THE FINANCIAL CRISIS AND THE EUROPEAN STATISTICAL SYSTEM

Transcription

1 THE FINANCIAL CRISIS AND THE EUROPEAN STATISTICAL SYSTEM Kurt Wass, EFTA Statistical Office ABSTRACT: THIS PAPER FOCUSES ON HOW THE FINANCIAL CRISIS WAS DEALT WITH BY THE ESS, OR MORE SPECIFIC WHAT WAS THE MAIN RESPONSE AND FOCUS OF EUROSTAT TO THE FINANCIAL CRISIS. IT SEEMS THAT IT WAS URGENT FOR THE ESS TO KEEP THE POLITICAL NOISE AS LOW AS POSSIBLE, BUT THE QUESTION IS WHAT THE COSTS OF THIS WERE. THE PAPER ALSO SUGGESTS THAT THE STATISTICALLY COMMUNITY CANNOT BE HELD RESPONSIBILE FOR NOT PROVIDING EARLY WARNING SIGNALS ON THE CRISIS, BUT IT IS MAYBE NOT TOTALLY INNOCENT EITHER. 1. SOME HISTORICAL BACKGROUND 1. Throughout history there have been a number of financial and economic crises differing in their causes, sizes, impacts and geographical coverage. In the last century the "Great Depression" in the 1930s is certainly the most known crisis and in current debate the one most referred and compared to the current crisis. After a couple of financial and economically "peaceful" decades following the World War II, the 1973 oil crisis was perhaps the first crisis since the 1930s to have a major global economic effect. In the 1980s and 1990s several crises occurred such as the debt crisis in Latin America in the early 1980s, the collapse of the real estate market in Japan in 1990, the world wide "Black Monday" on the stock markets in 1987, currency crises due to short selling speculations leading to devaluations of several currencies, the Mexican economic crisis in 1994/95, the Asian Financial Crisis in 1997, the Nordic Banking Crisis in the late 1980s/early 1990s etc. 2. Some of these crises also had consequences for statistics. Even though history can be read and interpreted in different ways, it seems generally agreed on that the "Great Depression" paved the way for the economic thoughts and ideas of Keynes. Keynes's theory not only affected economic thinking and policies, but also the development of national accounts was largely motivated and inspired by Keynes's theory and these statistics soon showed their usefulness. The statistical work of American economists such as Simon Kuznets showed in the early 1940s that the American economy was running far below its capacity. Kuznets s statistics showed how much the economy could produce for both civilian and military purposes of unused capital and labour. It is commonly believed that the building up of statistical systems in both USA and the UK significantly improved the ability to use their resources, and consequently this was of great importance to the Allied powers in World War II. 3. The shortcomings of the financial systems were also addressed after the Mexican crisis in The Mexican crisis led to a closer surveillance by the IMF of national policies through the establishment of new international standards of timely reports of economic and financial statistics - SDDS (Special Data Dissemination Standard) and GDDS (General Data Dissemination Standard). The statistical transparency awareness was addressed and further strengthened by the IMF after the Asian financial crisis in 1997 by

2 introducing the Code of Good Practices on Transparency in Monetary and Financial Policies where one issue was that the government sector should clearly be distinguished from the rest of the economy. 4. In the current crisis the so-called Interagency group on economic and financial statistics has been established, which under the chair of the IMF also includes the Bank for International Settlements, the European Centralbank, Eurostat, OECD, the UN, and the World Bank. This group has established a website showing relevant, timely and comparable economic and financial indicators for a group of important/systemic countries already available in these international agencies as well as links to other relevant websites. One of the conclusions of the Interagency group so far seems to be that the financial crisis is not the result of a lack of proper macroeconomic statistics but that some gaps in the statistics have been revealed, and the group is working on an inventory of these gaps. Moreover, the Interagency group has also pointed out that economic statisticians should be more attuned to the legitimate data needs of policy makers. 2. EUROSTAT S FOCUS 5. Eurostat is not an ordinary statistical office or the statistical division of an international organisation. Eurostat is one of the Directorates Generale of the European Commission having specific tasks in monitoring statistical data reported from the Member States where some of those data also serve specific administrative needs within the European Union. Eurostat can also make amendments or put reservations on data reported from the Member States. This double role of Eurostat to serve politically oriented needs in the EU come together with the tasks to organise and coordinate the statistical methodological development in the European Community has created some tensions during the financial crisis. 6. Budgetary rules and access to various financing in the EU like structural funds are to a large extent based on statistical information. Economic and financial statistics like GDP, GNI, VAT and the deficit and debt of general government are key statistics for administrative purposes, and the latter was, and still is, Eurostat's main focus in relation to the financial crisis. The background for this is the so-called Excessive Deficit Procedure (EDP). In order for the Economic and Monetary Union to function smoothly a need for requiring fiscal discipline was recognised. The EDP was established in the Treaty and specified in the dissuasive arm of the Stability and Growth Pact (SGP) where criteria to ensure the desired fiscal discipline were defined. These criteria state that the deficit and debt of the general government sector should not exceed a deficit-to-gdp ratio of 3% and a gross nominal debt-to-gdp ratio of 60%. The Member States report biannually EDP notifications to Eurostat. If it is decided that the deficit is excessive in the meaning of the Treaty, the Council issues recommendations to the Member States concerned to correct the excessive deficit and gives a time frame for doing so. Non compliance with the recommendations triggers further steps in the procedures, including for euro area Member States the possibility of sanctions. 7. The statistical indicators outlined in the EDP framework are based on the European system of Accounts, ESA95, which is the conceptual reference framework for national accounts legally binding the EU. The ESA95 is one of the cornerstones of the European Statistical System (ESS). 2

3 8. On behalf of the Commission, Eurostat has the operational responsibility for the figures in the EDP notifications. This means in practice that Eurostat monitors the countries' EDP reports on government deficit and debt, and it is also Eurostat's task to ensure that these reports are in line with the methodological statistical guidelines as stated in ESA95. The level of details in the EDP notification reports is very high showing links between government accounts and the statistical concepts of government deficit/surplus and debt. The examination done by Eurostat is often on the level classification of single transactions and institutional units. Since ESA95 is basically a framework regulation not all detail that can occur in practice are foreseen. Therefore, to complement ESA95 a flanking guidance has been developed, the ESA95 Manual on Government Deficit and Debt. Even though, disagreements between Eurostat and Member States do occur, and disputes between Eurostat and individual Member States have been brought three times into the European Court of Justice on the initiative of the Member State in question. Eurostat s positions have prevailed in all cases so far. Because the issues at stake might have political consequences, discussions often involve not only national statistical institutes. In some cases also ministries and the political parts of governments have been involved as large single transactions sometimes can significantly impact government finances, typically in the context of large infrastructure investments. 9. This specific part of the ESS plays an important role not only in developing national accounts methodology in Europe, but also in practical policy making. There have been few cases where political decisions in Member States have been changed or postponed due to Eurostat decisions on the statistical treatment of planned economic policy measures. Member States frequently ask Eurostat for advice, both ex post and ex ante. It might be argued that providing ex ante advice is institutionally questionable, since Eurostat will potentially be in a difficult situation judging later on its own advice. 3. EUROSTAT S RESPONSE 10. Towards the end of 2007 Eurostat became aware that some banks in distress in Germany and the UK could significantly affect the governments' debt and deficit. These first observations were discussed internally in Eurostat late 2007/early 2008, and a methodological paper, that on a general basis discussed the limited number of cases so far, was presented to the countries in the Financial Accounts Working Group (FAWG) in June Eurostat proposed in this meeting to establish a dedicated task force, which was agreed on. The first Task Force 11. The first task force, a Eurostat task force, met in July 2008, with the participation of only Germany, the UK and Greece in addition to ECB and DG ECFIN. A number of countries that were invited asked to be excused and did not meet. It quickly became clear in this first meeting that it was difficult to reach conceptual agreements on the so far limited number of cases observed. The objectives for the task force were to discuss involvements of governments and the statistical treatment of those. The cases were difficult as the roles of governments' were only indirect, contingent (guarantees), or in cooperation with nongovernment entities. The second meeting of the task force took place in August, and in the end a conceptual agreement was reached on two important cases in Germany. In one case it was agreed that some operations by a public owned unit in order to rescue a private bank were conducted on behalf of government, and in the statistical treatment agreed on was to reroute the support through government. This was reflected in the following EDP notification turning an otherwise small government surplus into deficit. 3

4 12. One important acknowledgement in this first task force was that the methodological framework, ESA95 and its complimentary manual, was not sufficiently developed to clearly guide statisticians in these new and rather complicated cases. ESS Action Plan 13. During September 2008 and the following months after the Lehman Brothers collapse, the financial turmoil escalated and governments and central banks intervened massively worldwide to an extent never observed before and several Member States had now become affected. On the background of these events in the world economy in September 2008, the Summit of Euro Area countries made on 12 October a declaration on a concerted European Action Plan for these countries, and this plan was also endorsed by all EU Member States. A follow up reaction from the statistical community was the so-called ESS Action Plan. The ESS Action Plan underlined the need to safeguard the existing surveillance framework ESA95, but indicated that the framework lacked guidelines to deal with the situation. The plan also made references to the rules on the internal market, claiming that rules on state aid legislation and the SGP included provisions for this kind of situation. This plan established some basic principles to be followed that only seemed to repeat already well established fundamental principles of statistics. The operational principles in the plan indicated that the national accounts rules as defined in ESA95 should be applied but with an appropriate interpretation. 14. One of the most important considerations in government finance statistics in general is to determine whether a government capital injection is to be classified as an expenditure impacting the deficit, or if it should be classified as a financial investment with reasonable expectations to receive a return on the investment in the near future. The ESS Action Plan signalled that government capital injections into so-called fundamentally profitable private banks that found themselves in a liquidity squeeze should be classified as financial investments without impact on the government deficit. By this the ESS Action Plan signalled a direction of the following practical methodological work of experts, and it can be questioned whether such a signal of direction was appropriate. It created also an impression of that whilst administrative rules are inflexible, national accounts data reported through EDP where expected to be flexible. The second (CMFB) Task Force 15. In this extraordinary situation a new task force as part of the ESS Action Plan was established, this time with a much broader participation and under the control of the socalled Committee for Monetary, Financial and Balance of Payments statistics (CMFB). 19 countries were represented in this task force in addition to DG ECFIN, ECB. The IMF also met as observers. Switzerland was represented as the only non-eu country. The task force met twice, late November 2008 and early January Prior to the first meeting, the countries had been asked to provide a list of "financial turmoil events" and the list totalled to more than 150 events of government interventions in relation to the financial crisis. 16. The mandate of the task force was to prepare a questionnaire asking the CMFB for its opinion on a number of key questions identified by the task force. The final note did not discuss individual cases (except the so-called Bank of England liquidity scheme), but asked the CMFB for its opinion on some fundamental methodological questions developed by the help of the national experts in the task force. The task force agreed that 4

5 one should concentrate on some key issues like; a) what type of intervention is observed (e.g. recapitalisation, lending, purchase of assets, exchange of assets, guarantees, debt cancellation/assumption etc.), b) what institution undertook the intervention (existing government unit, central bank, existing public corporation or a new institution). In addition, the opinion on issues concerning valuation of non-traded (toxic) assets and time of recording of events were asked for. 17. The CMFB provided its opinion to Eurostat in March Eurostat Decision 18. On 15 July 2009, based on the advice of the CMFB, Eurostat published its decision on the statistical recording of public interventions to support financial institutions and financial markets during the financial crisis. The Eurostat decision diverged from the advice of the CMFB, and in the Eurostat news release it could be read; In this context the Eurostat Decision does not fully take the same approach as the CMFB opinions in the case of the classification of financial bodies in which non-government units hold a majority stake. In some circumstances, such entities will be classified, at inception, outside general government. See also: These circumstances referred to in the Eurostat decision were cases where majority privately owned entities receiving government guarantees should be recorded outside the general government sector if the expected losses they would bear were small in comparison to the total size of their liabilities. The first Task Force meets again 19. The Eurostat decision did not address all practical questions, and the first Eurostat task force from 2008 was brought back to life by a meeting convened on 27 August. The task force discussed some topics Eurostat wanted to bring clarity into before the EDP notification round in October. From the outcome of the meeting two issues could be mentioned in particular. First, it was agreed on that a new supplementary table should be reported with the EDP notification showing all government interventions in relation to the financial crisis, irrespective if they had an impact on government deficit and debt or not. In addition, Eurostat reported that the statistical treatment as stated in the 15 July 2009 decision would only be applicable under the financial crisis, and that Eurostat would give notice when the financial crisis will be over for statistical purposes. 4. POLITICAL STATISTICS 20. Traditionally, the direct users of macroeconomic statistics are forecasters who need national accounts data, whose forecasts are used by economic policy makers. In the context of the EU another important use of statistics is for administrative purposes, as mentioned in the first section of this paper. It can be argued that administrative needs and uses of statistics give a much higher importance to statistics than otherwise, which can be seen as a pro in the sense that statistics are subject to much greater attention and exposure and probably more resources are attached. On the other hand, this leads to a need for legislation that can impose a rigidity and constraint on the development of statistics. The effect of this is that the role of the statistical system in Europe faces the risk to be politicised. However, one can argue that this importance given to statistics in the EU has strengthened the need to fully harmonise statistics in Europe, which is a goal in itself to obtain coherent and comparable statistics across countries. 5

6 21. During the financial crisis, the processes within Eurostat and the Commission as well as in the dialogues with the countries have shown that the ESS is indeed subject to political attention. It would perhaps be naive to believe that the basis for how to evaluate government spending in a scale never seen before and with a potential to undermine the reliability of the SGP should be the sole stage for statisticians. On the other hand, one should be critical when the independence of statistical systems and national statistical institutions seems to be under pressure due to political responses in given economic circumstances. 22. In national accounts, a fundamental principle is that the statistics should show economic realities and not reflect legal or administrative arrangements. EDP should be based on the well established principles of national accounts, and not the other way around. The ESS Action Plan indicates that the rules on the internal market, on state aid and SGP can include provisions for this kind of situation. This is difficult to justify and it threatens the independence of ESA95 and thereby the ESS. It is undoubtedly true that statisticians in Europe have been given through the role of statistics in Europe a unique position to influence important policy making, and some might even argue that statisticians also can curtail political decision makers. But the distinct role of statistics in Europe should never be used to undermine the integrity of the statistical system. If politicians find a national accounts framework inadequate for administrative needs, this is not a reason to change national accounts but rather to question if national accounts is suitable for such purposes. It is hard to see good alternatives. On the contrary, the ideas of national accounts were in fact partly a brain-child of the crisis in the 1930s, and despite its weaknesses, it has proven to be one of the major tools to provide information necessary in the conduct of economic policy making. The role of national accounts has been of great importance for the development of modern economies. If the principles of statistics were stretched to fit political desirers in various circumstances the value of statistics would be very limited. However, one should not be surprised that political pressures occurred in the financial crisis. 23. Not only has Eurostat been subject to pressure in the financial crisis. As the economic recession hit, the government finances in several countries have deteriorated dramatically, and the need to cut government expenditures as a consequence. Many national statistical institutes have experienced large cuts making it difficult to maintain the same level of providing high quality statistics. It can be seen as a paradox since crises are often the result of insufficient information. 5. CONSEQUENCES OF THE FINANCIAL CRISIS ON ESS 24. The financial crisis cannot be explained by lack a of macro economic statistics. In terms of early warning signals it seems more appropriate to point at the lack of transparency in business accounting preventing sufficient scrutiny of financial surveillance authorities. From a statistician's point of view, it is interesting to observe international debate on the need for harmonisation and cooperation on financial surveillance since this need has been addressed for a long time within the international statistical community. 25. Although it is unjust to hold statistics responsible for the crisis, there is probably room for improvements. It would be an idea to emphasize the need for balance sheets for the different sectors in the economy. For example, statistical balance sheets for the private corporations sector in Spain have for some years revealed that the economic growth in 6

7 Spain has been significantly investment led, and the sustainability could have been questioned many years ago. This information was already available, and one question to ask is if the ESS has been able to understand and communicate the importance of existing information. 26. In several countries, the growth in residential and real estate prices compared to the general economic growth should probably have been a much bigger concern. Even if economists for decades have warned about the building up of bubbles in these markets, little has been done by politicians. One reason for this can be lack of good statistics, e.g. price statistics for housing combined with balance sheets for the wealth of the households. This would have given valuable information to assess the vulnerability of the household sector for changes in housing prices. One could question why this very important issue has not been raised within the ESS. 27. The success of a statistical system is its capability to meet the needs of users. Statisticians are often introverted and busy discussing internal statistical quality issues, and are usually incapable to propose for users to set negative priorities. A key to improve statistics in this respect is to enhance the communication with users. One future challenge for ESS lies in this. 28. As also mentioned above, many national statistical institutes have experienced big cuts in their budgets due to the financial crisis s impact on government finances. This is very unfortunate, but it is also a challenge for NSIs to convince politicians that statistics is part of the solution and should not be cut on the contrary statistics should be boosted to provide more, better adequate information to help prevent new crises and to help explain reasons for and consequences of the current crisis. 29. The political dimension of parts of the ESS was known before the financial crisis, and confirmed by the political pressure on ESS in particular related to government finance statistics, i.e. the aforementioned EDP. Pressures towards changing the rules of national accounts should never be accepted unless there are good reasons for this. It is the role of Eurostat in cooperation with the other members of the ESS to develop statistics to reflect changes in the societies, economies and environments. The fact that some statistics are used for administrative purposes should not change the way these statistics are developed. In current ongoing and very important work on the review of the ESA95, which in time coincidence with the financial crisis seems to take place in a sound environment without political interference. If administrative users get both the power of setting the agenda and premises for the development for statistics, the alarm bells should ring. If statistical systems cannot be safeguarded for such reasons, it would be necessary to discuss if other monitoring systems should replace the current role of ESA95. But, to put things in a more positive way: if the political processes are reflecting important developments in the societies and economies, there is also a need for the statistical system to respond to this development. Statisticians in Europe are not passive spectators, but are given an important role to meet the need of a complex and dynamic reality. Given this importance of statistics in Europe, it should not be a surprise, even for statisticians that the statistical system is subject to the interest of not only statisticians. Political pressure means political importance. 30. To conclude; first and in general, would it be fair to criticise the international statistical community that a system was not in place to give early warnings on what was building 7

8 up? The answer would be mostly no, since the problem was rather in the accounting books of the companies that did not reflect the real risks. Often were these risky investments linked to e.g. asset backed securities not consolidated in the accounts of the companies but rather off balance sheet in structured investment vehicles often abroad. But, some statistics were available but not highlighted, such as balance sheets of the private sector. It can also be criticised for not delivering full balance sheets for households and good price statistics for the housing enabling the possibility to monitor the changes in net wealth of households. The statistical system as described in ESA95 foresees this, but it is not reported to Eurostat by the Member States. 31. Second and in particular, to what extent was and still is the focus and response of statistics in Europe to the crisis affected by the linkages to the political framework and processes in the EU? The focus of Eurostat can be questioned, as all the efforts seem to have been in the direction of government finances more specifically on the so-called EDP and the SGP. It seems that it was more important to agree on the recording of the government interventions by stretching the methodological framework as much as possible to avoid questions about the SGP. It can be questioned if this is not very prudent by trying to make government finances to look as good as possible. It is also questionable if it can be justified to decide on rules that will apply only to the time of the financial crisis. Moreover, what would be the criteria be for when the financial crisis is over? Accordingly it could be asked when the financial crisis really started. Did it start in autumn 2008 with the collapse of Lehman Brothers, or did it start when banks and other financial institutions started their hazardous investments? 32. What would be the indicators for Eurostat to say that the financial crisis is over, and will it be over at the same time in all countries? Many banks are now reporting surpluses, maybe because of interventions of governments and central banks. Does that mean that the crisis is over? It seems more important for Eurostat to consolidate the functioning of the EU, and in this respect it seems that it has done a good job. 33. One last remark concerns that crises can create opportunities. One might wonder why the European statistical community has not tried to initiate a debate on what statistics are needed for the future. It is true that Eurostat has been involved in the Interagency group as already mentioned in the paper, but the European statistical community in its entirety has not been activated except to solve the statistical recording of government rescue interventions. In this sense a two step strategy could be envisaged. First, it should be looked into carefully whether information in already existing statistics could be utilised better and if the quality of these statistics are satisfactory, and then in a second step the need for new statistics it could be looked into. In addition, it seems of utmost importance to enhance the cooperation with main users. Statisticians are hopefully those most knowledgeable in making statistics, but are maybe not those most knowledgeable about what kind of statistics are most needed. Therefore to cite the wisdom from economic theory - it is important to do the things right, but it is also important to do the right things! 8