Welcome to the dark side: GDP revision and the non-observed economy

with Pia Huettl

Back in 2009, the United Nations Statistical Commission endorsed a revision to the System of National Accounts (SNA), which sets the international standards for the compilation of national accounts. As a consequence, Eurostat has amended the European equivalent of the SNA, the European System of Accounts (ESA) leading to a revision of GDP figures.

The changes come from the accounting treatment of some items. Research & Development (R&D) purchases and military weapon systems have been reclassified from intermediate consumption to investments, which increases value added (the difference between output and intermediate consumption), and thus GDP. Additional changes have been introduced in the accounting of pension entitlements, directly affecting the computation of compensation of employees and households’ savings rate. Other measures, such as changes in measurement of financial services, and the classification of head offices, holding companies and Special Purpose Entities, have little or no impact on the GDP numbers.


Figure 1 shows the average difference between GDP computed with the new and the old standard, retrospectively over the period 2000-2013. The reclassification has had a positive effect on GDP, increasing it on average by 3.5 percentage points for the EU and the Euro area as whole. Country variation is however significant; the impact of the reclassification ranges from 0.3 percentage points in Luxembourg to 9.3 percentage points in Cyprus. Although the revision may have had a visible impact on GDP levels, growth rates are generally less affected.

Unfortunately, Eurostat does not provide a breakdown of how the different accounting changes contribute to the final number. However, the OECD published this month a report disentangling the effect of the different factors for all OECD countries in year 2010 (Figure 2).



Having a breakdown is important because most countries have used the opportunity of the changeover in standards to also introduce a new statistical benchmark estimate, introducing new sources and methods. The OECD report shows that in some countries this has an important impact, most notably in the Netherlands and the UK (see the light red bar in Figure 2).

Other than that, R&D reclassifications tend to be the item with the largest impact on GDP recalculation, whereas reclassification of military weapon systems has very limited impact, with the exception of Greece. In Europe, the impact of R&D reclassification on 2010 GDP ranges between 0.5 percentage points and 4 percentage points, compared to an OECD average of 2.2.  The effect of the change is highest in Finland and Sweden – which have among the highest gross spending levels on R&D in EU – and lowest in Poland, the Slovak Republic, Luxembourg and Greece (Figure 3).


The impact of including illegal activities varies across countries. Both the new and the old standards for the compilation of national accounts stated that illegal activities should be included in GDP, but many countries did not explicitly include estimates for these activities, also because the definition of illegal activities was only streamlined by the decision in September 2014. In contrast with the previous system, EU States are now required to comply with common methodological guidelines how to account for prostitution, the production and trafficking of drugs and alcohol and tobacco

This decision was expected to have a differentiated impact across countries, and the OECD breakdown of GDP change between the old and new system make clear the extent of this divergence.

In Figure 3, Italy and Spain stand out as two special cases. The inclusion of estimated illegal activities results in an increase in 2010 GDP by 1 percentage point in Italy and 0.9 percentage points in Spain. This is about five times the average for the OECD as a whole, which was 0.2 in 2010. Perhaps even more striking is the fact that in these two countries the impact of including illegal activities is only slightly smaller than the impact of reclassifying R&D (which increased 2010 GDP by 1.3 pp in Italy and 1.2 pp in Spain).

Apart from illegal activities, GDP numbers can also be influenced by the share of countries’ ‘legal’ shadow economies, defined as all market-based legal production of goods and services which are deliberately concealed from public authorities.

A report on the shadow economy in Europe shows that its size reached a 10-year low in 2013. Several reasons might explain this development: improving economic conditions, the dramatic contraction of the construction-sector in several crisis-hit countries (which is historically a sector with a large shadow economy), a crack-down on tax evasion in recent years and generally stricter law enforcement.


Interesting to note is the comparatively small size of the shadow economy in Western Europe, with Southern and Eastern Europe exhibiting substantially higher shadow economy activities in % of GDP. The euro area is characterized by a great heterogeneity, with the size of the shadow economy in % of GDP in 2013 varying from just 7.5% in Austria to 27.6% in Estonia. In absolute numbers, the shadow economies in Italy and Germany were as large as Austrian GDP in 2013. Statistical offices make adjustments to account for shadow economy in order to arrive at exhaustive estimates of GDP and national accounts (see OECD)

The rationale for streamlining guidelines for both the inclusion of legal and illegal activities in the national accounting – which has been extensively debated in the media – is that GDP is supposed to be comparable across countries, independently of differences in national law. As recently pointed out by the OECD, international comparability has become even more important, as contributions to international organisations (including e.g. the European Union) are based on levels of Gross National Income, which is linked to GDP.


This piece was also published at Bruegel



Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s