Consistency between national accounts and balance of payments statistics

Transcription

1 Consistency between national accounts and balance of payments statistics Statistics Explained Data extracted in April Planned article update: September Absolute discrepancies in the European non-financial accounts over time (selected quarters), sum of EU-28 Member States, (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) Author: Robert Obrzut In 2014, the process of harmonising the methodological standards for compiling European national accounts and balance of payments statistics (BOP) came to an end. Hence, in applying both methodologies, the European System of Accounts 2010 (ESA2010) and the Balance of Payments and International Investment Position Manual in its 6th edition (BPM6) suggest a high degree of comparability and consistency between the rest of the world account (ROW) of national accounts and BOP statistics 1. As a consequence, both statistics should lead the user to the same conclusions. Where this is not the case, the interpretation of data becomes quite difficult, as diverging conclusions would generally question the reliability of these statistics. Such discrepancies could arise from e.g. the application of uncoordinated compilation practices (including different interpretations of the standards), from different access to data as well as from different use of data sources and primary statistics. Depending on the national peculiarities in organising statistical compilation processes, some countries could be more concerned than others. However, consistency between the two statistics should not be overemphasised as a measure for statistical accuracy. There are more dimensions for a comprehensive assessment of statistical quality. 1 BPM6 Appendix 7, ESA2010 Chapter 18 Source : Statistics Explained ( - 18/05/2018 1

2 Eurostat systematically analyses any inconsistencies when they occur, based on quality reports, compiler surveys or regular monitoring exercises and emphasises the importance of comparable statistics in Europe, both at national and international level. In this report Eurostat informs about the current state of consistency in the non-financial accounts of the EU-28 Member States 2. This analysis is exclusively based on available quarterly statistics, where systematic vintage effects are expected to be less prevailing due to the high update frequency of the underlying statistics. Services are most exposed to discrepancies Current measures (Table 1) confirm continuously high inconsistencies in some components of the non-financial accounts of the EU-28, but with some improvements between 2012 and Consolidated total absolute discrepancies in the EU-28 3 assumed EUR 233 billion for 2016 (i.e. 1.6 % of EU-28 GDP) and 198 billion for 2017 (i.e. 1.3 % of EU-28 GDP), while in October 2015 general levels of inconsistencies were considerably higher between EUR 250 to 300 billion. In relative terms the extent of inconsistencies remained between 1.3 % and 1.7 % of EU-28 GDP. In particular, the services accounts still show elevated discrepancies over time, followed by goods and the primary income. Otherwise, secondary income as well as the capital account remain only moderately exposed in absolute terms. It appears noteworthy that the consistency of the primary income accounts has experienced considerable improvements in the recent past in both statistics. Table 1: Absolute exposure to discrepancies, sum of national data, non-financial account, by BOP item, (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) High concentration of inconsistencies around six Member States A look at the underlying country data reveals a highly diversified picture about the geographical breakdown of discrepancies in the EU-28 (Figure 1). Depending on their exposure to components of the non-financial accounts, six Member States showed the highest absolute discrepancies within the EU-28, contributing together to currently 80 % of all measured discrepancies. This applies to France, Luxembourg, Germany, Poland, Portugal 2 The ECB will launch similar initiatives concerning the financial accounts. 3 We sum up the measured absolute discrepancies of Member States to a consolidated EU total. For analytical purposes this helps to monitor developments in the underlying data seen from a European rather than national perspective. Consistency between national accounts and balance of payments statistics 2

3 and Greece. Almost 43 % of mean discrepancies occurring during can be currently attributed to France alone, which is the major contributor to inconsistencies at this moment in the EU-28 in absolute terms. On the other hand a few Member States show no or only low inconsistencies in their non-financial accounts: United Kingdom, Cyprus, Denmark, Estonia and Latvia. Figure 1: Mean annual discrepancies, non-financial accounts, by Member State, (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) The measured discrepancies in the components of the non-financial accounts vary among the most exposed countries. While France and Luxembourg show high levels of discrepancies in services, Germany faces most of its inconsistencies in the primary and secondary income accounts, and Poland in its secondary income accounts. France, Greece and Portugal also show elevated discrepancies in their goods account. However, in regard to the countries GDP, these discrepancies appear relatively smaller, except for Luxembourg (Figure 2). A particular situation applies to Portugal where the discrepancies in goods and services are almost equal, pointing at systematic reclassifications between these two components for the sake of accommodating the conflicting consistency requirements of the country s Input-Output tables. This applies to some lesser extent also to Greece. Consistency between national accounts and balance of payments statistics 3

4 Figure 2: Mean annual discrepancies, non-financial accounts, by Member State, (% of GDP) Source: Eurostat (bopc6q) and (nasq10nftr) Revisions have effectively contributed to reduce discrepancies Revisions clearly play an important role in making both statistics more consistent. According to Figure 3 there was a significant downward shift in overall discrepancy levels during the past years in the EU-28. Between October 2015 and April 2017 the most comprehensive revisions took place over the entire period Ever since, compilers dedicated their attention to the more recent data with considerable improvements for In absolute terms, overall levels of discrepancies in the non-financial accounts dropped in their multi-annual average from around EUR 274 billion in October 2015 to EUR 207 billion in April This is an overall improvement by around 25 % since October 2015, reducing the overall relative share of discrepancies from around 3 % to below 2 % of the EU-28 GDP. Consistency between national accounts and balance of payments statistics 4

5 Figure 3: Absolute discrepancies in the European non-financial accounts over time (selected quarters), sum of EU-28 Member States, (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) When comparing the evolution of consistency related to the 2014 vintage, the majority of the EU-28 Member States (at least 24 Member States) managed to achieve improvements in their levels of discrepancies since October 2015 (Table 2). Total levels of discrepancies were on average reduced by almost 50 % in the Member States, which illustrates the ambitious work that European compilers are currently dedicating to this issue. Some countries generally achieved significant reductions in their discrepancies - among those were most notably the Netherlands, Germany, Belgium and Greece. Consistency was particularly improved in the goods accounts of Germany, Greece and Hungary, the services accounts of France, Germany and Belgium, the primary income accounts of the Netherlands and Belgium, and the secondary income accounts of France and Ireland. Germany, Greece and Italy showed significant improvements in their capital accounts. As the two most notable examples, the Netherlands reduced their discrepancies in primary income since October 2015 by more than EUR 66 billion, and Belgium (once a main contributor particularly to inconsistent primary income accounts) has produced a highly consistent primary income account for As a consequence, we may currently conclude on a general trend for ongoing improvements in all components; more recently the most promising in the primary income. In at least 2 countries, however, no sustainable trend towards improvements could be established so far since October 2015 (Luxembourg and Lithuania). While Luxembourg s deteriorating situation relates particularly to the services account, Lithuania s secondary income account appear affected by increasing inconsistencies ever since. Consistency between national accounts and balance of payments statistics 5

6 Table 2: The impact of revisions, by component of non-financial accounts and Member States, 2014 (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) Directional inconsistencies - severe limitations to statistical comparability in some Member States Opposite signs in the account balances of both statistics pose a considerable challenge to the interpretation of the data by the user, and have a severe impact on political conclusions: in Table 3 we compare the balances 4 of the non-financial accounts in both statistics for From these statistics France produced a tiny net surplus in BOP services, but a net deficit of EUR 14.7 billion in ROW services (provisional data for 2017: BOP net surplus of EUR 17.9 billion against ROW net deficit of EUR 14.6 billion). Poland reported a negative balance in its BOP secondary income account of EUR 1.4 billion, but a positive balance in the corresponding ROW accounts of EUR 2.5 billion in 2016, while Germany reported a capital account net surplus for 2016 (EUR 3.5 billion) in its BOP and a net deficit (EUR -1.1 billion) in its corresponding sector account balance. Slight inconsistencies occurred also to Greece in its primary income for BOP balance = BOP credits minus BOP debits; ROW balance = ROW payables minus ROW receivables. Quarterly data have been annualised for easier reading. Consistency between national accounts and balance of payments statistics 6

7 Table 3: Balances of components, BOP and ROW non-financial accounts, 2016 (EUR million) Source: Eurostat (bopc6q) and (nasq10nftr) Source data for tables and graphs Consistency NA BOP: tables and figures Data sources Eurostat monitors developments in consistency between BOP and National Accounts statistics with regular data comparisons of quarterly BOP (QBOP) with the ROW of the quarterly sector accounts (QSA). According to the methodological standards these data are fully comparable and should be reconciled with little or no inconsistencies. The time span of the analysis ( ) was chosen due to the broad availability of BOP time series, which were compiled in regard to the BPM6 standard. Meanwhile provisional data for 2017 became available which have been included in the analysis. For easier interpretation all quarterly data have been annualised. The analysis focuses on gross transactions data, allowing discrepancies to be identified to BOP credit/row payables and BOP debit/row receivables respectively. In the lack of underlying gross transactions in one component for the capital account net transactions are compared. The analysis unfortunately has to cope with a few data gaps, as some Member States at this stage do not publish quarterly sector accounts, or others have not been validated due to quality concerns. Austria currently only produces annual sector accounts. Luxembourg does not report quarterly data on primary and secondary income, or on capital account to the QSA, but recently has started to report these data in its annual sector accounts. To a lesser extent, gaps apply also to the QSA data of Malta as regards the secondary income account, where neither annual nor quarterly time series are currently available. A full set of quarterly data for 2017 are currently missing in the sector accounts of Czech Republic, Poland and Romania. The QSA data of Croatia for have been withheld from dissemination due to quality concerns by Eurostat. In the current release the Netherlands have so far reported only estimations figures for primary income of its quarterly sector accounts, Consistency between national accounts and balance of payments statistics 7

8 which appear incomparable with the corresponding BOP figures, and thus have been excluded from the analysis. The United Kingdom has not sent any comparable data during this production round in the sector accounts. The underlying principles of reconciliation in the non-financial accounts have been documented and refer to the suggestions of the methodological standards (Table 4) 5.This analysis is based on the data publications of April 2018 and it is thus implicitly assumed that the published time series are directly comparable in all Member States. QSA data usually become available three weeks after the QBOP data release. Due to the different publication dates, revision and vintage effects cannot be entirely excluded. Table 4: Reconciling the current/capital (BOP) and the ROW non-financial account (QSA)Source: BPM6, ESA2010 Context The debate among statisticians why discrepancies between BOP and National Accounts persist although the methodological standards clearly suggest full consistency, has drawn analysts to the causes of inconsistencies. The following causes have been identified that appear essential to the statistical production processes in the EU-28. They always apply in a country-specific situation, and thus cannot be generalised: The organisational setting of national compilation processes plays a prominent role in explaining the occurrence of inconsistencies. Decentralised statistical compilation systems lead to institutional coordination issues, which have not been adequately addressed in some Member States due institutional autonomies. Further, this has also resulted in different interpretation of the methodological standards. Different access to (micro) data sources or source statistics generates discrepancies, in particular for items that can be measured from a heterogeneous spectrum of data sources. It has been further shown that contagion effects arising from different (vintages of) source data, could import inconsistencies into the final statistical product (e.g. financial data for the calculation of investment income). 5 Obrzut, (2016) p.111 Consistency between national accounts and balance of payments statistics 8

9 Items difficult to measure through surveys or administrative data sources are naturally subject to estimations or extrapolations (e.g. Financial Intermediation Services, Indirectly Measured [FISIM]). This paves the way for discrepancies, when applied by more than one counterpart and without coordination. Due to the specific objectives in each statistics and the foregone investment in IT infrastructure, (automatic) compilation systems are less flexible for being redesigned or adapted to new needs. Further, manipulation of underlying compilation processes requires back data revisions, which challenge data stability of longer time series. As a consequence, national counterparts generally appear less inclined to challenge already established and effectively working operational processes, even when their statistical products diverge from each other to some extent. Institutional peculiarities foster discrepancies arising from different delineations of economic sectors (e.g. captive financial institutions, government sector). International organisations can play an important role in clarifying identified issues in a coordinated manner. Different institutional progress in fully adopting the corresponding statistical standards BPM6 and ESA2010 also explained to some extent the occurred discrepancies in the past. International recommendations have so far focused on the consistent treatment of data in BOP and NA, but abstracted from the dilemma that through their recommendations the consistency to other macroeconomic statistics (e.g. Input-Output tables) is challenged. This exposes the compiler to the dilemma of choosing between conflicting priorities (e.g. Portugal). Revision and vintage effects persist as statistical noise due to different publication calendars and revision practices, which hampers full consistency. Consequently, zero absolute discrepancies appear achievable only from fully integrated production systems (e.g. the United Kingdom) or systems with a high degree of coordination (e.g. Cyprus). Eurostat maintains its ambitions to reduce overall discrepancies in the EU-28 Member States National Accounts and BOP statistics and has supported investigations to the fundamental reasons of inconsistencies. It encourages compilers to regard better comparability of their statistical products with high priority. At the same time Eurostat is aware that statistical comparability is not alone guaranteed by fully consistent national data, but also by symmetric data with its major international partners. At this moment European statistics still show high asymmetries within the EU-28 Member States, which pose an additional challenge to the comparability. The European Statistical System as a supranational network can effectively address these issues and contribute to better quality of data from its Member States in the medium-term. Other articles Balance of Payments statistics Quarterly sector accounts - households Quarterly sector accounts - non-financial corporations Sector accounts Building the System of National Accounts Consistency between national accounts and balance of payments statistics 9

10 Publications Trade asymmetries and Consistency between National Accounts and Balance of Payments, in: Selected papers from the 2016 Conference of European Statistics Stakeholders, Luxembourg 2017 Consistency between national accounts and balance of payment statistics an updated view on the non-financial accounts, Luxembourg 2017 Obrzut, R.: Consistency between national accounts and balance of payments statistics, in: Eurona Nr. 1/2016 Eurostat Statistical Working Paper: Consistency between national accounts and balance of payments statistics, Luxembourg 2016 Database Balance of payments - International transactions (BPM6), see: Balance of payments statistics and International investment positions (BPM6) (bopq6) National accounts (ESA 2010), see: Quarterly sector accounts (ESA 2010) (nasq10) Annual sector accounts (ESA 2010) (nasa10) Methodology Balance of Payments and International Investment Position Manual (BPM6) European system of accounts (ESA 2010) From National to European Accounts Legislation Regulation (EC) No 184/2005 of the European Parliament and of the Council of 12 January 2005 on Community statistics concerning balance of payments, international trade in services and foreign direct investment. Commission Regulation (EU) No 555/2012 of 22 June 2012 amending Regulation (EC) No 184/2005 of the European Parliament and of the Council on Community statistics concerning balance of payments, international trade in services and foreign direct investment, as regards the update of data requirements and definitions. Regulation (EU) No 2016/1013 of the European Parliament and of the Council of 8 June 2016 amending Regulation (EC) No 184/2005 on Community statistics concerning balance of payments, international trade in services and foreign direct investment (Text with EEA relevance) Regulation (EU) No 549/2013 of the European Parliament and of the Council of 21 May 2013 on the European system of national and regional accounts in the European Union External links International Monetary Fund Balance of Payments and International Investment Position Manual 6th edition, 2009 European system of accounts (ESA 2010) Notes View this article online at http: // ec. europa. eu/ eurostat/ statistics-explained/ index. php/ Consistency_ between_ national_ accounts_ and_ balance_ of_ payments_ statistics Consistency between national accounts and balance of payments statistics 10