3.20. This development is contrary to the objectives behind the GNP resource, namely the creation of a resource based on reliable and exhaustive data that are comparable from both the conceptual and practical points of view. It offends against the principle of the impartiality of Community statistics (29), which implies that they should be produced in an objective and independent way. The GNP resource is therefore in danger of becoming a purely financial contribution, ending up as a formula for budgetary cost-sharing among the Member States. If it is not based on the most refined calculation of the GNP (the 1995 version of the ESA), this GNP resource will introduce a bias in the measurement of the contributory capacity of the Member States. The Commission nevertheless intends (30) to accompany in future any proposal aimed at modifying the statistical definition of the GNP used for the calculation of own resources with an estimate in percentage terms of the difference between the GNP according to the 1995 ESA and the GNP according to the 1979 ESA taken into account for the own resources.
3.21. As a result, and in connection with the considerations under 3.9. above, the existence of two GNPs (relating to the 1979 and 1995 ESAs respectively) poses the problem of deciding which aggregate is to be used in other Community fields, such as, for example, decisions on which areas are to be eligible for structural measures.
3.22. Furthermore, calculating the GNP twice does not make for any qualitative improvements in the national accounts. On the contrary, it represents an inefficient use of available resources and is contrary to the general principles of Community statistics.
Correcting budgetary imbalances
3.23. Whilst considering expenditure policy to be the main tool for the eventual solution of the problem of budgetary imbalances, the Fontainebleau European Council of 25 and 26 June 1984 nonetheless decided that any Member State that was bearing excessive budgetary costs in relation to its relative prosperity could benefit from a correction (31).