The EU Budget - Public Perception & Fact

The European Union - how much does it cost, where does the money go and why is it criticised so much?

Terry Wynn, MEP
Chairman of the Committee on Budgets of the European Parliament

Table Of Contents


  1. Revenue

  2. Expenditure

  3. Budgetary balances


  1. The Statement of Assurance

  2. The Court's findings

  3. The problems of the DAS

  4. The system works


The mechanisms of the EU Budget are complex and understood by few people. The revenue or income side is far from transparent for the average taxpayer to understand, and the decision making process on the expenditure side is usually left to those directly involved i.e. the Budget Committees of the European Parliament and the Council of Ministers.

Those who do understand it usually specialise in the subject, and that is one of the reasons why reporting on matters of the EU budget are seldom positive.

Many media articles concerning the EU budget are negative and invariably appear in November of each year, when the European Court of Auditors publish their annual report.

As someone who has been involved with EU budgetary and budget control issues for almost 15 years, I never cease to be amazed at what I read in the press and what I know to be the reality of a situation. Put the words "Europe" and "Fraud" together and a good story is guaranteed.

No one is saying that fraud does not occur in the EU, but does it occur on the scale of each November's headlines (2 bn, 3 bn, 4 bn - take your choice)? Such headlines began in 1995 when the European Court of Auditors (ECA) published the first Statement of Assurance (DAS) for the 1994 Budget year. I was the Rapporteur, or person responsible, in the Budgetary Control Committee for making recommendations to the European Parliament as to whether we should grant the Commission "discharge" or not (i.e. to discharge them of their responsibilities or give an okay to the accounts). It meant analysing the ECA's report with a fine toothcomb and then drawing conclusions from that analysis. The second part of this document tries to give some insight into how the ECA and the DAS reach conclusions and how these are interpreted by the press. It is done in a style that is hopefully understood by those who are prepared to give half an hour of their time to reading it, although the complexities of the EU budget haven't made that an easy task.

The first part attempts to explain how the system operates, and tries to give the positives of the EU budget. No doubt opponents of the CAP will (to my mind quite rightly) say that too much is still spent on Agriculture, but it could be pointed out that CAP expenditure has gone from being 95% of the budget to 45%, and within that amount is included expenditure on such things as environmental improvement and afforestation.

The EU budget is not a wasteful exercise in misusing taxpayers' money, but attempts to be an instrument for bettering the EU and its influence in the wider world.

I hope this booklet helps to explain some of the facts about the EU budget, and at the same time, dispels some if the common perceptions.

Part One:


The European Union budget usually only makes headlines when it comes to bad news. The Court of Auditors' refusal of the Statement of Assurance has quickly found its way on to most European front pages. Fraud and mismanagement are what people associate with Community money.

However, irregularities uncovered by the Court of Auditors, bad as they may be, only concern a small part of the European Union funds. Despite all the complaints, justified or not, one should always keep in mind the "full picture of the EU budget".

Did you know, for example, that

  • Nearly half the money spent to help poor countries comes from the European Union and its Member States, making it the world's biggest aid donor?

The EU and its Member States pay out more than 30 billion euro a year in public aid to developing countries. Of this amount, about 6 billion euro is channelled through the EU. The Union has committed itself to raising the annual total from 30 to 39 billion euro by 2006.

Over the years, the EU has funded thousands of development projects across the third world. Often relatively small amounts of cash go a long way. Recent success stories include a project to equip and train silk weavers in Cambodia, funding to help small businesses in Peru to export more, support for Namibian farmers to set up a lobby to defend their interests, a grant to a business venture in Senegal to improve the quality of local manufactured products, and many more. The ultimate objective of Union policy is to give disadvantaged people in the third world control over their own development.

  • EU structural assistance raised the GNP of Ireland in the late 1990s by 3 to 4 percentage points above what it would otherwise have been? And 213 000 jobs were created on the Emerald Isle with support from the Structural Funds?

Among other countries, Ireland has been a major beneficiary of European Community Regional Policy aid. By doubling the level of Community funding for the Union's less prosperous regions in the period 1989 - 1993 and keeping it at a high level ever since, the European Union with its Structural Funds has had a significant effect on reducing disparities in economic performance across the Union. Thus, Ireland's GNP expanded by 140 percent between 1987 and 2000, compared to an expansion of 40 percent in the US and 35 percent in the EU15 and the UK.

  • In order to cut unemployment, improve job quality, make it easier for people to balance the demands of work with their personal life, and to ensure that everyone has an equal chance of employment, the European Union will spend 60 billion euro between 2000 and 2006?

The European Social Fund provides finance for programmes to develop work skills and social skills.

Special attention is paid to funding for areas of the EU with particularly high levels of unemployment or low average incomes. Almost 300 million euro per year is reserved for the EQUAL programme, which tests new ways of tackling discrimination and inequality experienced by those in work and by job-seekers.

  • The two principal European Union programmes in the field of education, training and youth, Leonardo da Vinci and Socrates spend together around 400m euro annually?

Erasmus, the oldest and probably best known sub-programme under Socrates, has more than 100m euro annually available for grants for students and teachers to spend time at universities in other countries. By the end of 2002, more than one million students had studied at another university thanks to Erasmus. By 2007, that figure should be two million and by 2010 three million.

These examples hopefully show that EU spending might be useful after all.

Very often, when it comes to the question of benefits and advantages arising from EU membership it is only the money that counts. National politicians assure their fellow country men and women that they are trying "to get the most" out of the EU budget. What seems to be most important is not political and economic stability in Europe but rather the question of budgetary balances. Whoever gets most money back from the EU pot wins. Sometimes one gets the impression that every country's main wish is to become a net receiver of EU funds...

The benefits of EU membership should not be measured in money alone, however, it is quite interesting to see how much the European Union actually costs and how the money is spent. EU funds do not disappear somewhere in the Brussels bureaucracy. On the contrary, some 85% per cent of all money spent by the Union is spent in the Member States. For these amounts it can clearly be identified which Member State gets how much money for which group of its population, for which policy area, for which action.

In this context, it is also interesting to note that, contrary to public opinion, EU spending has not skyrocketed from year to year. After a period in the nineties, when payments did go up, reaching a peak in 1993 (1.18 per cent of GNP1), expenditure (as a percentage of GNP/GNI, not in cash terms) has gone down again. The total of payments foreseen in the EU budget for 2004 will only reach 0.98 per cent of GNI, which is the lowest figure since 1990. Similar to developments in the Member States the EU has also drafted a prudent budget - contrary to its spendthrift reputation.

Budgetary balances certainly do not give the full picture of EU membership. They have their shortcomings and inconsistencies but, nevertheless, a closer look at the EU budget will show that the way EU spending is perceived by the public is not always consistent with the facts.

1. Revenue

Before having a closer look at how and where the European Union spends its money, it is also important to get a clearer idea on how the Union is financed. Who pays how much into the EU budget? What are the criteria for defining the Member States' due share? Is the financial burden really distributed as unevenly as people sometimes think?

It is the revenue side which is the starting point for all evaluations of what are budgetary balances - or imbalances.

1.1. Own resources

There has been a long and steady development from the "membership fee" system of the Community's early days to what is now a rather complicated system of own resources. At least in theory, the Union now spends its own money.

1.1.1. History of the own resources system

From 1958 to 1970 the Community budget was financed exclusively by contributions from the Member States. In 1970, the Luxembourg European Council introduced for the first time a system of own resources for the general budget of the Community taking effect in 1971. One objective was to enhance the Community's financial independence from Member States' transfers. The first own resources were customs duties and agricultural levies which were to become known as traditional own resources (TOR).

In 1979, a VAT-linked resource was introduced as an own resource after a base for its assessment in the Member States had been defined. The inability to contain spending on the Common Agricultural Policy (CAP) and two enlargements led the Fontainebleau European Council in 1984 to increase the maximum rate of call for the VAT resource to 1.4 per cent (and to establish a correction mechanism for budgetary imbalances which has since applied only to the UK2).

Since CAP spending still could not be sufficiently controlled and the revenues from TORs continued to decline the Fontainebleau decisions soon proved to be insufficient. In consequence, in June 1988, the Brussels European Council introduced a new own resource based on the Member States' GNP. It also set up an overall ceiling to the total amount of own resources which could be called to finance the Community's spending.

The Edinburgh agreement of December 1992 increased this overall ceiling to 1.27 per cent of Europe's GNP, while at the same time introducing steps to further decrease the importance of the VAT resource. This agreement entered into force at the beginning of 1995.

In 1999, the Berlin European Council called on the Commission to prepare a new Own Resources Decision which should provide the Union with adequate resources for the period 2000 - 2006 while at the same time adhering to strict budgetary discipline. The new system should be "equitable, transparent, cost-effective and simple", and based on criteria which express best the member States' ability to contribute to the financing of the Union. This latest own resources decision of 29 September 2000 entered into force on 1 March 2002.

The revision clause it contains calls on the Commission to undertake a general review of the own resources system before 2006 in the light of the effects of enlargement. The review is also to explore the possibility of modifying the structure of the own resources system by creating new autonomous resources and examining the correction of budgetary imbalances granted to the UK. But of course it will need an unanimous decision of the Council to agree to any changes.

1.1.2. Distribution of the own resources

As we have seen, the own resources system as it is today is made up of three components:

  • Traditional own resources (TOR)
    - Agricultural duties
    - Sugar levies
    - Customs duties
    - (minus collection costs to be retained by Member States of 10% before 2001 and 25% since then)

  • VAT based resource

  • GNI based resource

In 2002, the composition of own resources was the following:

In general, the GNI based resource has become increasingly important accounting for 59.1 per cent of all resources in 2002 (this means the GNI resource doubled its share over a period of six years, having amounted to only 30 per cent in 1996). The share of traditional own resources has shown a trend to decline dramatically accelerated by the fact that the percentage Member States were permitted to retain to compensate for their collection cost was raised from 10% to 25% with effect of 1 January 2001. This has led to a decrease in the amount of the TOR of 25% between 2001 and 2002, when TOR only contributed to own resources with a share of 11.9 per cent. VAT contributions have also declined due to policy reforms limiting the VAT base of the Member States to 50% of their GNI. Thus, VAT contributions in 2002 have gone down to 28.8 per cent.

The development in the structure of own resources is illustrated in the following table:

With the introduction of the GNP resource and the reduction in the VAT call-up rate, Member States' contributions have become more closely correlated with national GNPs; this means improved equity3 in gross budget contributions. Under the current system perfect equity cannot be expected since the VAT resource will continue to yield revenues which will not be correlated with national income.

However, while this development might lead to a closer correlation of the Member States' ability to contribute with their actual levels of contributions, the decline of traditional own resources has resulted in a loss of financial autonomy for the European Union.

Traditional own resources are the only true own resources of the Union (although Member States tend to also regard them as national contributions). The lower their share becomes, the more the EU will depend on intergovernmental transfers. This, in turn, can cause Member States to maximise concepts of the national benefit from the EU budget.

The more Member States (and their citizens) feel it is THEIR resources that are transferred to Brussels, rather than the EU's OWN resources, the more they will concentrate on what they get in return. This is why the concept of "budgetary balances", the roles of "net payers" and "net receivers", has become so important. However, it can often be misleading to concentrate purely on input/output calculations, as we will see later.

1.2. Member States role

In 2002, Member States' payments (these figures include the TOR, following the "traditional" way of depicting Member States contributions) to the EU budget were the following (in € million):

As was already mentioned before, the system of these own resources payments should take into consideration the ability of each Member State to contribute to the EU budget. It has to be noted, however, that the presence of the UK rebate considerably reduces this desired correlation. Its complexity also diminishes the transparency of the whole own resources system. But as long as CAP expenditure remains as it is the United Kingdom will always insist on keeping the rebate.

In the light of this, and in order to be able to see whether these shares are "fair" or "unfair", one also has to have a closer look at the "wealth" of the Member States as expressed by their gross national product.

If you look at percentages of the EU budget and at the relation of the Member States' payments (in million €) to their GNI you arrive at the following picture (2002):

Another interesting way to look at the EU budget is Member States' contributions in relation to their population, i.e. the contribution of the European citizens to the EU budget:

Surprisingly enough, if you look at the Member States' own resources payments in relation to their population, it is the citizens from Luxembourg, Denmark and Belgium who pay the highest share (I bet you anything you'd have thought it was the Germans or the British). Portugal, Greece and Spain have the lowest per capita contributions to the EU budget, which is a bit more of what one would have expected.

  1. As from the year 2002, the concept of gross national product (GNP) has been replaced by the concept of gross national income (GNI) in the EU budgetary and own resources area. Thus, percentages for years before 2002 relate to the GNP, later ones to the GNI. For example: The EU expenditure ceiling of 1.27% of GNP corresponds to 1.24% of GNI using the new statistical approach.

  2. Fontainebleau European Council, Conclusions of the Presidency: "Expenditure policy is ultimately the essential means of resolving the question of budgetary imbalances. However, it has been decided that any Member State sustaining a budgetary burden which is excessive in relation to its relative prosperity may benefit from a correction at the appropriate time".

  3. In the sense of "proportionality of gross contributions to income across Member States"

  4. The balance does not add up to zero because of exchange rate differences


Now that we have had a look at where the European Union gets its money and which payments under which title come from which Member State, let's turn towards the other side of the coin: expenditure. This is where most public criticism is aimed: The European Union is depicted as some kind of black hole, far away in Brussels, which pulls increasing amounts of the Member States' (and their citizens') money into its centre, never to be seen again.

But again, this perception, popular as it may be, is not substantiated by fact. In 2002, some 91%, or 72.7 bn euro, of the EU's total executed operating expenditure was allocated among the Member States. If you exclude expenditure like money spent on external action, reserves, and pre-accession aid, which, by definition, cannot be allocated to a Member State, then 98.8% of (allocable) operating expenditure can be broken down to the Member State level in 2002.

So, the money sent to Brussels doesn't disappear forever in the EU's bureaucratic jungle, nor does expenditure go up unrestrained.

Before the details of EU expenditure are explained, you need to be familiar with the French system of describing expenditure in the EU budget for a given year. It's done in two ways: as commitment appropriations (CA) and as payment appropriations (PA). The difference (official definition) is the following:

2.1. Development of EU expenditure

Contrary to public perception, EU expenditure as a percentage of GNP/GNI has continuously gone down over the last couple of years. Especially the 2004 EU budget which could be called a real "austerity budget" with payments only amounting to 0.98 per cent in relation to GNI, and this although the 2004 budget is the first of its kind for a European Union of 25 Member States.

A look at the development of payments since 1986 shows this increasing trend towards economy:

This chart clearly shows that the EU budgetary authority has acted quite responsibly when it comes to the European Union taxpayers' money. It is not only the Member States that have had to economise, the European Union has done the same.

2.2. The headings of the EU budget

EU expenditure is classified in six different categories:

Heading 1: Common Agricultural Policy (CAP)
Heading 2: Structural operations
Heading 3: Internal policies
Heading 4: External action
Heading 5: Administrative expenditure
Heading 6: Pre-accession aid
(Heading 7: Reserve)

Categories 1, 2, 3, 4 and 6 are operational expenditure, heading 5 covers all administrative expenditure of the European Union institutions.

A multiannual financial framework, a "financial perspective", lays down annual expenditure limits for the different categories. In 1999, as part of the negotiations on "Agenda 2000" relating to the Union's future priorities, a financial perspective for seven years was adopted (covering the period 2000 - 2006).

The following pie chart shows the distribution of EU expenditure between the different headings as foreseen in the budget for 2004 (figures in billion €):

Oh yes, these figures are for commitments of the 2004 budget and since the budget totals some 99.529 billion euro, they also represent more or less the shares of the different headings in percentage points.

2.2.1. CAP

Under the agriculture heading of the EU budget two main actions are financed:

Firstly, there is support to the markets, which takes the form of guaranteed prices, export refunds or contribution to storage costs. This "traditional" CAP must now be fully compatible with sustainable development, in particular by promoting environmentally friendly production methods and the effective use of resources.

Secondly, EU funds are also used to further the development of rural areas. Thus, CAP money is also spent on non-agricultural activities, trying to enhance the competitiveness of rural areas and preserving the environment through, for example, the financing of afforestation activities, etc. The EU's rural development policy supports rural heritage in order to secure the future of rural areas and promote job-maintenance and creation.

Although there has been a slight but constant decrease in CAP expenditure over the years, the agriculture heading of the EU budget is still the biggest spending area, with an amount of 44.761 billion euro in 2004, which represents some 45% of the overall EU budget.

The latest CAP reform, adopted in 2003, is a step in the right direction, with its principles of "decoupling" (establishment of one single farm payment independent from production, i.e. farmers can produce what consumers want and not what is subsidised), "cross-compliance" (payments linked to the respect of environmental, food safety, animal welfare, health and occupational safety standards) and "degression" (reduction in direct payments to bigger farms to generate additional money for rural development). The real effects of this reform will only be seen in the future.

As for the past, how was the agricultural expenditure allocated in 2002 (in million €)?

One can certainly see that, as expected, large agricultural countries like France, Germany, Italy and Spain got the biggest portions of the money in heading 1. The United Kingdom's share is also quite considerable. However, the clear message this graph seems to give has to be considered with some caution.

For example: In the case of agricultural expenditure, export refunds are allocated to the Member State from which the goods leave the Union. This Member State is not necessarily the same as the one in which the producer, and ultimate beneficiary, resides. This practice might lead to an overestimation of payments to Member States with major ports.

Consequently, the distribution of the CAP funds as officially registered is just a rough indication on where and how the major bulk of the money has been spent.

2.2.2. Structural operations

The aim of the EU's structural operations is to give support to less, and lesser, developed regions in the Member States of the European Union. In the course of the period of 2000 - 2006 an amount of 210 billion euro will be spent under the structural operations heading of the EU budget. The structural operations heading comprises payments from the structural funds (on objective 1, 2 and 3 areas) and the cohesion fund as well as other specific structural measures and operations.

By their nature structural operations should result in differences in expenditure between Member States because they aim at achieving an economic and social balance throughout the whole Union.

A look at the expenditure on structural operations in 2002 indeed reveals some differences in the Member States' shares:

One speciality with spending under the structural funds is the fact that it varies considerably from year to year. Payments to Member States from heading 2 of the budget depend on a lot of factors such as how late in the budgetary year Member States have introduced requests for funding, etc.

For example: If you compare the budgetary years 2002 and 2001, you will find a noticeable fall in structural operations expenditure allocated to Greece. This is partly due to the fact that payments under outstanding appropriations from the previous programming period (1994 - 1999) are finally beginning to decline.

Expenditure under the common agricultural policy and the structural operations headings can virtually be completely allocated to Member States, i.e. you can determine practically 100 per cent where and how the money under headings 1 and 2 are spent.

But here again: Structural expenditure accruing to one Member State has wide spill-over effects reflecting largely the enhanced interdependence characterising the EU. Transfers appearing in the EU budget under this heading offer a limited view of the overall gains generated across the Member States.

There is also a "time" aspect to transfers made under the structural expenditure heading in a given year, i.e. EU spending programmes not only have short term effects but also a lot of effects in the long run that cannot be easily measured in money terms.

If you take structural assistance to Ireland for example, a country that has profited immensely from Community funds under heading 2, you can see that until 2003 transfers from the Cohesion Fund have contributed over 2 billion euro to the improvement of transport and environment infrastructure in Ireland. So the input is quantifiable on a short term basis.

However, besides the short term effects spending programmes have, like rising employment and income through, for example, the financing of big infrastructure projects, their long term impact is even more important: Because, when the initial round of spending is finished, the consequences of the improved stock of human capital and physical infrastructure will continue to show up in the form of higher productivity and enhanced profitability. - And these are effects of Community assistance which do not show in any net payer - net contributor graph.

2.2.3. Internal policies

Under heading 3 of the EU budget money is spent on a variety of internal policy activities such as training, youth, culture, audio-visual media and information; energy and environment; consumer protection, internal market, industry, and trans-European networks; research and technological development.

Unlike the common agricultural policy and structural operations, only around 90 per cent of expenditure under heading 3 can be allocated to Member States. This is mainly due to factors like payments that benefit non-member countries as well as cross-border measures that are difficult to allocate to individual Member States.

In 2002, those funds that could be allocated were distributed between the Member States as shown in the following chart:

Compared to 2001, internal policies expenditure increased by 24.7 per cent. This increase was mainly due to payments under the European Union Solidarity fund, which was set up in 2002. 444 million euro out of this Solidarity fund were allocated to Germany to alleviate the effects of the floodings of that year, which also explains why this country scores so well on the chart above.

The disparity between budgetary costs of spending programmes related to internal policies of only about 8% of total EU operating expenditure, and their impact in terms of growth and employment once more highlights the fact that measuring benefits from EU membership in budgetary terms alone does not always work.

2.2.4. External action

Category 4 of the EU budget is dedicated to supporting the objectives of EU external policy by means of cooperation, development aid, conflict prevention and human rights programme and projects. These objectives include, alongside development cooperation, promotion of the EU's identity on the international stage, notably through implementation of the common foreign and security policy.

Budgetary appropriations as a means to achieve the above goals are broken down according to geographical areas ("Relations with Asia", "Relations with Latin America", etc.). They also cover urgent humanitarian actions and support to non-governmental organisations (NGOs).

All in all, the Budget 2004 foresees an amount of 5.177 bn euro to be spent under heading 4, "External action", which is about 5% of the total budget of that year for EU 15. This sum includes 95m euro from the so-called "flexibility instrument", an additional fund to be used for unexpected budgetary requirements, following a strict procedure for its employment. The flexibility instrument was used in the budget 2004 in order to allow the Union to spend 160m euro on reconstruction in Iraq without having to severely cut back financing its other priorities in the field of external policy.

In addition to the appropriations under category 4, there is also the European Development Fund (EDF) which provides assistance to 77 ACP countries and amounts to about 13.5 billion euro for the period 2000 - 2007.

By definition, expenditure under heading 4 cannot be allocated to the Member States - so, sorry, no table this time!

2.2.5. Administrative expenditure

Expenditure under heading 5 of the EU budget can also not be allocated to the Member States because it is used to cover the administrative costs of the EU institutions.

In 2004, heading 5 of the budget comes to a total of 6.157 bn euro which is about 5.3% of the overall appropriations for EU 25 for that year. This amount will have to cover the total operational expenditure of all the European institutions in 2004, i.e. salaries, buildings and equipment.

The total staff of all the European Union institutions is about 32.000 officials which just about equals the number of public service officials in a city such as Vienna or the staff of the French Ministry for Agriculture. Over the last few years, administrative expenditure of the European Union has always been around 5% of the total EU budget.

This is quite remarkable given the additional administrative needs of the EU institutions in connection with EU enlargement (staff - new languages!, office space, etc.), which had to be catered for without having to increase the ceiling for heading 5 of the financial perspective. A series of exceptional measures (advanced payments, re-allocation of payments, etc.) have allowed the budgetary authority to finance the EU institutions' enlargement needs within the existing ceiling, which was something of a challenge but quite successfully mastered in the end.

Last but not least: Please bear in mind that 25 to 30 percent of all EU staff are engaged in linguistic work. So, over one quarter of the 32.000 EU officials work as translators and interpreters or in other jobs connected to EU multi-lingualism, safeguarding equal treatment of all official EU languages.

2.2.6. Pre-accession aid/Cost of enlargement

Before their accession in 2004, the ten new Member States received pre-accession Community assistance under programmes such as SAPARD, PHARE and ISPA to help them get "fit" for EU membership. By their nature these appropriations cannot be allocated to the "old" EU-15 and are therefore not part of our allocation exercise.

However, the cost of EU enlargement has been at the centre of speculation for a long time. So let's give you some figures on the financial consequences of enlarging the European Union with Poland, Hungary, the Czech Republic, Slovakia, Slovenia, Estonia, Lithuania, Latvia, Malta and Cyprus.

To meet the enlargement related expenditure needs, the financial perspective has been adjusted in headings 1, 2, 3 and 5 and by creating a new heading 8 - Compensation. These adjustments added a total of 40.953 billion euro to the current financial perspective up to the end of 2006.

The new heading 8 has been created for the new Member States, to keep them, at least in 2003 figures, from becoming net payers too soon, since they will start to contribute to the EU budget as from the first day of their accession.

In order to cater for the needs of EU enlargement the financial perspective has been increased as shown in the following table:

a) adjustment of existing headings of the financial perspective (figures in billion €)

b) creating a new heading in the financial perspective - "Compensation"

If you break down this amount to the level of the European citizens, enlargement will cost every citizen of the current Member States 36 euro per year. In reality however, costs per citizen will be lower since the new Member States will also fully contribute to the EU budget as from their first day of accession.

5. As from the year 2002, the concept of gross national product (GNP) has been replaced by the concept of gross national income (GNI) in the EU budgetary and own resources area. Thus, percentages for years before 2002 relate to the GNP, later ones to the GNI. For example: The EU expenditure ceiling of 1.27% of GNP corresponds to 1.24% of GNI using the new statistical approach.
6. Fontainebleau European Council, Conclusions of the Presidency: "Expenditure policy is ultimately the essential means of resolving the question of budgetary imbalances. However, it has been decided that any Member State sustaining a budgetary burden which is excessive in relation to its relative prosperity may benefit from a correction at the appropriate time".
7. In the sense of "proportionality of gross contributions to income across Member States"
8. The balance does not add up to zero because of exchange rate differences
9. All tables used in this chapter, unless indicated otherwise, reflect the actual payments made during the financial year 2002 = utilised appropriations for payments. Basic expenditure data comes from the Commission's accounting system. Operating expenditure excludes all administrative expenditure of all institutions. The exercise of allocating expenditure to Member States includes all operating expenditure, except for expenditure benefiting recipients outside the Union (expenditure under heading 4) and expenditure that is too difficult or too impractical to allocate. In the exercise, payments are allocated to the Member State in which the principal recipient resides; subsidies for Community agencies are allocated to the Member State in which they are located.
10.   see footnote 1 (page 6)
11.   African, Caribbean and Pacific States
12.   In addition to this there are also 9.9 billion euro available which have not been spent in earlier periods.


As you have already seen from the examples given so far, conventionally measured budgetary balances fail to adequately represent the benefits of EU membership. All in all we can summarise three aspects circumscribing why simple in-put / out-put calculations do not work:

3.1. Allocating expenditure to Member States

In consequence of what was concluded above, allocating expenditure to Member States as a mere accounting exercise gives only a very limited view of the benefits that each Member State derives from the Union.

In order to give a correct picture - within the limits explained above - expenditure on External Action, Administration, Pre-accession Aid and Reserves (headings 4, 5, 6 and 7) of the Financial Perspective needs to be excluded from this allocation exercise since it either does not directly benefit the current Member States or is virtually impossible to attribute with accuracy to any of the current Member States in particular.

Bearing all this in mind let us have a look at how the operational expenditure was allocated in 2002 (in € million and %):











As was to be expected, the five large Member States get the biggest share followed by the Cohesion Fund countries (Greece, Ireland, Portugal).

Tables and figures used in the net balance debate can often be misleading and this is nicely illustrated by what happens if you include spending under heading 5 in your expenditure table. We said before that administrative spending was excluded from the allocation exercise. However, often tables are used which do include administrative spending meaning that Belgium, Luxembourg and France as the three main seats of EU institutions get all the administrative expenditure allocated amongst them.

In doing so, the figures from the above table (which only included operating expenditure) change dramatically for these three countries, as you can see from the following table which includes heading 5 expenditure (total EU expenditure allocated in 2002 in € million and %):












This, in turn, strongly affects net balances or the figures of EU contributions per citizen of different Member States, which the media also like to use.

In the case of Luxembourg 87% of all expenditure allocated to that country falls under administrative expenditure (heading 5). Of course, the majority of people benefiting from this amount of administrative expenditure are not even citizens of Luxembourg. So, the picture is clearly distorted. Actually, if you exclude administrative expenditure again, the citizens of Luxembourg become the strongest sponsors of the EU with contributions of 197 euro per year each.

On the other hand, there are big indirect benefits for Luxembourg and Belgium arising from hosting EU institutions, which are also not reflected in the budgetary balances...

One would hope the point has been made how net balances may give a distorted picture. So let's leave administrative expenditure out of it for the time being.

To give you a final overview over the full allocation of 2002 operating expenditure you can find a breakdown by sector and Member State in the following table:

3.2. Concepts of budgetary balances

Would anybody in a Member State ever question budgetary transfers within his/her home country, let's say from a Southern province/Land/district to a Northern one, or the other way round? This is done very rarely, but of course with the European Union things are different...

And then there is also the "Rotterdam effect", which has not been explicitly mentioned yet but which also contributes to distorting all budgetary balances if you include Traditional Own Resources in the Member States' contributions towards the EU budget. Countries with major ports like Rotterdam (or London), where customs duties are paid in the incoming port although goods are destined for other countries, naturally have a large amount of TOR to transfer to "Brussels". Their official payments into the EU budget are enormous but of course it was never the Netherland's (or the UK's) money anyway and does not come out of the pockets of their citizens.

After all the warnings already given to readers about the usefulness of comparing budgetary balances let us finally give you some tables on net payers and net receivers. We hope you have been cautioned enough to see the limits of this exercise. Even so, there are still some interesting details coming to light.

3.3. Net contributors - net payers

Finally, here comes the well known chart on net contributors and recipients - but just to make it clear once more: These figures - on the revenue side - are based on the full own resources payments of the Member States, including the traditional own resources (and the Rotterdam effect); and - on the expenditure side - they include administrative expenditure:

(figures for Member States, 2002, in million euro)

It's what you expected all along, isn't it?

But it is still misleading, because, if, as stated above in the chapter on own resources, traditional own resources are to be considered the only REAL own resources of the European Union, then TOR should be left out of the picture if you look at the total of Member States' contributions. If you deduct the TOR from these contributions (because they never were the Member States' money in the first place) you arrive at a very interesting picture of REAL net balances, and of who is a net payer and a net receiver within the European Union:

(figures for 2002, in euro million):

If you do this for 2001, for example, it gets even more interesting because the UK suddenly turns into a net receiver:














And this was not a single occurrence: The same was true in the year 1997, when the United Kingdom also had a real net balance (after deduction of the TOR) of +1.548 billion euro, meaning it got over one and a half billion euro more from "Brussels" than it had transferred there.

And finally, you get another astonishing picture if you break down net balances to the level of the individual citizen. Then the people from the Netherlands and Luxembourg turn into the biggest contributors, whereas it is the Irish citizens who get the most out of it.

Have a look at our final table about net contributors and recipients in 2002 (in euro per citizen):











But don't forget: This table does not show you the whole picture of the EU budget but just one limited part of it.

Part 2:


Every November, the European Court of Auditors publishes its annual report on EU financial management of the year before. Every year this report leads to the same critical headlines:

[EU Budget 1994] Auditors blame EU for £2bn fraud and waste
"Lax management and fraud cost European taxpayers up to £2bn last year, the European Union's auditors estimated yesterday. Their report attacked the European Commission and member states for failing to curb their spendthrift ways. ... Exercising new powers conferred on them by the Maastricht treaty, the auditors refused to certify the overall 'reliability' of the Commission's accounts because of faulty and inadequate figures. ... The auditors estimated that 4% of the budget, or £2.02 billion, was subject to 'substantial errors'." The Times, 15 November 1995, on the EU Budget 1994

[EU Budget 1995] Auditors criticise EU fraud
"European Union member countries will be criticised today for failing to get to grips with fraudulent claims on the EU budget." Financial Times, 12 November 1996, on the EU Budget 1995

[Budget 1996] EU members lose almost £3bn to fraud and waste
"The annual report by the Court of Auditors, the EU's financial watchdog, says about 5.4% of the union's total £52 bn budget in 1996 was marred by 'substantive errors'. There are so many discrepancies in the accounts that the court will also announce, for the third year running that it cannot certify their legality. The extent of the waste uncovered ... looks set to add to demands for a reform of many of the EU's spending programmes, which have been described as an invitation to waste and fraud." The Sunday Times, 16 November 1997, on the EU Budget 1996

[Budget 1997] Auditors Find Evidence Of 'Irregularities' in EU - Report Stops Short of Alleging Fraud in Budget
"The European Union's Court of Auditors reported widescale 'irregularities' on Tuesday that cost taxpayers some $5.9 billion .... The court estimated that about 5 percent of the budget was mismanaged or improperly accounted for ..." International Herald Tribune, 18 November 1998, on the EU Budget 1997

EU watchdog barks at £3bn fraud
"The European Union's financial auditors will refuse to approve its accounts next week because of irregularities and missing money amounting to more than £3bn - 5% of the EU's £63bn budget. The annual report of the Court of Auditors ... blames fraudulent grant claims and inadequate accounting in member states and the EU for the losses." The Guardian, 14 November 1998, on the EU budget 1997

[Budget 1998] EU chaos over GBP 3bn fraud gag
"Jan Karlsson, head of the EU's court of auditors, explained how, for the fifth year running, he was unable to sign off the EU accounts for 1998 because GBP 3 billion of taxpayers' money had been mis-spent. He revealed a new catalogue of financial abuse and criticised the EU's executive Commission for serious mismanagement of funds." The Express, 17 November 1999, on the EU Budget 1998

[Budget 1999] Mismanagement and fraud cost EU taxpayers £4 bn
"The European Commission wasted up to £4 billion of taxpayers' money during 1999, according to a damning annual report by independent auditors. ... The auditors refused to state the exact sum squandered but one of the report's authors said that the figure was 'between 5.5 and seven per cent' of the commission's £60 billion budget - an even higher level of waste than before." The Sunday Telegraph, 19 November 2000, on the EU Budget 1999

[Budget 2000] Clampdown fails as EU loses £2.5 bn to fraud and mismanagement "The European Union lost £2.5 billion - equivalent to £15 for every family in Europe - through mismanagement and fraud last year, according to an auditors' report. The losses, 5% of the EU budget, were found by the court of auditors, its financial watchdog." The Sunday Times, 11 November 2001, on the EU Budget 2000

[Budget 2001] Financial watchdog discloses EU fraud and error
"The European Union's top financial watchdog refused to certify the EU's £63 billion budget yesterday for the eighth year running, admitting it can only guarantee that five per cent of taxpayers' money is being spent properly." The Sunday Telegraph, 6 November 2002, on the EU Budget 2001

EU auditors uncover fraud and complacency
Only "5 per cent of the EU's expenditure - mainly relating to internal administration - was legal and regular. The remaining 95 per cent, including the sprawling farm and regional aid budgets, was not given a positive statement of assurance 'due to the incidence of errors found'". Financial Times, 6 November 2002, on the EU Budget 2001

[Budget 2002] Nine years of fraud
"Fresh evidence that fraud and corruption remain rampant in Brussels emerged yesterday. The European Union's own spending watchdog has refused to give its annual accounts the all-clear for the ninth year running. ... In yet another damning assessment of the EU's handling of its £65 billion budget, the Court of Auditors questioned the accuracy of the last year's spending figures." Daily Mail, 19 November 2003, on the EU Budget 2002

Something is still rotten in the state of Europe
"For the ninth successive year, the European Court of Auditors has refused to sign off on the EU's £65 billion budget, saying the accounts are still plagued by 'significant errors' and 'irregularities'. … Concerns about rampant corruption involving EU money … have been around Brussels and Strasbourg for years. Yet nothing is done about it and no one is required to take political responsibility." The Daily Telegraph, 19 November 2003, on the EU Budget 2002

As these examples show, every year the same figures are given, the same amounts quoted and the same allegations made. Thus, the figures of 5% of the EU budget or 3/4/5 billion euro (increase following the growing EU budget) that are "wasted by Brussels through mismanagement or fraud" have been soundly imprinted on the minds of the European public. They come up in every EU round table talk on television or radio, they are constantly quoted in newspaper and magazine articles, they have become "public knowledge".

Certainly, it would be unrealistic to assume that no irregularities occur with regard to the EU budget. Like in every other administration, mistakes are bound to happen, even the odd case of fraud. And indeed, the occurrence of financial irregularities is far from being a uniquely European Union or European Commission problem - it is world-wide, and so is the problem of assurance.

At a conference organised by the Commission's Internal Audit Service9, Phil Calder, Chief Accountant of the United States General Accounting Office (GAO), said that the situation in Europe regarding the DAS reminded him very much of the situation in the US. In 2000, 19 out of the 24 most important Federal Executive Agencies had only received a qualified audit opinion. Similar to the state of affairs in Europe, the GAO had also not been able to express a clean opinion on the consolidated accounts of the US Federal Government. So, Europe is not alone10.

But are several billion euro really being wasted in Brussels through irregularities or fraud? And worse: has the situation not improved in all those years?

In order to get a clearer perspective let us look at the basis of all these figures given in the articles above. It is a long story, which started in 1994.

1: The State of Assurance

1. The Statement of Assurance

Every year, in the month of November, the European Court of Auditors publishes a report on the results of its audit of the European Community accounts of the year before. Since the Treaty of Maastricht these Annual Reports also need to contain a Statement of Assurance (DAS)11, certifying the reliability of the accounts and the legality and regularity of the underlying transactions. This DAS was expected to render the discharge procedure more meaningful and the Commission more responsible in implementing the budget. With the Maastricht Treaty coming into effect on 1 November 1993, it was the 1994 financial year that saw the DAS fully implemented for the first time.

In order to be able to give the DAS as required by the Treaty, two types of audit are carried out by the Court of Auditors: an audit as to the reliability of the accounts and an audit of the legality and regularity of the underlying transactions.

1.1. Audit of the reliability of the accounts

The objective of the audit of the accounts is to gain sufficient evidence that all transactions, assets and liabilities have been completely, correctly and accurately recorded in the accounting records and presented in the financial statements. The audit opinion will indicate whether the consolidated accounts accurately reflect the revenue and expenditure and give a true and fair view of the European Communities' financial situation as at 31 December.

The reliability of the accounts is not dependent on the transactions being legal and regular, but whether they are completely and accurately recorded and disclosed.

The audit of the accounts has always led to a result that was - though including some reservations/qualifications - overall POSITIVE, i.e. a positive Statement of Assurance.

1.2. Audit of the legality and regularity of the underlying transactions

The objective of the audit of the legality and regularity of the underlying transactions is to gain sufficient evidence that funds have been received and spent in conformity with contractual and legislative conditions and are correctly and accurately calculated. Questions to be answered by this audit are for example: Does the underlying transaction exist? Is the recipient/beneficiary eligible? Do the costs/quantities claimed exist? Are they accurate and eligible?

The opinion resulting from this audit consists of an overall conclusion on revenue and expenditure transactions. This audit not only includes expenditure at the European Union level itself but also payments on a national (Member States, 3rd countries), regional, local or even individual level for which authorities at these levels are responsible.

It was this audit of the legality and regularity that mainly caused the above newspaper headlines. Its results have led the Court of Auditors to refrain from giving a positive Statement of Assurance on payments made from the EU budget ever since the DAS was introduced in 1994.

9.  "Verstehen" Conference 2001
10. However, there was a consensus amongst participants of this conference on one big difference between the situation in the EU and the US: In the US the auditing focus was principally on financial information and not on the regularity of payments. So, other than in Europe, the political reaction in the US to the failure to obtain a clean audit opinion is only "a big yawn", as Mr. Calder put it.
11. Art. 248 of the Treaty stipulates: "The Court of Auditors shall provide the European Parliament and the Council with a statement of assurance as to the reliability of the accounts and the legality and regularity of the underlying transactions which shall be published in the Official Journal of the European Union."

2: The Court's Findings

First it is interesting to know that only for the first three years of the DAS, in its Annual Reports concerning the financial years 1994, 1995 and 1996, did the Court of Auditors publish global error rates, i.e. actual figures on the amounts likely to be concerned by irregularities. As from 1997, the Court ceased to do so. But the figures dating back to those three years still have their repercussions today.

With regard to the legality and regularity of the underlying transactions, since 1994 the Court's audit has revealed:

  • An acceptable situation, as a whole and within the scope of the audit.

    • For own resources and commitments (which are the responsibility of the Commission);

    • For staff expenditure payments (since 1999) and

    • For all administrative expenditure payments (since 2000) but

  • An unacceptable incidence of error in operational expenditure payments (85% of which take place in the Member States) for all other areas.

2.1. Financial year 1994

After having carried out its audits, on 14 November 1995 the Court of Auditors presented its Annual Report concerning the financial year 1994 that contained the first DAS.

While the Court declared in its report that "the accounts ... accurately reflect the revenue and expenditure, as well as the financial situation of the Union" and "the Court's examination of the revenue entered in the accounts in respect of traditional own resources did not reveal any significant errors" it arrived at other results regarding expenditure: "There are too many errors in the transactions underlying the payments entered in the accounts for the Court to be able to give a positive global assurance as to their legality/regularity".

As to the financial implication of these errors the Court stated that "The most probable cumulative amount of the very numerous substantial errors concerning payments, which resulted in the sums concerned being wrongly charged to the budget, is of the order of ECU 2 386 million".

For the first time, this statement made by the Court in its report on the financial year 1994 gave a figure and a percentage: It said that ECU 2.386.000.000 or 4.0% of the EU Budget had not been spent correctly. Stories of fraud and waste went through all the European media and the public outcry was enormous. This being two weeks before the Norwegian referendum on EU membership , little wonder they voted "No"12.

2.2. Financial year 1995

The situation was not much different in the following year. The Court of Auditors' DAS for 1995 concludes with the statement that "The most probable cumulative amount for the very numerous substantial errors concerning transactions underlying the payments which had measurable implications for the amounts financed from the general budget was around 4 000 million ECU (5.9% of the amounts involved)." As in the year before, the Court of Auditors considered this error rate being far too high and refused to give a positive global assurance as to the legality and regularity of transactions.

2.3. Financial year 1996

An interesting thing happened in 1996: Though still giving a global error rate, in its DAS for that year the Court of Auditors explicitly cautioned the readers of the report against taking the figures at their face value:

The "audit has to cover all stages in the management of Community funds, starting with the Commission's central accountancy and ending with final recipients in the Member States and elsewhere. The Court's audit for the purposes of the Statement of Assurance is aimed at establishing whether the rate of error detected in its sample of transactions verified down to the level of final recipients is low enough to enable it to provide a positive global assurance that the accounts are reliable and the underlying transactions legal and regular. On the basis of extrapolations of its sample results, it provides estimates of the rates of substantive error detected in order to support its conclusions. These estimates should be regarded as indications of orders of magnitude, associated with upper and lower limits at a level of 95% confidence, rather than as precise measurements."

However, after these remarks, the Court of Auditors did state that "the overall rate of substantive errors - quantifiable errors directly affecting the amount of the transactions underlying Community funds disbursed - is estimated to be in the range of 3.5% to 8.3% of total payments: the most likely estimate is 5.4% of total payments."

And it again concluded that "the estimated level of error in the transactions underlying the payments for the financial year 1996 is so high that the Court declines to provide positive global assurance as to the legality and regularity of the transactions concerned."

These statements of the Court of Auditors for the first three years of the DAS sound bad. The amounts involved are huge, "Brussels" had a bad reputation anyway and there is very little that annoys citizens all over the world more than having to watch how public funds are administered incorrectly.

However, the Court's warning against taking the figures included in the DAS literally and "for real" gives a first indication that at a closer look these "clear" figures might not be what they seemed to be.

12. In the referendum of 28 November 1994 some 52% of Norwegians rejected EU membership of their country.


As pointed out by the Court of Auditors itself, its findings must not be interpreted as a mathematically exact statement of the scale of the errors identified, as it was done by most commentators. The DAS figures are the result of applying a statistical method, they are an extrapolation made on the basis of a sample. As with all statistics you need to be very careful in judging on what the figures provided actually prove or do not prove.

It is a "speciality" of the EU budget that financial transactions do not only take place at one level. Instead, many of the transactions within the EU budget have an important "multiple layer" or "cascade aspect". I.e. large payments are made by the Commission to the Member States, more payments by the Member States to their agencies, to different regional authorities, etc., even more payments by these intermediate recipients to the final recipients, often via additional other stages in between.

Therefore, the organisation of the audit saw a methodological dilemma because of the huge number of systems and subsystems (many thousands) and transactions (many millions) to be covered. The transactions can only receive a very thin random coverage. Their auditing on the spot in Member States/at the final beneficiaries' premises was often insufficient due to the limited resources available.

With millions of EU budgetary transactions at several levels going on within a financial year, the Court of Auditors has simply not been able to check every single one of them. Instead, it has had to use statistical methods based on sampling to calculate the error rates so relevant for the DAS - at least in its early stages.

Therefore, when trying to interpret the DAS correctly, there are several important factors to be taken into consideration:

  • the limitations of statistics
    Do statistical results provide a true mirror of the real world?

  • the nature of the irregularities involved
    Are all irregularities observed intentional attempts at omission, deception or even fraud?

  • the question of who is responsible
    Is it really Brussels bureaucracy that wastes billions of taxpayers' money every year?

At this point: Reader beware! As one senior EU budgetary control official so nicely put it, "The trouble is the thing is so damn complicated, involves the use of numerous technical terms and is inherently hard to explain. I'm personally aware that I needed a full day's seminar in the Court of Auditors to get to grips, without really understanding, the statistical methods behind the DAS." Bearing this in mind, let us meet the challenge of statistics.

3.1. Statistics

The decision by which criteria and method the Court of Auditors would arrive at the DAS was left to its discretion and not prescribed by the legislator.

In order to be able to assess the legality and regularity of millions of individual transactions at EU, national, regional and local level down to the final recipient, the Court chose a sampling method called MUS = Monetary Unit Sampling. According to MUS, the higher the financial impact of a transaction is on the budget the more likely this transaction is to become part of the sample (PPS approach - "probability proportional to size"). Statistics, on the basis of the MUS, would then allow the Court to conclude with a reasonable degree of certainty that the gross error in the transactions does not exceed a reasonable level.

By applying the MUS method, samples were drawn from the SINCOM system, the Commission's book keeping system, which records each and every transaction with an impact on the EU budget.

Regarding the outcome of the audit, the Court decided to grant a positive Statement of Assurance (DAS) if testing led to 95% certainty that the errors in the transactions did not exceed a materiality threshold of 1% (i.e. errors concern 1% of the total amount involved).

To begin with, this all sounds very reasonable and well founded. However, here is exactly where the problems start. Critics said from the very beginning that the method chosen by the Court made a positive Statement of Assurance practically impossible. The author as the rapporteur in the European Parliament's Budget Control Committee for the first DAS, after analysing the DAS and questioning the Court of Auditors and the Commission, publicly stated that we would never see a positive DAS with the way the systems operated.

A closer look will show that the Court's approach did indeed favour negative statements, due to the statistics involved and the way the EU budget was structured.

3.1.1. Parameters

There were three interdependent parameters involved in the sampling process:

Confidence level


Materiality threshold


Sample size


The 95% confidence level/certainty is the standard value used in statistics for this kind of calculation.

The 1% materiality threshold was set by the Court of Auditors within the scope suggested by the INTOSAI13 Auditing Standards which consider any value between 0.5% and 2% an "appropriate materiality threshold".

The critical parameter in the approach used by the Court of Auditors was the very small sample size of 300 compared to the many million budgetary transactions that made up the whole "population". The small sample was of course due to practical restraints in connection with the human resources available. The Court simply did not have enough manpower to pull more samples and check them.

But according to the Court itself, the sample needed to consist of at least 6000 transactions14 to have a chance - even in the presence of error - to reduce the confidence interval to 1% or under, i.e. under the materiality threshold. The small sample size of 300 made it practically impossible to get below the materiality threshold of 1% (i.e. to get a positive Statement of Assurance) if any error at all is discovered in the sample.

3.1.2. Method

Statistical sampling is used for drawing conclusions about a "population" that is too large to be checked in its entirety. If you pull a sample from a population you have to extrapolate your sampling results in order to be able to arrive at overall conclusions. Because any sample, be it as representative as it may, is after all still a "sample", these extrapolating calculations do not result in one value (i.e. one single figure) but rather in a range, or an "interval", in which the real value is likely to lie (i.e. between figure a and figure b). In the case of the DAS, when you are looking for errors, this interval reaches from a "lower error limit" to an "upper error limit" with "the most likely error" somewhere in between.

But it is even more complicated because there is another step involved. The error found in your sample (in case of the DAS a sample of 300) has of course an element of chance in it which is all the bigger, the smaller the sample: Chance might have it that if you had pulled another sample of 300 different transactions the error found might have been totally different. So, your "chances" that you actually arrived at a valid result with just one of these "likely error intervals" are actually rather low. Accordingly, any interpretation of where the lower and upper limit of your interval lie has to be done very cautiously.

A confidence level of 95%, as chosen by the Court, is based on the following interpretation: If statisticians would calculate 100 intervals from 100 different samples pulled from the same population, then 95 of these intervals would contain the most likely error.

Given this background, when interpreting your one existing interval the following assumption has to be taken into consideration: If you actually had calculated 100 intervals then these, of course, would have different lower limits and different upper limits. Consequently, if they are all joined up (to get at a 95% certainty), the result is one larger interval because it extends from the lowest lower limit to the highest upper limit. Your existing interval would only cover a small part of it.

Therefore, the intervals found and published by the Court do not have exact lower and upper error limits, i.e. the intervals tell you even less than expected because their borders are anything but sharply defined.

You had your warning at the beginning of this chapter - so, are you still with us at this point?

Let us go back now to our 1% materiality threshold. Since this parameter is a constant, fixed by the Court of Auditors according to international standards15, any likely errors range larger than 1% is bound to go above the threshold. With a certainty level of 95% and a small sample size of 300, it is practically impossible to arrive at a confidence interval of less than 1%. The sample would more or less have to be without the slightest error in order to stay below the materiality threshold, which is of course near to impossible.

In order to reduce the confidence interval to 1% or under (i.e. to allow for it to go below the materiality threshold), you would need to increase the sample size16. Put simply: Increasing the size of the sample would provide you with more information/data. This would make the confidence interval smaller, while keeping up the certainty level of 95% (the third parameter of the calculation). And we already mentioned above that the Court itself thought a sample of at least 6000 was needed to have a chance of getting below the materiality threshold ...

3.1.3. Results

On the basis of what has just been said, let us have a look at the figures published for the financial years of 1994, 1995 and 1996, those three years when actual error rates were published by the Court:

The underlying presumption is clear: If the upper error limit lay below the materiality threshold of 1% the underlying transactions could clearly be accepted. If the lower error limit lay above the materiality threshold of 1%, the legality and regularity of the underlying transactions would clearly have to be refused17.

The figures in the table above mean that in 1994, 1995 and 1996 the audit of a sample of 300 transactions taken from the whole budget caused the Court to conclude that, with 95% probability, the lower limit of the probable error rates was above materiality.

A global error rate of even 2.9 (lowest likely error rate calculated/1994) was considered unacceptable by the Court, the assurance for the years in question was refused.

3.2. Definition of error

We saw above that the statistical method applied was an important factor in making a negative Statement of Assurance very likely. However, some of the errors found in the sample and contributing to the negative DAS also deserve a closer look.

Regarding legality and regularity of the underlying actions, testing for error was done from the Commission decision down to the final recipient of the funds mainly on

  • the reality and eligibility of the activity concerned

  • the accuracy and eligibility of the costs declared and

  • the accuracy of the quantities claimed.

In doing so, the Court of Auditors differentiated between substantial error and formal error18. In principle, substantial error is an irregularity with a measurable and quantifiable effect on the EU budget. Formal error is an irregularity (usually a failure to comply with the applicable regulations) with a possible effect on the EU budget that is, however, not quantifiable19. Because its effects could not be measured, formal error was not included in the extrapolation for the global error rate but was nevertheless included in the Court of Auditors' reports that listed number and nature of real cases of formal error found in the sample.

According to its official definition20, "'irregularity' shall mean any infringement of a provision of Community law resulting from an act or omission by an economic operator, which has, or would have, the effect of prejudicing the general budget of the Communities or budgets managed by them, either by reducing or losing revenue accruing from own resources collected directly on behalf of the Communities, or by an unjustified item of expenditure."

This definition does not differentiate whether the irregularity was caused by an intentional act or omission (fraud) or unintentionally, by a mistake for whatever reason possible. Therefore, while "fraud" may be the reason for some irregularities, it is certainly not the reason for all of them, as can be seen from its rather stringent definition:

In respect of expenditure, fraud affecting the Communities' financial interests shall consist of "any intentional act or omission relating to

  • the use or presentation of false, incorrect or incomplete statements or documents, which has as its effect the misappropriation or wrongful retention of funds from the general budget of the European Communities or budgets managed by, or on behalf of the European Communities,

  • non-disclosure of information in violation of a specific obligation, with the same effect,

  • the misapplication of such funds for purposes other then those for which they were originally granted."21

There were certainly some cases of fraud discovered by the Court of Auditors in the samples they pulled for the DAS. However, some other cases that also contained a "substantial error" and thus became part of the calculation for "most likely error", were not as clear cut, as the following examples22 will show:

Example 1: United Kingdom

This case concerns a farmer who in 1996 applied for the extensification premium for beef for 21 Livestock Units. The extensification premium is granted provided that the stocking density is less than 1.4 livestock Units per ha of forage area. Hence, to fulfil the condition the farmer should have at least 15.00 ha of forage area. The farmer had declared having 15.03 ha.

The Court of Auditors made an on-the-spot check to verify the declared area. The first measurement carried out by the Court using a so-called measuring wheel showed an eligible area less than that needed for the farmer to be eligible for the extensification premium. However, as the differences were small, the Court asked the regional inspector to accurately measure the whole farm. According to this result the forage area was 14.65 ha and some doubts were still left whether the area of a stream had been deducted from the eligible area. Similar doubts remained concerning some hedgerows.

The Commission will not deny that there can be doubts whether there were 15.00 ha of forage area. But the fact that the Court had to ask for a second more precise measurement shows that we here are dealing with a case where it is very difficult to judge with certainty who is right. The British Authorities have replied to the case that "... discrepancies are minimal and within acceptable tolerances ..."

The Commission is of the opinion that the problems in this case are due to the fact that the UK authorities do not apply any technical tolerances at all, whereas all other Member States do apply such tolerances. Therefore, such insignificant differences would not have, with the greatest likelihood, resulted in a refusal in other Member States.

Example 2: Spain

This is a road construction project involving adding a second carriageway to an existing road. The Court claims that such a project requires the prior carrying-out of an environmental impact assessment.

Under the relevant Directive in application at the time, such an assessment is required for an "autoroute" or a "voie rapide", but not for other categories of road. We [the Commission] claim that since slow vehicles were not excluded from the improved road, it was not a "voie rapide".

Moreover, "voie rapide" is defined (by reference to the European Agreement of 15.11.75) on Main International Traffic Arterials) as a road accessible only via regulated interchanges or junctions; whereas the road in question still has some traditional direct-access junctions.

Because of uncertainties as to what categories of road project should have to have a prior environmental impact study, the Directive was later amended to clarify the situation.

The Court claims that in two other cases (roads in Murcia and Valencia), the Commission has taken the opposite line and stated that an assessment is required. But
a) the Commission has not taken a final decision on these,
b) it is not clear that this road has as high a design standard as these other two.

This seems a very insecure case for the Court to choose as a major "error".

These examples show that some of the irregularities found in the sample and qualified as substantial error because they did have a measurable effect on the EU budget were certainly contestable. Irregularities like those above affected the result of the sample with direct consequences for the extrapolation and the resulting figures. With the statistical method used, which only allowed for an absolute minimum of error found for a positive DAS, taking aboard such irregularities had an even bigger effect.

So, two questions arise: Was it fraud? Was the Commission to blame?

3.3. Responsibility

The example of the UK farmer quoted above brings us to another factor that has to be taken into account when looking at the negative DAS. Who was it that actually claimed too much money (supposedly the measurement of 14.65 ha was correct) and who paid it?

It was not some eurocrat in Brussels but a final recipient of EU money who was paid by one Member State. If the Court of Auditors only had to audit payments made directly by the EU Commission or other EU institutions the error rate found would be considerably lower. But Article 248 of the Treaty stipulates that "The audit shall be … performed on the spot ... on the premises of any body which manages revenue or expenditure on behalf of the Community and in the Member States, including on the premises of any natural or legal person in receipt of payments from the budget."

With 80 - 85% of all EU funds spent by the Member States in the Member States it does not come as a surprise that 90% of the errors recorded by the Court of Auditors were made at the level of - again - the Member States.

3.3.1. Where the money is spent

The vast bulk of this EU expenditure concerns payments in the fields of agriculture and structural funds. The money is simply transferred by the Commission to national paying agencies which are required to observe rules established centrally for disbursing the funds.

Though the European Commission is legally responsible for the entire EU budget, it has no practical means of ensuring that these rules are respected by those spending the money. This is even the case with the new Financial Regulation because the Member States refused to grant the Commission such means ...

3.3.2. Where the irregularities occur

It was almost exclusively the financial management of the paying agencies in the Member States, which the Court of Auditors found fault with. Still, the problem for the European Commission is that it is legally responsible for the implementation of the EU-budget but, at the same time, a large number of discretionary decisions are left to the agencies of the Member States. Very often these agencies make mistakes with regard to the eligibility of projects for EU-funding.

While the Commission considers its role to be that of ensuring that Member States have in place a system of financial controls that meet standards set centrally, this role cannot include ensuring that all decisions taken in Member States which have financial consequences for the EU-budget are correct.

Although, judging from the irregularities described in the DAS, some Member States seem to have more problems in ensuring proper financial management of their paying agencies than others, repeated requests by the European Parliament to the European Court of Auditors to publish "black lists" of Member States that have a particularly bad record ("naming and shaming"), were persistently refused by the Court of Auditors. The Court pointed out that, following the DAS, drawing up a list of errors per Member State and using it to come to conclusions about the relative performance of the Member State would be misleading and professionally incorrect23.

13. International Organization of Supreme Audit Institutions
14. The effect of the sample size on the confidence interval can for example be seen in traditional opinion polls before elections, etc. Often, the sample size in these polls is 2000, which leads to a standard confidence interval of about 4%. Conclusion for example: "Party A lies at 39%, +/- 2%", i.e. a confidence interval of 4%. What the standard confidence interval of a sample of 300 would be, one can only guess ... but certainly more than 1%.
15. When setting the threshold at 1%, the Court of Auditors decided to apply the INTOSAI recommendations of 0.5-2.0% in a very strict way. A higher threshold of 1.5% or more might have had a noticeable effect on the outcome of the audit.
16. There is, of course, another possibility for reducing the confidence interval. It uses the other parameter of the calculation, the certainty level of 95%. If you were to decrease it, the likely errors range would also get smaller but the conclusion as a whole would become less exact.
17. Despite the refusal of the DAS, Council recommended discharge in all three years. The European Parliament refused discharge for 1996 but other political factors were involved. Granting discharge means in fact that Council and Parliament obviously consider even 8.5 (highest likely error rate calculated/1995) as acceptable. While the DAS is widely considered an essential tool for the discharge debate, opinions differ as to whether the Court of Auditors' decision not to certify the legality and regularity of the underlying transactions constitutes sufficient reason for refusing discharge. The precedent established in 1999 for the EU allows for closure of the accounts to be carried out independently of the granting of the discharge.
18. The exact classification of error for the purpose of the DAS is stipulated in the Court of Auditors' "Practical guidelines for the DAS". Unfortunately, this being an internal document, the Court of Auditors saw it not fit to provide a copy of the "Guidelines" to the European Parliament.
19. For example failure to comply with the proper call for tender procedure might have an effect on the EU budget (like making a service or good more expensive). However, this effect cannot be measured.
20. Council Regulation (EC, Euratom) No 2988/95 of 18 December 1995 (Article 1, paragraph 2)
21. Council Act of 26 July 1995 drawing up a Convention on the protection of the European Communities' financial interests
22. Taken from the DAS 1997, Commission's reply
23. The sample of transactions is not drawn using the Member State as a sampling criterion. Therefore the transactions selected for testing are not representative at the Member State level. The overall combined sample size necessary for providing useful comparable information linked to Member States would be counted in thousands and thus impossible for the Court given its present resources.


Since the first DAS in 1994, press and public have focused on errors and shortcomings only. Nobody paid much attention to the improvements that have been made as a consequence of Statements of Assurance that were not "clean" but qualified (regarding reliability of the accounts) or even negative (regarding legality and regularity of the transactions).

"On account of human nature", as Egbert Kaltenbach, Chief of the UNHCR Audit Service, once put it24, "it is impossible to achieve a 100% fraud-free environment". Therefore it simply is utopian to expect "perfect assurance" with millions of transactions taking place budget year after budget year.

Statements of Assurance that pinpoint problems, highlight weak points and monitor improvements are proof that the underlying control system, the "master control system" at EU level as provided for by the Treaty, works. It is understandable that qualified and negative statements give rise to public criticism year after year. However, if there was no Court of Auditors and no statements - then journalists and citizens really would have to worry.

Over the years, the Court of Auditors' qualified or negative Statements of Assurance and the European Parliament's positions in the discharge procedure have had a tremendous effect on changing the system. With these statements, the Court and Parliament largely contributed to initiating the reform of the European Commission and have closely monitored the progress made ever since. It's not only on the European level but also on the level of the Member States (but due to the increased pressure to do so exerted by the European Commission) that control systems have been developed, introduced and strengthened in order to better monitor financial transactions in those areas where most irregularities were bound to happen: agriculture and structural funds.

"The Statement of Assurance ought to provide us with data on changes in management, in other words, it should tell us whether the Commission and the Member States have made the necessary improvements or whether, on the contrary, the shortcomings persist and the errors are being repeated", confirmed Juan Manuel Fabra Vallés, President of the Court of Auditors, in his presentation on the court's work programme for 2003 to the EP Budgetary Control Committee.

4.1. Reform of the Commission

The regular qualified (although overall positive) Statement of Assurance on the reliability of the accounts was one of the main triggers for a profound reform of the European Commission's accounting system. The negative Statement of Assurance on the legality and regularity of the underlying transactions, together with some other related events, led to an overhaul of financial management at the European level, with an action plan set out in the White Paper on Reform and a deadline for its implementation.

The three most important Commission initiatives for future prevention or detection of irregularities and/or fraud in the management of Community funds are

  • the new Financial Regulation

  • the introduction of mandatory Internal Control Standards and

  • the New Staff Regulations.

A new and thoroughly modernised Financial Regulation which governs the entire expenditure of the Community budget came into effect on 1 January 2003. This Regulation and its Implementing Rules provide clear guidance on all major budget execution and internal control matters. They considerably strengthen the internal control system.

The Commission has introduced 24 Internal Control Standards, which are to ensure a comprehensive internal control system. All services had to comply with the baseline requirements of these standards by 31 December 2003. Standard 1 requires that "staff are fully aware of the rules governing staff conduct and prevention of fraud and irregularities". The standards also foresee a risk management system, clear procedures for reporting improprieties, the establishment of appropriate supervision arrangements including ex-post control of transactions, and the recording of internal control weaknesses.

The New Staff Regulations, due to enter into effect on 1 May 2004, contain new rules protecting officials who report concerns (whistleblowing). Since February 2002, these rules already apply at the Commission. They offer protection to whistleblowers and identify "secure" internal reporting channels and also make it possible in certain circumstances to report concerns to the President of the European Parliament or to the Ombudsman.

These new initiatives are considered to be an important step forwards for the Commission in its efforts to prevent and detect irregularities and fraud. Once the reforms are fully implemented, events like those currently unveiled at Eurostat will become highly unlikely, the Commission is convinced. And it will be the Court of Auditors again that will monitor implementation and effects of these reforms.

4.1.1. Accounting system

Since 1994, other than the audit of the underlying transactions, the audit of the accounts had resulted in a positive DAS although with qualifications, meaning it had been stated that the accounts were reliable "except for ...". Over the years numerous such reserves were included, some of which were repeated year by year like

  • overstatement and understatement of fixed assets

  • insufficient information on advances and payments on account

  • overstatement and understatement of commitments

  • overstatement of commitments still to be settled (RAL)

In the DAS 2001 the Court emphasised once more that most of its reservations and observations were matters that regularly occurred: "These weaknesses stem to a large extent from the Community accounting system which was not defined to provide an assurance that the various components of the Commission's assets have all been recorded." In conclusion, the Court considered that urgent in-depth action in the Commission's departments to cope with the risks arising from the shortcomings in the accounting system was required.

This recurrent criticism by the Court has been met with new developments like the new Financial Regulation, the new "Accounting framework for the European Communities" and the subsequent action plan concerning in particular the inter-operability of the management information systems.

4.1.2. Financial management

After the resignation of the Santer Commission in 1999 and the report of the Committee of Independent Experts ("wise men"), the Commission adopted its White Paper on Reform in April 2000. It was largely inspired by the experts' report, the discharge resolutions of the European Parliament and - the conclusions of the Court of Auditors, which all had recommended a profound change of the Commission's financial management and internal control system.

Consequently, in 2001, the DAS focused on the progress made in this respect by the European Commission, with the central sample being 450 transactions at the level of the Commission, and a pre-evaluation of the Commission reform, then already initiated, as an additional element of the audit. The new internal control system of the European Commission was evaluated to find out its strengths and weaknesses. The objective was to arrive at first conclusions on the capacity of the new system to ensure legality and regularity of the underlying transactions.

If you look at these developments, the Court of Auditor's pointing out of irregularities together with the European Parliament's persistence on reforming the EU Financial Regulation have certainly been instrumental in bringing about a new and more transparent system of financial management on the Commission level.

Besides: We should not forget that the original systems were designed for six Member States - and not for a European Union of our current size and competencies.

24. Statement at the " Verstehen " Conference 2001 of the Internal Audit Service of the European Commission


"The EU budget - Public perception and fact" was written in order to give the interested reader some information on how the European taxpayers' money is dealt with by the European Union, going a bit beyond the picture usually conjured up by the media.

During my five year's experience as chairman of the European Parliament's Committee on Budgets, I have had the privilege of taking part in the processes shaping the "real picture" of the EU budget. Assisted by an excellent secretariat, for me this has been the best job in the world.

On the basis of these five years, I would venture four conclusions as to this "real picture":

  1. Contrary to public opinion, Brussels is not the European centre of fraud and waste. The funds that - according to regular media reports - allegedly "disappear through fraud and waste" do not "disappear" in the corridors of the European Commission. EU money is almost exclusively spent in and by the Member States, this is where most irregularities occur. Admittedly, the European Commission might be blamed for not having adhered to some administrative financial rules to the last letter, but the European Parliament's and the Court of Auditors' criticism have led to a profound reform process within the Commission, creating clearer responsibilities, systems that are more transparent and more efficient control procedures.

  2. Very often it was the Member States that did not pay enough attention to whether all transactions with EU money were a hundred percent in line with regulations. Yes, we all know that EU rules can indeed be very complicated, so genuine mistakes are bound to happen (even if it is only a slight mis-measurement as in the case of our UK farmer mentioned above). But the criticism of financial irregularities aimed at the European Commission and the reforms that followed even had some repercussion in the Member States.

  3. The way the Statement of Assurance was done by the Court of Auditors over the first couple of years practically made it impossible for the European Commission to end up with a clean sheet. With the statistical method chosen it was next to impossible for the Court of Auditors to end up with granting a positive Statement. Because of this, the Court of Auditors itself changed its way of arriving at the DAS, moving from a method purely based on - far too few - samples to an evaluation of the efficiency of control systems. Now, the Statement of Assurance has developed into a means to locate weaknesses of the systems favourable to the occurrence of irregularities and to suggest possible remedies.

  4. The European Parliament through its close monitoring of EU expenditure has certainly contributed substantially to this development. The Budgetary Control Committee, in constant contact with the Court of Auditors, assesses year after year whether to grant discharge to the EU institutions. And it does not take this task lightly. The Budgets Committee, on the other hand, looks at the results of the expenditure of previous years, where there are problems, where there are under spends and where the best use of EU resources can apply.

Our role is to ensure that EU taxpayers' money is prudently allocated to meet EU priorities and at the same time ensure the accountability of the Executive to the public. Hopefully this is done in a democratic and reasonable manner on behalf of the people of Europe who finance the budget.


List of useful links in connection with the EU budget

Back to Budgets