Publication revision National accounts, reporting period 2021 erratum

3. Background and reason for the revision

No large-scale introduction of new international guidelines apply in this revision for reporting year 2021; the statistics are still compiled according to the same European System of Accounts (ESA 2010). However, this revision does introduce some methodological changes that better describe sections of the Dutch economy within the existing guidelines. These are partly due to additional European guidelines on interpreting ESA 2010.
For instance, this revision is in line with an updated version of the European Manual on Government Deficit and Debt (MGDD 2022).

The revision adjustments are spread across all economic activities and across almost all transactions, assets and liabilities. Nevertheless, many of the adjustments can be explained based on four overarching themes: revisions based on recommendations or action points following verification of the figures by Eurostat and ECB, revisions resulting from using new sources, revisions resulting from realignment with existing sources and revisions resulting from reclassifications of enterprises and institutions.

3.1 Data verification by the European Commission (Eurostat and ECB)

Because of the importance of macroeconomic statistics, extensive checks are carried out on the figures. Recommendations or action points can flow from this, which must be implemented during the revision. The national accounts were reviewed extensively recently in the framework of GNI, EDP and MIP verification. These verifications are explained below.

3.1.1 GNI verification

The level of gross national income (GNI) of EU member states is used to determine each member state’s contributions to the EU. The European Commission (Eurostat) therefore conducts extensive verification of the correct use of ESA 2010 to determine GNI. This verification has led to two types of action points to improve the Dutch methodology and calculations.

First, it is possible that areas of the economy are identified in which the methods used by different member states are not sufficiently comparable to enable GNI to be determined. European agreements are made regarding the desired method with respect to these areas, to which all member states are obliged to adhere.

Second, the estimates of individual member states’ national accounts are reviewed extensively in terms of data, methods and calculations used. Part of this review included an information visit by Eurostat to CBS in July 2023.

Together, this resulted in six improvement areas for the Dutch figures. These have been incorporated in this revision.

  • The methodology used to determine trading margins on financial assets has been aligned with newly developed guidelines;
  • For a select group of multinational companies, the registration of cross-border transactions between divisions of these companies was compared with those of other member states and, where necessary, the registration was brought in line with those of other member states;
  • New level estimates have been produced for areas of the economy for which the current estimates are older than five years;
  • The remuneration received by Dutch residents as employees of international organisations was added to the remuneration received by employees from abroad;
  • In health and welfare, the cost of bank services was added to intermediate consumption;
  • Activities undertaken by non-resident leaders of Dutch criminal organisations are no longer included in the Dutch national accounts.

As the previous revision is six years old, the third action point in particular, of updating level estimates older than five years, led to a large number of adjustments. This is partly because these new estimates regularly involve changes of method. This is the case for housing services, software investment and cross-border employment.

3.1.2 EDP verification

European Union member states must meet standards for public deficit and government debt. Deficit and debt are determined in the framework of the national accounts. As these figures are of major importance, Eurostat conducts extensive verifications of government figures. This includes making information visits, known as EDP dialogue visits, to member states once every two years. An information visit may result in a member state adjusting the registration of certain transactions. If the adjustment leads to a significant revision of government deficit and/or government debt, the figures should generally be revised immediately. If the impact on the general government balance and debt are limited, the adjustments can be incorporated in the revision of the national accounts.

This revision included the following adjustments:

  • The energy tax credit, in which a fixed amount per electricity connection is no longer netted against taxes, but is recorded separately as a subsidy or income transfer;
  • The extraction of natural gas by Energie Beheer Nederland is included as an activity within the government sector;
  • The student transport pass is recorded as government consumption;
  • The waste management contribution paid by companies is registered as tax;
  • Statutory social security contributions deducted from benefits by social security funds were previously netted against benefits. Starting from this revision, the benefits and associated statutory social security contributions will be recorded as gross proceeds;
  • The registration of capital contributions to multilateral development banks takes place when the Netherlands commits to an unconditional capital injection;
  • Project developers’ contributions to municipalities for public utilities are recorded as an asset tax;
  • Various institutions have been included in the general government sector, which has increased government consumption (see also Section 3.4).

With respect to GDP and GNI, particularly the first item on energy taxation led to a major adjustment. Section 4.1 addresses this further.

3.1.3 MIP verification

The macroeconomic imbalances procedure (MIP) is a set of European regulations that aim to prevent and correct risky economic developments in European member states, such as large current account deficits, or high private sector debt. In the framework of this procedure, the ECB and Eurostat conduct analyses of member states’ national accounts and balance of payments. In this context, a visit to DNB and CBS took place in December 2023, during which the used sources and methods were reviewed. Based on action points arising from this visit, the following adjustments were made during the revision:

  • All household-owned financial holding companies are included in the figures;
  • The Bank for International Settlements (BIS) statistics were used for deposits held by households with foreign banks;
  • ECB statistics were used for securities held by households with foreign management companies in other euro area countries.

3.2 New sources

The data sources on which economic figures are based change over time: some cease to exist and new sources are also continuously being added. These new sources are already used to calculate macroeconomic figures wherever possible, especially to determine growth rates. However, in a revision, the new sources are fully incorporated into the national accounts and the figures from the new sources are also incorporated. Updated versions of the sources are also used where available, and the latest statistical insights are applied to the sources to better reflect the concepts used in the national accounts.

In recent years, various new sources have become available to both CBS and DNB. Several of these new sources were used for the first time for this revision. Various other new sources had started to be used to calculate developments in the economy in recent years, but in accordance with the revision policy, these were used for the first time in this revision to determine the new levels.

  • European obligations for supplying statistics on companies was expanded as of reporting year 2021. From this reporting year, statistics on all market producers in private education, health and welfare, and culture, sports and recreational services must be provided in accordance with the new European Business Statistics (EBS) regulation. CBS has therefore prepared new statistics, with data on these companies’ profit and loss accounts; the production statistics.
  • As part of the collaboration between CBS and DNB, DNB has prepared survey-based statistics over the past few years for various types of financial institutions that, until then, had not yet been collected or had only been collected to a limited extent. This concerns sections of the other financial intermediaries, financial support companies and intra-group financial institutions and lenders sectors, for which annual reports and models were previously used.
  • From this revision onwards, corporate tax declarations are used to produce figures for financial and non-financial holding companies for which no proprietary statistics are available within CBS and DNB. As this source was already being used for other statistics within CBS, this concerns and additional use of this source.
  • New and improved data have been used with respect to the holding and issuance of securities, such as listed shares and bonds. DNB has removed reporting on securities, which includes listed shares and bonds, from the existing statistics and has established a more comprehensive autonomous survey; Monthly Securities Reporting (MER). Data from the ECB’s Centralised Securities Database (CSDB), a reference database, were also used more extensively for securities. This database provides information on securities issued. This information will also be used to determine the balance sheet value of the issued securities.
  • Several adapted CBS and DNB sources were also used, including the updated Statistics of Finances of Enterprises (SFO); a joint CBS and DNB survey for non-financial corporations (NFCs). This combines previous separate CBS and DNB surveys on non-financial enterprises. Another revised source is DNB’s Macroeconomic Statistics Reporting (MESREP), which in parts provides improved monitoring of financial sectors than previous sources.
  • One of the external sources used is the Employee Insurance Agency’s (UWV) policy administration. The policy administration includes such things as all payroll tax returns from employers in the Netherlands. Since the previous revision, this source has been expanded to include a variable that enables severance and transition payments to be reported separately. This additional variable replaces a model that was used previously for these remunerations.

3.3 Alignment with existing sources

Even for those parts of the economy where no new sources or methods have been introduced, the revision can sometimes lead to substantial adjustments as a result of realignment with the levels and structure of the source statistics. The main focus when compiling national accounts in the years between revisions is to describe macroeconomic developments, such as GDP growth. This means that it is not always possible to align with the current figures due to, for instance, series breaks, methodological changes, or a new data collection design. Over time, this may result in a divergence between the levels published in the national accounts and the source levels. This has been rectified in this revision.

There are multiple reasons why re-alignment with the source can result in adjustments. First, following further analyses of the definitive economic figures that have already been published, irregularities are sometimes found in the source data in the national accounts, or in the processing of these data. Such minor irregularities that surface after the publication of the final figures of the national accounts and that only have a limited impact on the economic picture, are generally only dealt with during a revision. During Eurostat’s verification of the figures, the rule applies that known errors greater than 0.1 percent of GNI must be adjusted in the interim. For more minor errors, there is an option to wait until the revision. With GNI at around €900 billion in 2021, this means that errors up to approximately €900 million will in some cases only be corrected during the revision.

Second, new insights can emerge regarding how to improve the consistency of various sources. There can be major differences between the sources, even when describing the same phenomenon. This occurs, for example, when output and domestic consumption of goods or services do not match the trade balance derived from international trade statistics. Choices were made during the 2015 revision regarding the corrections that needed to be made to ensure consistency of figures from the various sources.

These are known as integration adjustments and are automatically incorporated in subsequent years to ensure comparability between the years. An integration adjustment in reporting year 2015 will result in a difference between the source and the national accounts until the next revision (reporting year 2021). Alignment with the source takes place as far as possible during a revision.

Finally, for some sections of the economy, the source used is not available in time each year. This is the case for the Budget Study, for instance, which surveys household expenditure. As this statistic is only produced once every five years, an extrapolation method is used for the interim years. Alignment with the source when this becomes available can sometimes lead to major adjustments. This issue is closely related to the action point from the GNI verification, in which the basis for an extrapolation must not be older than five years.

3.4 Reclassifications of enterprises and institutions

The national accounts use two systems to classify enterprises and institutions: one by type of economic activity, known as the NACE Rev 2 classification, and one by institutional sector. Both are maintained in the CBS general business register (GBR). Classification according to economic activity makes a distinction between economic sectors such as manufacturing, trade or business services. Classification according to sector distinguishes between, for example, non-financial corporations (NFCs), government and households (self-employed persons). For example, a company is then classified as a non-financial institution in business services.

There has been a considerable focus in recent years particularly on improving the sector classification in the GBR, partly based on European laws with respect to statistical enterprise registers. These improvements are being implemented in one go in the current revision.

EBS’s new delivery obligation for market producers in several service types (see new sources) has increased the importance of distinguishing between market producers and non-market producers. There is actually no need to provide statistics on non-market producers. In response to this, an improved method was developed to distinguish between market and non-market producers. In this revision, this has resulted in reclassifications between enterprises (market producers) and institutions (non-market producers). During this process, the number of institutions increased on balance. In accordance with international guidelines, as the output of non-market producers is determined based on the sum-of-the-costs rather than market prices, this also affects macroeconomic totals.

As part of the collaboration between CBS and DNB, there has been a major focus in recent years on further alignment of the distinction between financial and non-financial corporations (NFCs). This included research on separate registration of financial business units from the non-financial corporations (NFCs) of which they are part. In some companies, this has led to including financial business units separately in the statistics, while in others, financial business units have instead been merged with the non-financial company to which they belong. Reclassifications of business units between sub-sectors of financial institutions have also been made within the financial institutions themselves.

In the framework of the revision, approximately 400 companies and institutions have been transferred to the government sector. These transfers took place based on analyses of the extent to which these institutions’ income originated from the government and on the government’s control over general policy.

Finally, economic activities were also classified differently in the revision, without changing sectors. This was the case for sheltered employment facilities, for example. Prior to this revision, sheltered employment facilities were accounted for in the manufacturing sector. From this revision onwards, these workplaces are included in general public administration activities.