Official Statistics for a Global Economy
Eugenio Domingo Solans, Member of the Governing Council and of the Executive Board of the European Central Bank, Speech delivered at the 54th Session of the International Statistical Institute, Berlin, 20 August 2003. Introduction and overview
I feel both honoured and privileged to have been invited to address you at this conference today. The International Statistical Institute (ISI) is seen as the top international professional association of statisticians and covers, in principle, all statistical disciplines. It is also one of the oldest international scientific associations. In this regard I would like to congratulate Germany on hosting the world forum of statistics for the second time, exactly one hundred years after the first such event was also held in Berlin in 1903.
Modern day economies can only function efficiently if the public at large, researchers, markets and policy-makers are well informed about economic and social developments. Official statistics developed, compiled and disseminated in accordance with the ISI Declaration of Professional Ethics are by far the most important and reliable source of information. Moreover, the methodology of official economic and financial statistics for a global economy should be imbedded in international statistical standards, as this enhances the cross-country comparability of official statistics. It also facilitates their interpretation by trading partners, financial markets and international and supranational institutions such as the International Monetary Fund (IMF) and the European Central Bank (ECB).
While there are many uses and users of official economic and financial statistics, particularly in a market economy, I would first like to illustrate the intensive use central banks make of official statistics, taking the example of the ECB. I shall also elaborate on the ECB's medium-term priorities for economic and financial statistics for a global economy. The detailed implementation of these medium-term priorities must strike the right balance between the merits and costs of additional statistics and between their timeliness and reliability for central bank decisions. At the very centre of a monetary union there is the need to harmonise the methodologies used by the various countries for the collection and production of statistics to enable the compilation of meaningful area wide aggregates. The methodology therefore follows the international statistical standards, which in turn may be reviewed with the aim of enhancing their application. The international statistical standards will also be strengthened by the ongoing harmonisation of the international accounting rules. It is my objective to encourage the international community of statisticians, in particular those working at universities and research institutes, to actively participate in all of these developments.
The importance of official economic and financial statistics for central banks
The business area in charge of statistics in central banks – the Directorate General Statistics, in the case of the ECB – should be responsible for the statistics relevant for all functions and tasks of the central bank. Besides monetary policy, which is certainly our main "customer", the central bank statistical department should provide services to other "clients" within the central bank (financial stability and banking supervision, payment systems, international, etc.) and to other users outside the central bank (market players, researchers, general public).
The different functions and tasks of a central bank are clearly related, and statistical information required for policy actions in these different areas has related or even common sources, requires similar methodological treatments and benefits from an integrated communication network.
The previous point concerns one important principle: the independence of statistics and statisticians from policy makers. Independence and fairness in providing the data must not only rely on the professionalism of the experts, but should also be based on an appropriate organisational framework. Statistical information is so powerful that regardless of the good will of the persons involved – which I take for granted – institutional arrangements should make it impossible to speculate on what, how much and when statistical information is disclosed. Against speculation, standardisation is the specific name that fairness takes when we consider the appropriate treatment that statistical information should receive. Independence, of course, does not mean isolation, lack of communication. Statisticians need a good understanding and co-operation with other central bank business areas, market players and policy makers and other users.
Allow me to turn now to the relationship between monetary policy and statistics in the case of the ECB.
The importance of official economic and financial statistics for the stability-oriented monetary policy of the ECB can hardly be overestimated. The monetary policy strategy was announced by the Governing Council of the ECB in October 1998 and, following a pre-announced review, was confirmed and clarified in May 2003. It consists of a quantitative definition of price stability and a comprehensive analysis of the risks to price stability, comprising an "economic analysis" and a "monetary analysis".
The quantitative definition of price stability is based on the Harmonised Index of Consumer Prices (HICP) for the euro area, as developed, compiled and disseminated by Eurostat, the Statistical Office of the European Union, in close co-operation with the national statistical institutes. The HICP for the euro area is available monthly and published with a reasonable degree of detail halfway through the following month. In addition, an HICP flash estimate for the euro area is already published at the end of the reference month.
The "economic analysis" focuses mainly on the assessment of current economic and financial developments and the implied short to medium-term risks to price stability from the perspective of the interplay between supply and demand in goods, services and factor markets. The economic and financial statistics supporting the "economic analysis" include, for example, national accounts' main aggregates, government finance statistics, short-term business and labour market indicators, exchange rates, the balance of payments for the euro area, financial market statistics and the financial balance sheets of euro area sectors. All of these statistics help to assess the dynamics of the real economic activity within the euro area and of the global economy, and the likely development of prices over shorter horizons.
The "monetary analysis" focuses on money and liquidity considerations and mainly serves as a means of cross-checking, from a medium to long-term perspective, the short to medium-term indications resulting from the "economic analysis". The statistics supporting the "monetary analysis" include, for example, the sufficiently detailed consolidated balance sheet of the euro area banks, in particular the monetary aggregates and counterparts, the balance sheet of euro area investment funds, securities issues statistics and the financial balance sheets of the non-financial sectors, including any changes.
The "economic analysis" and the "monetary analysis" of the ECB's monetary policy strategy have been deepened and extended in line with the increasing availability of euro area statistics and ongoing improvements in technical tools. They will continue to be developed in the future. Many of the supporting euro area statistics are published in the statistics section of the ECB Monthly Bulletin and on the ECB's website. The ECB's Statistics Pocket Book, first published during this ISI conference, presents the key euro area economic and financial statistics in a handier format.
All modern central banks typically use a wide range of official statistics for preparing and explaining monetary policy decisions. The twice-yearly Monetary Policy Report to the Congress of the Board of Governors of the Federal Reserve System is a further prominent example. Following a simplified metaphor, statistics may be seen as the raw material entering the monetary policy process, leading to an interest rate decision which either confirms or changes the rates.
Moreover, central banks are not only intensive users but also producers of statistics. The Treaty on European Union assigns a role to both the ECB and the Community institutions, which in practice means Eurostat. The respective areas of responsibility in economic and financial statistics at European Union level were detailed in a Memorandum of Understanding, first signed in 1995 and amended in 2003. Its aim is to avoid duplication of statistical work at European Union level, to prevent conflicting requests for data from the ECB and the Community institutions and to promote high-quality statistics for the use of policy-makers, markets, researchers and the public at large.
ECB medium-term priorities for statistics
Those who are familiar with the development, collection and compilation of official economic and financial statistics know that many years of preparatory work are needed before data with a sufficient reliability can be published. The Directorate General Statistics of the ECB has therefore established medium-term priorities for statistics. They refer not only to areas of statistics for which the ECB takes responsibility at European Union level, but also set out requirements for statistics used intensively by the ECB and the financial markets but produced by other parties, mainly Eurostat. I like to call this the catalyst function of the ECB for official economic and financial statistics.
The most important items on the agenda of the ECB's medium-term priorities for statistics are: the implementation of the Action Plan on economic, monetary and financial statistics for the ten acceding countries joining the European Union in May 2004, a statistical framework for financial stability, a full system of quarterly financial and non-financial accounts by economic sector for the euro area, the Principal European Economic Indicators including the first-for-Europe principle, and an enhanced dissemination of statistics including quality indicators.
Before outlining these main medium-term priorities, I would like to announce that the ECB and the NCBs of the euro area will release a new set of bank interest rate statistics before the end of this year. The harmonised set of monthly statistics will cover interest rates applied by the banks to deposits and lending business of different maturities vis-à-vis households and non-financial corporations. It relates to outstanding amounts and new business for both the euro area and the participating countries. The new statistics will provide a comprehensive and harmonised picture of the level and volume of bank interest rates and any changes over time.
The Action Plan on economic, monetary and financial statistics for the acceding countries was jointly prepared by the ECB and the Commission. It shows that the completeness and timeliness of the statistics for the acceding countries is improving. It also shows that the national central banks and the national statistical institutes need to make concerted efforts to comply with the main requirements. The statistics supporting the assessment of nominal and real convergence are top priority. Concerning balance of payments and international investment position statistics, while data are available in most countries, the timeliness and accuracy should be improved in a number of cases. Many national central banks have already completed or aim to complete the implementation of the requirements on monetary statistics in the course of 2003 and on interest rate statistics in the course of 2004. However, only very few national central banks are currently in a position to provide quarterly financial accounts. Improvements in this area are urgently needed and must be supplemented by quarterly non-financial accounts.
The Memorandum of Understanding on economic and financial statistics agreed between the ECB and the Commission may also guide the acceding countries in organising co-operation between their national statistical authorities. Similarly, the acceding countries should implement the Code of best practice on the compilation and reporting of data in the context of the excessive deficit procedure, which assigns the compilation of the government finance statistics to the national statistical institutes, and where applicable to the national central banks.
A further medium-term priority is the development of a statistical framework for financial stability. Central banks and governments around the world are paying increasing attention to macroeconomic and institutional developments that pose potential risks to financial stability. It is worth recalling that, historically, one of the main reasons for establishing a central bank was its function as lender of last resort. Unlike the objective of monetary stability, in other words price stability, the conceptual and statistical framework underlying the central bank objective of financial stability is largely underdeveloped.
Financial stability as used by the ECB means "a condition whereby the financial system is able to withstand shocks without giving way to cumulative processes which impair the allocation of savings to investment opportunities and the processing of payments in the economy". We therefore do not confine financial stability to banking stability, but the financial stability of the banking sector is crucial for the soundness of the financial corporations sector as a whole. Against this background, the ECB relies on macro-prudential indicators, for example with respect to credit developments and competitive conditions. Many of these macro-prudential indicators are derived from the consolidated banking statistics. Structural banking indicators are another important data set. The development of harmonised profitability, asset quality and capital adequacy data is a challenge for a multinational economy such as the euro area.
The statistical framework for financial stability is still in its design phase. In a global economy, this work is best done in close co-operation between central banks, governments and international financial institutions. In this respect I would like to welcome the initiatives taken in recent years by the International Monetary Fund and the Bank for International Settlements.
While the official economic and financial statistics supporting the ECB monetary policy strategy are already fairly developed, the statistical framework for monetary stability is still incomplete. Quarterly euro area accounts by economic sector are the most important omission. They present the financial and non-financial accounts and balance sheets of households, financial and non-financial corporations and government, as well as the external transactions and financial investment positions, in one consistent framework. The statistical concepts themselves are already available and described in the System of National Accounts 93 and the related European System of Accounts 95. Certain parts of the quarterly euro area accounts are also available, such as the table of financing and investment of the non-financial sectors, the balance of payments and quarterly data on the government sector. However, an additional effort at European Union and national levels will be needed if a full set of euro area accounts by economic sector is to be compiled regularly and not more than 90 calendar days after the reference quarter.
These macroeconomic accounts provide three major advantages. First, economic and financial statistics from different sources are placed in one single framework and their reliability can therefore be compared. Second, the published euro area accounts present one single and consistent set of macroeconomic information covering the economy as a whole, including its external economic relations, but distinguishing between resident sectors with different economic functions and behaviour. Third, and most importantly, the euro area accounts enable the financial and non-financial accounts and balance sheets of the households sector to be compiled, for which it is not possible to collect direct data.
The interaction between the households and the non-financial corporation sectors is essential in a market economy as large as the euro area for the analysis and projection of economic developments. It is important to understand portfolio shifts in households' financial balance sheets and the link to variations in interest rates, investment in housing, the impact of tax and social benefit measures on households' disposable income and consumption. It is equally important to understand shifts in the financing of non-financial corporations between equity and debt, corporate investment and profitability. The approach taken in the twice-yearly Monetary Policy Report of the Board of Governors of the Federal Reserve System to the US Congress is an outstanding example of such a type of analysis.
The ECB and the European Commission have taken a joint initiative, backed by the European ministers for economic and financial affairs, to develop, compile and publish a first set of quarterly euro area accounts by 2005. This is an ambitious but also very important goal.
The euro area accounts by economic sector are also part of a wider initiative, namely the Principal European Economic Indicators including the first-for-Europe principle. The euro area is a multinational economy with a single monetary policy. High-quality euro area statistics therefore require an agreement on the priorities, the methods and the timetable for publication and revision between all parties involved, namely users and producers, the European Union and the participating countries.
The Principal European Economic Indicators contain a list of statistics, such as monthly consumer prices, quarterly national accounts, monthly and quarterly business and labour market statistics and the monthly external trade balance. They refer to the euro area and the European Union as a whole and are essential for short-term economic analysis. The Principal European Economic Indicators follow the example of the US statistical system, which in 1985 adopted a revised directive on the compilation of Principal Federal Economic Indicators. The targets set for the euro area by the Principal European Economic Indicators are more limited in scope and aim to meet the standards for timeliness adopted in the United States in only a few cases. However, when comparing the statistical systems for the two economies, the different institutional and policy environment has to be considered. While the focus of most US statistics is on providing results for the US economy as a whole, most euro area statistics are compiled for both the euro area and the Member States.
The first-for-Europe principle means that the release calendar for the first publication and subsequent revisions of the Principal European Economic Indicators follows European policy needs, and is aligned with the release calendar of the respective national contributions. The first, early release of the Principal European Economic Indicators should be based on a sufficient, but not necessarily complete, coverage of national contributions. Ideally, first euro area releases should be published at the same time as the first data for the most important countries. Later releases will then cover almost all euro area countries. The implementation of the first-for-Europe principle will significantly facilitate communication compared with the first years of the single monetary policy, in which many monthly and quarterly euro area economic indicators implicitly or explicitly changed almost every day. The publication of the euro area GDP flash estimate and related publications at the national level 45 calendar days after the end of the reference quarter is a good example for the first-for-Europe principle. This principle has been applied to the monetary aggregates and the consolidated bank balance sheet for the euro area since the start of European Monetary Union.
Official statistics for a global economy have many world-wide uses and users in addition to policy-makers. While a few of these users are trained statisticians, most users deal with official statistics only occasionally. The key economic and financial statistics have an impact on the financial markets immediately after they are published. Furthermore, they provide important information for everyday life. If this information is wrong or is wrongly perceived, it has a significant welfare cost. The dissemination of statistics including quality indicators is as important as their development and compilation.
Modern website technology has greatly improved the timely access to official statistics. Many websites also present statistical tables in both the respective national language and in English. While this is a most welcome development, it is not enough. The press releases on quarterly GDP and its components for the euro area and the G7, i.e. Canada, Germany, France, Italy, Japan, the United Kingdom and the United States, may illustrate the case. The press releases do not only differ in terms of timeliness, coverage and detail, they also differ in terms of the concepts applied to GDP and its components, the presentation of the growth rates, the relationship between nominal and real aggregates, the revision policy, etc. Furthermore, quality indicators for the data published are not always provided, and if so, they differ between the G7 countries. An important consideration when assessing the overall quality of quarterly national accounts is the reliability and coverage of actual data that are used to construct them. Some countries may have actual data from surveys and administrative data sources that provide a high-quality input for a significant proportion of the components of GDP. Others may have much less actual data and rely on modelling techniques as applied for forecasts. Information on the amount of actual data used to compile the preliminary estimates of GDP and its components is rare. Even the comparison and interpretation of quarterly GDP data for the G7 countries requires time and expert knowledge.
The dissemination of statistics, in other words its marketing, is still in its infancy. As statistics are a public good, this is not surprising. The Directorate General Statistics of the ECB has identified the dissemination of its statistics as one of the medium-term priorities. Work has also started on quality indicators with a focus on balance of payments statistics. It will take some time to develop and establish appropriate indicators. Official statistics for a global economy developed and compiled for policy-makers, markets, researchers and the public at large also require efforts in terms of their dissemination.
Further topics for official statistics in a global economy
I have underlined the importance of official economic and financial statistics for central banks and outlined the ECB's medium-term priorities for statistics. As I mentioned in the introduction, I would now like to address three further topics for official statistics in a global economy: first, how many statistics do we need, or, alternatively, how to find the right balance between the merits and the costs of new and existing statistics; second, in what way do the international statistical standards, on which official statistics for a global economy are based, need further development; third, official statistics for a global economy are an interesting and rewarding field for research and development.
There is a balance to strike between the merits and costs of new and available statistics. Official statistics cost taxpayers money. Their production also requires resources from reporting agents. Official statistics therefore have an impact, albeit a very limited one, on the competitiveness of an economy. However, a lack of required statistics, their late provision, or not sufficiently accurate statistics can have significant welfare costs. It is therefore of the utmost importance to establish a procedure according to which the merits and costs of users, producers and reporting agents are balanced when deciding whether or not to introduce new statistics. It is even more important to review the existing statistics from time to time and to decide which of them to discontinue. In general, I have the feeling that the statistics we have are not always the right ones.
While there are many national and international users, the most important users must be actively involved in the decision-making process. Users will have to justify their requirements for new statistics, to confirm their requirements for existing ones and to rank both. The production of official statistics which are only very occasionally used, must be reconsidered. Users must also be made aware that new economic and financial statistics need several years of lead-time before reliable results can be disseminated. Producers must develop and publish a more integrated set of statistics. Not each new policy issue requires a new statistic designed solely to respond to a specific question. Moreover, producers must be given a higher degree of flexibility to respond to new and substantial political priorities, such as the single monetary policy. This also implies a sufficient degree of flexibility in human resource management. In the future, the development, cost-efficient compilation and dissemination of official statistics for a global economy will not require more staff, but a high degree of well-educated staff.
The ECB has established an internal merit and cost procedure when it comes to deciding whether or not to introduce new financial statistics. Similarly, the EU ministers for economic and financial affairs support a re-balancing of the data transmission programme to the European Union involving European users and producers of statistics. I am confident that these procedures will improve the availability of the statistics required and will phase out past priorities.
The higher the demand for official statistics for a global economy, the more important the international statistical standards are. The most important of them is the System of National Accounts, SNA 93, and its European version the European System of Accounts, ESA 95, which is implemented in the Union via a legal act.
A review of the SNA 93 was recently initiated and conclusions are expected around 2008. The review is currently focused on several methodological issues which need clarification in the light of new developments. I would, however, like to look at two other important issues. The SNA 93 was developed and written by producers. This was necessary at the time in order to establish a system which was both internally consistent and feasible to apply. Now that the system exists and is applied in practice, it is not appropriate to restrict the further development of the concepts underlying the economic and financial statistics to producers only. The review of the SNA 93 must also include a user section. This may, for example, contribute to a better terminology: the current terminology uses rather detailed and precise headings, implying that the SNA terminology differs from ordinary economic terms and is cumbersome to use in a plain text when analysing the statistics. Moreover, including more definitions of key economic indicators may enrich the SNA. A few examples are government expenditure and revenue, fiscal burden, corporate profitability and indebtedness of the private non-financial sector. Furthermore, the SNA refers to a national economy. There is a clear need for an amendment covering a multinational economy such as the euro area.
A tighter alignment between the international accounting standards and the international statistical systems is very desirable. Despite some differences, there are clear advantages in ensuring that international statistical and accounting standards are harmonised to the fullest extent possible. Such harmonisation will enable the same source data to be used for several purposes. This will increase the reliability of the macroeconomic statistics and will reduce the reporting burden for corporations. Similar consistency and efficiency gains are possible for the government sector. In this respect the Directorate General Statistics of the ECB is participating in an international initiative to align government accounting practices with international statistical standards. In general, statisticians should be closely involved in discussions on accounting standards in order to avoid unnecessary deviations between these standards and the statistical requirements.
My third topic looks at universities and researchers. Economic and financial statistics, and in particular national accounts, are an integral part of economic studies, just as business accounting is an integral part of business administration. It is astonishing that in Europe, at least, the number of graduates and post-graduates with an advanced education in economic and financial statistics is not only insufficient, but is actually decreasing. There is a need to reflect the growing importance of official statistics for a global economy better in university courses. Similarly, researchers may find it attractive to contribute to the further development and application of international statistical standards and their alignment with the international accounting rules, to a better presentation of statistics including the development of quality indicators or to a more efficient compilation of euro area data. The Directorate General Statistics of the ECB will be happy to co-operate closely with a select group of interested researchers at the leading edge of official statistics for a global economy.
In conclusion, the development of European economic and financial statistics has been so remarkable in the past ten years that I am tempted to call it a "silent revolution". I would even say that European Monetary Union itself would not be possible without those developments in the statistical field. However, this silent revolution must now continue at an international level. The ultimate goal would be to reach a similar degree of harmonisation among large economic areas as there is within them. This may certainly look like a revolutionary achievement, but I should like to add that it is also a must for a truly integrated international community.