandum of Understanding. This update was signed in March 2003 and has recently been re-examined in certain areas, such as the compilation of euro yield curves and annual financial accounts. In those and other areas of statistics, the ECB also stands ready to serve the needs of the European Commission and other users.
At the national level, the distribution of statistical tasks between a country’s central bank and its statistical institute is not the same across Europe. Solving the coordination issues at the European level is one of the main tasks of the Committee on Monetary, Financial and Balance of Payments statistics (CMFB), which brings together the senior managers responsible for statistics at Eurostat, the ECB, and the NSIs and NCBs of the EU Member States. This Committee not only has a pivotal coordination function for European statistics, but also must play a key role in one of the most sensitive and crucial areas of statistics in Europe, that is, the statistics as part of the Excessive Deficit Procedure (EDP). These statistics should be compiled in strict conformity with the European Regulation for compiling national accounts, the European System of Accounts 1995 (ESA 95), and the related Manual on Government Deficit and Debt.
The Code of best practice on the compilation and reporting of these data, which was endorsed by the ECOFIN Council in February 2003, states that the actual data of the ESA 95 government accounts must be compiled by the National Statistical Institutes, and where applicable by the national central banks. The NSIs must act in full scientific independence and the government must ensure that accurate basic data and other information needed for compiling reliable ESA 95 accounts are made available to the compiling authorities in time and with sufficient detail. All current and new EU Member States are encouraged to fully implement these statements in the Code of best practice and to build up the statistical capability in both the NSIs and NCBs that is needed for an adequate fulfilment of these tasks. The Code of best practice also refers to the procedure for resolving methodological issues and states that Eurostat must consult the CMFB in cases which are not adequately covered by the ESA 95, or cases that are particularly complex or of general interest. Eurostat must, as a general rule, publish its decision together with the CMFB opinion. Clearly, the CMFB should then form its opinion strictly in the framework of the ESA 95 Regulation and solely on the basis of its professional expertise in national accounts, in line with a transparent procedure, which is available on the CMFB website (www.cmfb.org). Obviously, this procedure requires all CMFB members to contribute a high level of professional expertise in national accounts. Improving the credibility of fiscal figures and the effectiveness of the monitoring and auditing process is still needed and can be done without altering the current fiscal framework.
As far as other areas of euro area statistics are concerned, the Action Plan on EMU Statistical Requirements, which was endorsed by the ECOFIN Council in September 2000, has stimulated substantial improvements in the timeliness and availability of crucial economic indicators. Major progress has been made in, for example, main aggregates derived from the quarterly national accounts, the HICP, quarterly government finance statistics, several short-term business indicators and foreign trade statistics. By contrast, the situation is not yet fully satisfactory for the quarterly national accounts for institutional sectors and labour market statistics.
Future priorities for euro area statistics
Looking ahead now at the future orientation of euro area statistics, our short-term priority should be to help the new Member States to fully catch up with the current euro area countries in terms of the quality and availability of economic and financial statistics. In this context, I am delighted to announce that the ECB, together with the Commission, will shortly publish, for the first time, long-term interest rates for the acceding countries. At the same time, I am aware that any enlargement of the EU and, subsequently, the euro area creates particular statistical challenges for the ECB, Eurostat and all of the countries concerned. For example, adequate resources must be available to re-create all euro area time-series each time one or more countries joins EMU. Among these statistics, the indicators that will be used to assess the convergence of these Member States towards EMU deserve particular attention.
Currently, the ECB’s Governing Council monitors twice a year the progress of the current pre-ins and the new Member States in their statistical preparations for EMU. Many countries have made very significant progress in a very short space of time, but full compliance with all legal requirements has not yet been reached. Since the compilation of statistics is an activity that is typically characterised by significant economies of scale, the situation in the smallest Member States, particularly those with sizeable financial sectors, deserves specific attention.
More generally, possibilities for economies of scale in the compilation of European statistics should be regularly re-assessed and exploited when appropriate. A good example in this context is the current creation of a Centralised Securities Database at the ECB, which should become available on line to the NCBs as well. Similarly, more comprehensive corporate data for conducting balance sheet analysis are widely considered as a priority in the period ahead, but organising a new data collection for this purpose at the national level would be very costly and would create high administrative burdens on businesses. The best way forward appears to be to integrate and expand the national Central Balance Sheet Offices (CBSO) that already exist within various NCBs and to set up a more comprehensive European CBSO. A last example of possible economies of scale at the European level refers to more international concentration of the data collection from multinationals. Obviously, progress in this area also depends on the interest expressed by the European multinationals themselves.
Among the other important new statistics that are in the pipeline, I would like to single out the forthcoming geographical breakdown of counterparties in the euro area balance of payments and international investment position. The ECB intends to launch these statistics within the next 12 months, and it is hoped that they may give some important new insights into the external economic relations of the euro area. Also over the next year, the frequency of the statistics on the international investment position will be increased from annual to quarterly. These statistics provide an overall picture of the euro area’s claims on and liabilities to the rest of the world. In the near future, the monetary presentation of the balance of payments will also be published monthly instead of quarterly, while work is already under way to update the trade weights of the ECB’s effective exchange rate statistics. In that way, the analysis of the euro area’s competitiveness remains up to date. Finally, the investment funds statistics will be enhanced by separating the changes in the outstanding amounts into two: the changes due to transactions on the one hand, and those due to valuation and other changes on the other.
All these statistics will become available on a much enhanced, easily accessible statistics section of the ECB’s website. Also, a statistical information service has been set up to answer specific queries from both external and internal users.
The priorities for ECB statistics in the medium term are very well reflected in the themes of this conference:
A full system of euro area quarterly accounts for institutional sectors, integrating both financial and non-financial accounts, would provide an integrated statistical basis for the ECB’s monetary and economic analysis.
More comprehensive statistics for the monitoring and analysis of financial stability, with explicit links to the statistics for monetary policy purposes, would make it possible to deliver a more thorough analysis of interactions between these policy goals.
Further development of external statistics (e.g. on the international role of the euro or concerning the details of the international investment position) would enable better monitoring of the continually increasing, global economic intertwining.
Better interplay between business accountants and statisticians would enable cross-fertilisation of the considerable expertise in these adjacent fields. This has an enormous potential for decreasing the administrative burden for businesses, while at the same time improving the quality of statistics.
An increasing focus on the various quality dimensions of European statistics as reflected, for instance, in the ECB’s intention to regularly publish quantitative output quality indicators (on revisions, consistency, etc.) and more comprehensive quality reports on its statistics.
At the same time, the ECB attaches great importance to improving further the availability, timeliness and other qualities of the Principal European Economic Indicators and to receiving more elaborate and consistent euro area statistics for assessing and analysing productivity changes, for service activities and for housing, including house prices. Moreover, incorporating price developments of owner-occupied housing in the HICP would be an important achievement. In any case, Member States may want to consider giving more prominence to national HICPs and less to their specific national Consumer Price Indices (CPI). By way of a supplementary measure, the analysis of any inflationary pressures would be served by the regular availability of an HICP-variant that excludes the direct effects of tax changes and, possibly, other administrative measures.
However, implementing the necessary improvements to European statistics can only be achieved if sufficient resources are made available. It may therefore also be useful to regularly review the need for existing statistics that were agreed upon many years ago.
To conclude, let me say that I wish to reflect today the importance that the ECB attaches to timely and reliable euro area statistics as a basis for policy-making. Statistics are like the glasses through which policy-makers and all other economic agents view macroeconomic reality. If the glasses are totally reliable, well polished and easy to handle, we may hardly notice that we are wearing them and we do not realise how vital they are for our clear view of reality. Surely, we only start to complain when the glasses are out of focus and we want to have a new pair when glass-making technology proceeds. This conference may be a good opportunity to put our glassmakers into the limelight, for a change, and to tell them, at the same time, how we would like our glasses to be even better polished in the future.
In any case, it is clear that producing euro area statistics is all about teamwork: time and time again, from the preparatory stage to publication. Over the past six years, Eugenio Domingo Solans has ably steered the statisticians of the European System of Central Banks: both quietly and persistently, directly and indirectly – but always in the right direction. I am convinced that the ground has now been laid for further excellent work to complete the provision of euro area statistics in the future.
Ladies and gentlemen,
I wish you all very fruitful discussions. It gives me great pleasure to declare the conference open.
Thank you very much.