92001E2003

WRITTEN QUESTION E-2003/01

by Robert Goebbels (PSE) to the Commission

(6 July 2001)

Subject: European statistics on per capita income

The European Union, in the shape of Eurostat, regularly produces statistics comparing factors such as per capita income in the different Member States of the European Union. A recent publication compared the differences in income between the Member States of the Union and the applicant countries. With the index for the European average set at 100, the comparison shows that the poorest country in an enlarged Europe would be Bulgaria (index 23), while the richest country would be Luxembourg (index 181). To achieve these results, Eurostat calculates the gross domestic product of each country in purchasing power parity and divides this figure by the number of inhabitants of the country. Is it not the case that this method leads to distorted results, particularly as regards Luxembourg where GDP is the product of the work of residents and non-residents, since the latter are not taken into account by Eurostat when it divides GDP by the number of inhabitants?

A factor which is insignificant in all the other Union countries is highly significant in the case of Luxembourg, where the total workforce of 275 000 (April 2001) includes some 95 000 non-resident frontier workers, in other words more than 37 % of the working population. Should Eurostat not review the methodology it uses for statistics of this kind, which ultimately are inaccurate?

Answer given by Mr Solbes Mira on behalf of the Commission

(18 September 2001)

The per capita gross domestic product (GDP) figures in euros and in purchasing power standards are disseminated quarterly by the statistical office of the European Communities (Eurostat) for the Member States, as well as for applicant and other countries.

Before providing a detailed answer to the Honourable Member's question, it would be useful to clarify some of the basic concepts concerning both GDP - which is the focus of the question - and gross national income, which is better known as gross national product.

GDP is the final result of the production activity of a country's producer units. It can be calculated from various viewpoints, the best known of which is the expenditure angle, defined as the sum of household consumption, general government consumption, gross fixed capital formation and the external trade balance.

The national income concept, on the other hand, relates to income rather than production and is obtained by deducting primary income paid to non-residents from GDP and raising GDP by the primary income received from the rest of the world.

In international comparisons, the main indicator used to measure wealth is, by convention, GDP which, by definition, is produced on the economic territory of a country either by residents or by non-residents. It is true that on the basis of this indicator, Luxembourg is the richest country in the world with an index as high as 190 (15=100) for 2000.

This can be explained on various counts:

- Luxembourg is a special case since it is defined, on account of its size, as a single region. As with all regions which are very strong poles of economic activity, it attracts a great number of commuters who make a very significant contribution to its GDP (some other regions in the EU are in the same position and the 1998 figures for regional GDP in two of these - Hamburg and Inner London - were 12 and 67 points higher respectively than in Luxembourg);

- Luxembourg's GDP is boosted by the significant volume of purchases, particularly of consumer goods, made on its territory by these commuters and other inhabitants of its neighbouring countries. These purchases are to be considered as exports and thus form part of GDP; they also benefit the State's coffers through the taxes levied on them;

- Given that the general level of prices in Luxembourg is only slightly higher than the Community average, per capita GDP expressed in purchasing power standards (PPS), where purchasing power parities are therefore used as conversion rates, is virtually the same as the GDP in euros, whereas in general when the euro index is high, the volume index is low (in Denmark for example, the index for per capita GDP in euros in 2000 stood at 146,4 but fell to 120,2 in PPS on account of the very high level of prices in this country).

With regard to the matter of whether Eurostat should revise its methodology for this type of statistics, the Commission is of the opinion that this is not possible, primarily for the following reason:

- Statistical data are compiled in accordance with international standards and conventions (the European System of Integrated Economic Accounts (ESA 95) in the EU, and the System of National Accounts (SNA) on the world stage) which cannot take the specific situations in each individual country into account.

- These standards are often drawn up after lengthy discussions among national experts resulting in Council regulations which are usually the fruit of hard-won compromises, especially in cases where they have financial implications for the Member States (e.g. their contribution to the Community budget).

- One point of contention at the moment, for example, is the treatment of financial intermediation services indirectly measured (FISIM), which certain Member States, headed by Luxembourg, would like to see broken down by user sector and final use of GDP (which would lead

to a further substantial increase in Luxembourg's GDP), whilst others oppose this type of breakdown. It should also be pointed out in this context that Luxembourg has, for some time now, published two versions of its national accounts: a national version which takes FISIM into account and a Community version which showed GDP to be around 10-15 % lower, depending on the data, than in the national version.

- To return to the use of GDP as an indicator for international comparisons, if the concept of national income (which is closer to the actual notion of income) were used in its place, there would be no fundamental change in the situation for Luxembourg, as its national income would only be slightly lower than its GDP.

- The Commission is, however, attentive to the question asked by the Honourable Member, which raises a fundamental problem relating to statistics: these are based on standards and conventions which apply internationally to all countries and thus they cannot give total coverage to all aspects of economic and social life.