Special Report No 14/2003 on the measurement of farm incomes by the Commission Article 33(1)(b) of the EC Treaty, together with the Commission's replies

Official Journal C 045 , 20/02/2004 P. 0001 - 0026

Special Report No 14/2003

on the measurement of farm incomes by the Commission Article 33(1)(b) of the EC Treaty, together with the Commission's replies

(pursuant to Article 248(4), second subparagraph, of the EC Treaty)

(2004/C 45/01)

CONTENTS

>TABLE>

SUMMARY

I. Pursuant to Article 33(1) of the EC Treaty, one of the five objectives assigned to the common agricultural policy (CAP) is to ensure a fair standard of living for the agricultural community, in particular by raising the individual earnings of persons engaged in agriculture. This has proved to be a real leitmotif running through the whole CAP.

II. The Commission uses three instruments to measure earnings in the agricultural sector: the Farm Accountancy Data Network (FADN), the Economic Accounts for Agriculture (EAA) and the statistics on the Income of the Agricultural Household Sector (IAHS). The main conclusions concerning these instruments are as follows:

(a) the FADN measures the income of farm holdings on the basis of harmonised accounts. However, substantial differences can be observed in the Member States concerning the determination of the field of survey, the selection of holdings and verification of the representativity of the sample. These differences affect the quality and comparability of the FADN results;

(b) as the EAA are not based on harmonised sources of information, the quality of the data provided by the Member States is very variable. An end-user of the EAA has no indicator which would make an assessment of the statistical accuracy of the data possible;

(c) the data drawn up in the framework of the IAHS also lack homogeneity; the latitude allowed the Member States concerning methodology is problematic as the differing approaches produce divergent results. In addition, the project has not been completed and the majority of the existing data are now obsolete.

III. The European Community has never defined the concepts of "agricultural community", a "fair standard of living" or "income" as they appear in Article 33 of the EC Treaty. There are therefore still no clear concepts or criteria which can be applied to the statistics intended to measure these variables.

IV. The Court found that the type of information required on the financial situation of farmers has changed over the years. Whereas, under the 1992 Mac Sharry reform, it was only a matter of compensating for income losses by direct payments, since then the Commission has also been interested in non-agricultural sources of income and in the distribution of income within the agricultural sector. Examination of these two questions requires information on the overall income of agricultural households.

V. The Court attempted to clarify the various concepts under Article 33 EC with the help of an expert. Examination revealed that, in the absence of other data, it is the overall disposable income of all the members of the household which may serve as an approximate indication of the standard of living. In addition, an agricultural household's standard of living can no longer be considered fair when the overall disposable income of all its members falls below a certain threshold as compared with the households of other types of entrepreneur.

VI. The Court studied the Commission's three tools for agricultural statistics from the point of view of their "pertinence" in view of CAP requirements. The aim of all these instruments is to measure income in the agricultural sector. However, it must be noted that the FADN only gives the profit of holdings run on a full-time basis without taking non-agricultural income or the income of other members of the household into account. As for the EAA, their main finding is the amount by which income exceeds costs for all agricultural products. These two instruments do not provide direct information on the standard of living of agricultural households. The IAHS project was designed to do this but has never produced reliable and comparable results.

VII. To conclude, at the present time the Community's statistical instruments do not provide sufficient information on the disposable income of agricultural households to allow an evaluation of the agricultural sector's standard of living. The Commission should therefore produce a new precise definition of Community requirements in this field and restructure the existing statistical instruments accordingly.

INTRODUCTION

1. According to Article 33(1)(b) of the EC Treaty, one of the objectives of the common agricultural policy (CAP) is to guarantee "the agricultural community a fair standard of living, in particular by increasing the individual earnings of people engaged in agriculture"(1). Although this is only one of the five objectives of agricultural policy expressly stated in the EC Treaty, the income of the agricultural community runs like a leitmotif through the CAP. This stands out, in particular, in the preamble of the Mac Sharry reform, Agenda 2000 and the mid-term review(2).

2. This is why the Court of Auditors devoted an audit to the methods used by the Member States for gathering data relating to agricultural incomes and the processing of these data by the Commission. The Court did not itself collect any data on incomes. Rather, it concentrated on verifying the procedures and evaluating the instruments used by the Commission to measure and estimate the earnings of the farming community.

THE COURT'S AUDIT

Audit approach

3. The standard of living of the agricultural community plays a major role in the discussions of the reform of the CAP. In order to be able to determine the consequences of a number of CAP reform measures on agricultural incomes, it is necessary to have reliable information on the composition of, and changes in, these incomes.

4. With regard to agricultural incomes, the Commission uses three statistical instruments: the Farming Accountancy Data Network (FADN), the Economic Accounts for Agriculture (EAA) and data from a special project devoted to the Incomes of the Agricultural Household Sector (IAHS). As the Member States form the basis of the statistical system and the Commission has a conceptual, coordinating and supervisory role in this system, the audit concentrated on the following points:

- the collection by the Member States of data for the three instruments mentioned above and,

- the pertinence of these instruments, i.e. their appropriateness to the needs arising in the context of management and development by the Commission of the CAP.

Implementation of the audit and audit criteria

5. The audit was carried out in six Member States (Germany, Greece, France, the Netherlands, Austria, Sweden) and at the Commission; the most recent available data examined concerned the financial years 1999 and 2000.

6. As the information concerning incomes constitutes statistical data, the quality requirements for statistical data developed by the European statistical system (Eurostat and the statistical offices of the community Member States) were used as audit criteria. ISO standard 8402 - 1986 defines quality as "the totality of characteristics of an entity that bear on its ability to satisfy stated and implied needs"(3). On this basis, the European statistical system has worked out a concept of quality applicable to statistics which takes the following aspects into account(4):

(a) completeness (complete coverage of the field);

(b) exactness (close estimate to the real (but unknown) value);

(d) comparability over time among the Member States and the various years;

(e) accessibility and clarity of information (for example user-friendly presentation, help for users);

(f) production in good time and punctuality (publication of the results at the appropriate time and within predetermined deadlines);

(g) relevance (usefulness of information for users);

(h) economy (relationship between costs and benefits of producing data). This criterion was added by the Court because it must always be asked whether the usefulness of a particular set of statistics justifies their actual cost.

STATISTICAL TOOLS FOR THE ASSESSMENT OF INCOMES

7. This chapter gives the main characteristics of the statistical instruments used by the Commission and the findings of the audits carried out in the six Member States visited.

The Farm Accountancy Data Network (FADN)

Legal basis and concept

8. Council Regulation No 79/65/EEC of 15 June 1965(5) constitutes the legal basis of the Farm Accountancy Data Network (FADN). The purpose of the FADN is to collect accounting data to permit "an annual determination of incomes on agricultural holdings" and "a business analysis of agricultural holdings"(6). This information comes from a sample of approximately 60000 farms, representing more than 90 % of agricultural production in the European Community.

9. The FADN is managed by the Commission. However, it does not collect the data directly, but relies on the Member States, which are represented within the Community FADN Committee by their liaison agencies. These liaison agencies are essentially responsible for the selection of the farms, the accuracy of the recorded data and the transmission of these data to the Commission within the established time(7). The representatives of the liaison agencies and the Commission meet at the Community FADN Committee approximately three times per year. This committee is consulted for the verification of the selection of the holdings, the critical analysis and evaluation of the annual FADN results and any other questions that are submitted to it(8).

10. The indicators calculated in the context of the FADN, based on the financial statements of the farms, concern gross and net value added and family farm income. Figure 1 explains how these three indicators are calculated.

Figure 1

Calculation of income indicators in the framework of the FADN

Final stock agricultural products

+ Sales

+ Correction livestock inventory values

+ Farmhouse consumption

+ Farm use

- Starter stock agricultural products

- Purchases of livestock

= Output (crop output plus livestock output)

+ Farm subsidies

- Intermediate consumption(9)

- Balance of VAT and farm taxes

= Farm gross value added (Gross farm income)

- Depreciation

= Farm net value added

+ Subsidies on investments

- Wages(10), rents and interest paid

= Family farm income

Source: Farm Incomes, Wealth and Agricultural Policy, Berkeley Hill, Ashgate Publishing Ltd, Hants 2000, third edition, p. 160.

Observations

Population coverage

11. Pursuant to Article 4(1) and (2) of Regulation No 79/65/EEC, the FADN field of survey covers agricultural holdings in excess of a given minimum size. A different minimum size is defined for each Member State in such a way that each field of survey covers at least 90 % of the standard gross margin (SGM)(11) of the farms of the Member State(12). This method makes it possible to ensure that the FADN includes only the data for agricultural holdings managed on a professional basis.

12. The audit nevertheless showed that the FADN field of survey is not defined in a uniform manner in all Member States; the following features, which are liable to affect the comparability of the results, were found:

(a) the economic size threshold of the holdings was set so high in one Member State, that, for 1995 and 1997, the field of survey covered less than 90 % of the SGM (Austria);

(b) farms in the form of corporations are proportionally under-represented (Netherlands);

Selection of the holdings to be included in the sample

13. In order to optimise the quality of the results obtained, the individual components of the sample must be selected at random(14). Nevertheless, in practice, two conditions stand in the way of genuine random sampling, the availability of suitable accounts documents and the fact that farmer participation in the FADN is voluntary (they cannot be forced to provide data). Disregarding these two factors, selection must be carried out as rigorously as possible in accordance with the principle of random sampling - the sample must be drawn at random from the subset of farms within the field of survey which fulfil the two conditions mentioned(15). As far as methodology is concerned, this is an essential condition for a valid extrapolation of the results of the sample.

14. The checks carried out in the Member States showed that the principle of the random selection of the holdings was not always complied with. In one of the Member States, the population of the holdings is divided up among several subsets, whose components do not all have the same probability of being taken up in the sample (France). In another Member State, a national selection committee examines the representativeness of each individual holding (Germany). In a third Member State, an official selects the holdings to be included in the sample at his own discretion (Greece).

15. Moreover, the period that the holdings remain in the sample varies significantly considerably from one Member State to another. An analysis of the Commission's data showed that the proportion of holdings that were selected in 1995 and still included in the 1999 sample ranged between 28 % and 88 %. This means that, in one case, only about 4 % of the holdings were replaced each year, while in the other, almost 24 % of the sample was renewed. These figures reflect two different concerns: whereas the "conservative" approach to the sample stresses the long-term development of the holdings contained in the sample, its rapid renewal aims at preserving its representativeness by monitoring the development of agricultural structures as closely as possible.

Representativeness of the sample

16. Pursuant to Article 4(2)(c) of Regulation No 79/65/EEC, the selected agricultural holdings must be representative of the FADN field of survey. The holdings are therefore divided up into cells, by type of farming, size and region in accordance with the Community typology. A sample is taken from these cells. From a technical point of view, the selection plan and the implementation report are the tools used for controlling the representativeness of the sample(16). Whereas the selection plan determines the number of holdings to be represented in the sample for each individual cell, the implementation report states the number of agricultural holdings actually selected.

17. An examination of the selection plans and implementation reports led to the following observations:

(a) in one Member State, the sampling rate is the only criterion applied to determine sample size; there are no statistical data on the stratification of the holdings according to the Community typology (Greece). It is therefore not possible to evaluate the representativeness of the sample in the various regions;

(b) in the same Member State, the selection plan was incomplete and had not been updated between 1983 and 1997, i.e. over a period of more than 14 years (Greece). Furthermore, this Member State did not draw up an implementation report for certain years and supplied incomplete reports for other years. The selection plan of another Member State did not follow the Community typology concerning economic size and farming type (Netherlands);

(c) there are sometimes considerable differences between the selection plan and the implementation report (Netherlands). These differences show that the selection initially envisaged was not properly implemented, or that the implementation reports do not accurately reflect reality;

(d) the Commission does not systematically make use of the selection and implementation reports. It does not regularly examine the representativity of the samples and there is no systematic reconciliation between the selection and implementation reports. According to the Commission, this comparison will be carried out automatically for the data concerning the financial year 2001, which will be available to the Commission at the end of 2003;

(e) the Commission has never clarified the concept of the representativity of the sample. The Member States rely on different indicators to assess representativity. The application of the indicator mentioned in Annex II to Regulation (EEC) No 1859/82(17) (the relationship between the number of holdings in the sample and the number of holdings in the corresponding population cell) led to substantial differences in several Member States. The Commission took no steps to harmonise the approach.

Deadline for the transmission of farm returns

18. The liaison agencies forward the farm returns to the Commission no later than nine months after the end of the accounting year to which they relate(18). A standard fee is paid by the Commission for each duly completed farm return that is forwarded to it within the period prescribed (132 euro in 2001(19)).

19. In practice, there are considerable delays in the forwarding of data. In 1997, an average delay of 133 days was noted; this was 238 days in 1998 and 139 days in 1999. In one of the countries visited, the data was regularly transmitted with a delay of approximately a year (Greece). It was therefore only possible to publish the results of the FADN for the financial year 1997-1998 in 2001, in the 1999 report on the situation of agriculture in the European Union, i.e. with a delay of approximately two years. The Commission has not penalised these delays by reducing the standard fee paid to the Member States concerned.

20. In July 2002, the Commission decided, in agreement with the Community Committee of the FADN, to have the period for the transmission of the data extended exceptionally from nine to 22 months for 2000 and 2001(20). The Commission justified this measure with the fact that foot-and-mouth disease and BSE had affected compliance with deadlines for forwarding the data. However, this extension might have led to Member States transmitting the data for 2002 before those for 2001, which would have caused administrative problems. The transmission period for the data concerning the 2002 accounting year was therefore extended, in turn, from nine to 15 months(21). For the data concerning the 2003 accounting year the maximum period of nine months will apply once more.

Verification of the data contained in the farm returns

21. The liaison agencies are required to collect the farm returns sent in by the accountancy offices and verify that they have been duly completed(22). A farm return is duly completed where its contents are factually accurate and the accountancy data given comply with the relevant Community provisions(23).

22. Until 2001 (data concerning the 1999 accounting year), the Commission itself checked the consistency of the data forwarded by the Member States. Amongst other things, the Commission's software checked whether the data contained in the farm returns were consistent and whether certain maximum and minimum thresholds had been complied with. The Commission reported any errors to the liaison agencies. It was the responsibility of these agencies to analyse the errors found, rectify them and forward the corrected data to the Commission. As the Commission only pays for farm returns which are completely free of errors, it was in the liaison agencies' interest to carry out the necessary corrections and send the corrected farm returns to the Commission. Under the new system for collecting and checking the data (FADN-1), the Member States carry out checks themselves with the help of programmes provided by the Commission. The liaison agencies are thus in a position to process the error reports directly.

23. The aim of the audit carried out in the Member States visited was to examine a sample of the errors reported by the Commission in order to establish the cause of the errors and ascertain that they had been corrected. As none of the Member States visited kept records of the reasons for the corrections made, it was not always possible to achieve this objective. In one Member State, where the national sample is much larger than required by the Community provisions, farm returns with errors are simply replaced with correct returns, but the errors are not rectified (Germany). Not all the Member States visited made use of the opportunities for cross-checking against other databases, for example the IACS(24) (Germany). A number of Member States do carry out such cross-checks with the aim of limiting, as far as possible, the risk of error but the Commission does not organise any opportunities for an exchange of information concerning the experience gained in this area.

Agricultural holdings organised as corporations

24. The Farm Accountancy Data Network (FADN) is wholly geared to traditional family farms. No separate methodology has been set out to deal with agricultural holdings organised as companies. This is clearly demonstrated by the FADN's main indicator, "family farm income". This indicator shows profit after deduction of interest, farm rent and labour costs. For the farmer and his family, this represents their return on land use, capital and labour. However, this indicator cannot be applied to legal entities, as it does not take into account the incomes of investors and those employed in agriculture(25).

25. Holdings organised as companies would not represent a major problem if there were only a few of them. But, actually, in two of the six Member States their role is by no means insignificant. In France, since the 1970s, the Government has created three types of company specially designed for the agricultural sector: the "groupement agricole d'exploitation en commun" (GAEC), the "entreprise agricole à responsabilité limitée" (EARL) and the "société civile d'exploitation agricole". The number of these companies increased from 3 % of agricultural holdings in 1979 to 17 % in 2000(26) and they accounted for 40 % of French agricultural production for 2000(27).

26. A similar situation exists in the new Länder in Germany. Many of the former agricultural cooperatives (Landwirtschaftliche Produktionsgenossenschaften) have been transformed into legal entities. The structure of these holdings (which are legal entities under private law and, in 2001, had an average agricultural area of 931,7 ha(28)) is fundamentally different from that of the traditional private holdings, whose average agricultural area is 27,9 ha.

27. It is not very meaningful to compare the profit of a limited company, after deduction of staff expenditure, with the income of an individual farmer. Nevertheless, the FADN remains wholly designed for traditional family holdings and does not provide any separate indicator for farms organised as companies. Provision has been made for differentiation between farm holdings on the basis of their legal status in the data for the 2002 financial year, which will be available in 2004.

The Economic Accounts for Agriculture (EAA)

Legal basis and concept

28. The EAA are drawn up within the framework of the European System of Integrated Economic Accounts (ESA) with the aim of providing comparable, up-to-date and reliable information on the economic situation of the agricultural sector and, more particularly, on the development of agricultural incomes(29). They have been published by Eurostat since 1964.

29. The aim of the EAA is to evaluate agricultural production and related costs. The point of departure is the farm holding and the entire agricultural production resulting from a full-time or part-time agricultural activity, which is defined on the basis of a list of activities. The holdings' secondary non-agricultural activities are not taken into consideration. These activities are nevertheless included in the value of production if it is impossible to distinguish them from the main agricultural activity.

30. The EAA are still based at present on a gentlemen's agreement, according to which the Member States have undertaken to comply with the concepts and rules of a common methodology for drawing up the EAA and a timetable for forwarding the data. In February 2003, the Commission submitted a proposal for a Regulation concerning the economic accounts for agriculture to the European Parliament and the Council(30). This proposal, which is currently passing through the parliamentary legislative procedure, is mostly based on the methodology which has been applied so far.

31. The production value is calculated by multiplying the quantities by the relevant prices. The values are provided by the statistics offices or ministries of the Member States on the basis of national statistics for each branch. Every Member State has its own particular source of information. Where statistics for the sectors are not available, other sources are used; for example, FADN statistics for previous years may be extrapolated.

32. Figure 2 gives an overview of the EAA and the indicators derived from them.

Figure 2

Income indicators derived from the Economic Accounts for Agriculture (AWU = annual work unit)

>TABLE>

Indicator A: Index of the real income of factors in agriculture per annual work unit.

Indicator B: Index of real net agricultural entrepreneurial income per unpaid annual work unit.

Indicator C: Net entrepreneurial income of agriculture.

Sources: Statistical Office of the European Communities, Income from agricultural activities in 2001 - European Union and Candidate Countries (2002 edition), Luxembourg 2003, p. 108; Manual on the Economic Accounts for Agriculture and Forestry EAA/EAF (Rev. 1.1), No 4.12.

33. Eurostat publishes a first estimate towards the end of the financial year concerned. It publishes a second estimate in the spring of the subsequent financial year and then presents final figures in the summer.

Observations

34. Two headings in the EAA 2000 were examined in the Member States audited in order to verify the exactness of the data - cereal production and "intermediate consumption". In the case of the first heading, the statistics available are generally of reasonably good quality. In the case of "intermediate consumption", it is more difficult to obtain reliable figures. The auditors retraced the various stages of the calculation of these two aggregates and checked the source of all the underlying data used.

35. Apart from two calculation errors (one negligible, the other significant) which were rectified immediately, the auditors found no inaccurate values. Nevertheless, they noted that the quality of the statistics varied considerably in the Member States visited. Whereas, almost all Member States visited had statistical data on production, either based on statutory provisions or collected voluntarily, in the case of "intermediate consumption" there were only estimates based on hypotheses, which varied from Member State to Member State.

36. Concerning the evaluation of the reliability of the various figures, Eurostat attempts to overcome the difficulties mentioned above in two ways. On the one hand, it has launched an initiative to draw up an inventory of the EAA, in which each Member State specifies the sources of its data and the methods it used to calculate the values forwarded to Eurostat. On the other hand, it has begun drawing up standard reports by country and financial year. Plausibility checks are carried out when these reports are produced and - where there are inconsistencies - the Member States are asked to provide explanations. Nevertheless, all the opportunities for cross-checking with data from other sources have not been exhausted and the end-user of the EAA has no indicator which would allow him to assess the statistical accuracy of the data.

Income of the Agricultural Households Sector (IAHS)

Legal basis and concept

37. In 1985, the Commission published a study of the prospects for the agricultural sector. In this study it estimated the number of farmers who would require direct income support because their income was insufficient, and the relevant costs for the Community budget(31). The discussion provoked by the Commission's Green Paper of 1985 showed that it lacked information on the total income of agricultural households. This is why Eurostat launched the "Total income of agricultural households" project. In 1997, the statistics relating to the total income of agricultural households were renamed statistics on the "Income of the Agricultural Households Sector" (IAHS) in order to give a more precise description of their contents and distinguish them more clearly from other statistics.

38. The aim of the statistics drawn up in accordance with the method based on the Agricultural Households Sector (IAHS) is to provide an instrument for measuring total income in order to:

(a) study changes, year-on-year, in the aggregate income of agricultural households in the Member States;

(b) track changes in the composition of this income, in particular in the percentage of this income that comes from agricultural holdings, other paid employment, property and social benefits;

(c) compare changes in the total income of agricultural households with those in the income of other socioeconomic groups by household, member of the household, and consumption unit;

(d) compare total farmers' income per unit with that of other socioeconomic groups(32).

39. The central concept of the IAHS is net disposable income. This includes, in addition to income from paid employment, rents and other transfer payments, the value of agricultural goods produced and consumed by the household and the rental value of owner-occupied dwellings. Taxes and social benefits, in particular, are deducted from the amount thus calculated.

40. The "Income of the Agricultural Households Sector" project has no legal basis; it is based on a gentlemen's agreement between the Member States of the European Community within Eurostat's Agricultural Statistics Committee (ASC). Figure 3 explains the calculation of net disposable income.

Figure 3

Calculation of net disposable income within the framework of the IAHS

Net operating surplus from independent activity

(a) from agricultural activity

(b) from non-agricultural activity

+ Compensation to members of agricultural households as employees, from agricultural and non-agricultural activity

+ property income received

+ non-life insurance claims (personal and material damage)

+ social benefits (other than social benefits in kind)

+ miscellaneous inward current transfers (paid)

= Total resources

- property income paid

- net non-life insurance premiums

- current taxes on income and wealth

- social contributions

- miscellaneous outgoing transfers (paid)

= Net disposable income

Source: Farm Incomes, Wealth and Agricultural Policy, Berkeley Hill, Ashgate Publishing Ltd, Hants 2000, 3rd edition, p. 184.

41. The unit of measurement covers all the persons living under the same roof. In order to distinguish agricultural households from other households, Eurostat has developed two definitions: in the narrow sense of the term, an agricultural household is a household where farming is the main source of income of the reference person - as a general rule, the head of the household. In the broad sense of the term, on the other hand, all households are considered to be agricultural households in which at least one member draws some income from an independent agricultural activity(33).

Observations

Microeconomic and macroeconomic approaches

42. As in the case of the EAA, the sources of data used for the IAHS in the various Member States have not been harmonised. To enable the Member States to apply the simplest method with regard to the information at their disposal, Eurostat allows them to choose between various approaches to determine the income of the agricultural households sector:

(a) the disposable income of the "agricultural households" sector may be obtained by the extrapolation of microeconomic data obtained from surveys of family budgets, tax records or surveys of the accounts of agricultural holdings;

(b) in the national accounts, the distribution of income account for the "households" sector is subdivided into separate sub-accounts for agricultural households and other socioeconomic groups. The distribution key is usually taken from a source of macroeconomic data (macroeconomic approach);

(c) a "hybrid" approach, where the income from agricultural activity is calculated on a macroeconomic basis but the other components making up disposable income are taken from microeconomic sources of information(34).

43. Of the six Member States visited, four had opted for the microeconomic approach. One Member State (Netherlands) compared the income calculated by the microeconomic approach, on the basis of a survey of households, with the income determined by the macroeconomic approach based on the EAA. The difference found between these two values was reduced by adjustments at both the microeconomic and macroeconomic levels (definitional difference). The figures for the 1997 and 1998 financial years are shown in Table 1.

Table 1

The Netherlands: difference between incomes calculated by the microeconomic approach and incomes calculated by the macroeconomic approach

>TABLE>

Source:

Centraal Bureau voor de Statistiek, The Hague (national statistics office of the Netherlands).

44. The latitude allowed the Member States to choose between the microeconomic and macroeconomic approaches to establish disposable income is questionable. In view of the differences in disposable income calculated for one Member State using these two approaches, 46 % and 48 % for 1997 and 1998 respectively, there must be doubts about the equivalence of the two methods. Differences of this kind also cast doubt on the possibility of comparing data supplied by Member States using different approaches. The Commission has not examined the possible sources of this anomaly with a view to drawing appropriate methodological consequences.

Difficulties in the implementation of the project

45. With regard to the calculation of the IAHS, using (unharmonised) microeconomic data, the Member States were repeatedly confronted with the problem of obtaining reliable data. For instance, one of the Member States visited extracted the data required from an annual survey of households, but the proportion of agricultural households was, in this case, so small, that this survey can no longer be considered representative. Faced with the impossibility, for financial reasons, of expanding the sample to include further agricultural households, the Member State concerned did not forward any more data to Eurostat (Germany).

46. Some Member States experienced difficulties in collecting the data from very disparate sources which were needed to be able to record all sources of income for the various members of the household. In addition, in some Member States, the rules governing tax secrecy and the protection of personal data hamper the use of the data. In one Member State, the statistics office can only use income data extracted from tax returns if they remain anonymous. These data cannot, therefore, be linked with certainty to agricultural holdings (France). Another Member State may use tax data but, as its tax system does not make any distinction between income from agriculture or from industrial or commercial activities, a significant percentage of the households covered by this survey cannot be linked with certainty to the agricultural sector (Netherlands). Yet another Member State avoids the difficulties by simply using IAHS farm data without making any allowance for methodological differences (Austria).

47. Table 2 shows the data forwarded as at 12 October 2001. The length of some delays demonstrates clearly the difficulties of implementing the project.

Table 2

Data provided for the calculation of the Income of the Agricultural Households Sector

>TABLE>

Source:

European Commission (Eurostat), Meeting of the "Economic Accounts for Agriculture" Working Party of the Agricultural Statistics Committee in Luxembourg on 28 and 29 November 2001, Doc. ASA/EEA/513, p. 37 ff.

48. In 2002, Eurostat and the Member States examined the difficulties affecting the collection of reliable data. On this occasion, four Member States stated that they were not interested in the project, five others mentioned major technical problems and six reported financial difficulties. Eurostat therefore acknowledged that the IAHS project had no priority and the contracts of the two experts who had been assigned to it until 2002 were not extended.

The cost of the FADN, the EAA and the IAHS

49. The Court asked the Supreme Audit Institutions (SAIs) of the Member States to forward information to the Court which would allow the cost of the FADN, the EEA and the IAHS and the benefits of using these sources of data to be evaluated. The information collected on the cost is shown in Table 3. The majority of SAIs emphasised certain advantages but the disparate nature of the information received allows no general conclusions.

Table 3

Cost of the FADN, the EAA and the IAHS

>TABLE>

Source:

Information forwarded by the SAIs of the Member States; Commission documents on reimbursements made under the FADN; investigations carried out by the European Court of Auditors; Annex I to the consolidated version of Commission Regulation (EEC) No 1859/82 of 12 July 1982 concerning the selection of returning holdings for the purposes of determining the income of agricultural holdings (OJ L 205, 13.7.1982, p. 5).

50. With regard to the costs, sufficiently accurate and comparable information is only available for the FADN and gives rise to the following conclusions:

(a) for most of the Member States, the Commission's reimbursements only represent a fraction of the total cost of the FADN (between 10 % and 20 %). However, one Member State receives a rather high rate of cover for its expenditure, amounting to approximately 50 % (Spain) and another a very low rate of approximately 2 % (Netherlands);

(b) the average cost, per agricultural holding, of acquiring the FADN data varies considerably from Member State to Member State, ranging from approximately 230 euro (Spain) to more than 2700 euro (Netherlands). Some Member States in fact commission accounting firms to draw up the annual accounts of the holdings in the FADN sample or, within the FADN framework, collect many more data than are required for the FADN itself, such as, for example, information concerning the use of fertilisers and pesticides.

THE PERTINENCE OF THE STATISTICAL TOOLS USED BY THE COMMISSION

51. The statistical tools applied to the agricultural sector are intended to monitor changes in the results achieved by the CAP in the context of the objectives of the Treaty. They are also applied to examine whether the agricultural community is achieving a fair standard of living in accordance with the objective laid down in Article 33 of the EC Treaty. If these instruments actually do provide suitable data for this purpose, thay may be considered "pertinent", i.e. as satisfying the audit criteria applicable to statistics (see paragraph 6).

52. The concepts "agricultural community" and "fair standard of living" contained in Article 33 of the EC Treaty nevertheless require precise definition before any evaluation of the "pertinence" of the statistical indicators relating to them can be carried out. For this purpose, the Court examined what had been said at the major points in the development of the CAP to clarify the content of Article 33 of the EC Treaty. In addition, the Court asked an agricultural expert to help develop a working interpretation of the concepts "agricultural community" and "fair standard of living".

The field to be covered by the statistics on agricultural income

Changes in the socioeconomic context of the CAP

The early days of the common agricultural policy

53. The early days of the CAP were characterised by the effort to emerge from the difficult situation of the post-war period and achieve self-sufficiency. These two objectives were to be achieved by increasing the productivity of farms(35). At the Stresa conference in 1958, the Agriculture Ministers and the Commission decided to encourage productivity by supporting the prices of agricultural products(36). The European Community gave this policy a legal framework by creating market organisations for the most important agricultural products.

54. Price support was meant not only to stimulate productivity, but also to result in an increase in the earnings of the agricultural community. By supporting prices, the Commission intended to ensure a fair income for people who had worked full-time over the years in well-managed and economically viable agricultural holdings - in other words by no means in all agricultural holdings, nor even in average holdings(37).

55. Already at the Stresa Conference, the Agriculture Ministers of the Member States had drawn attention to the relatively unfavourable situation of farm incomes(38). However, it was not specified what was to be understood by a "fair income" or by an "income level comparable with that of industrial workers". The question of knowing which income should be measured and what comparison should be carried out remained open.

56. Similarly, the expression "agricultural holding" was not clarified at the time. In 1956, the Spaak report, which heralded the Treaty of Rome(39), still regarded the family holding as the fundamental form of agricultural activity in the European Community, both past and future(40). In 1960, however, the Commission was already describing an agricultural holding more broadly as "an agricultural structure in which farms of various types and sizes complement each other, and in which the farm operating with paid labour - a farm that is sound from the socioeconomic point of view - will have its place"(41).

57. In short, the European Community has never specified what is meant to be understood by either "the agricultural community" or a "fair standard of living". Neither the Council nor the Commission have given these concepts any meaning that is binding for them(42). Nevertheless, it is clear from the work at the Stresa Conference that "fairness" is to be defined in comparison with other sectors.

The move towards a broader and more complex concept

58. In the 1985 Perspectives Green Paper, the Commission examined the need for income support for farmers with low or falling incomes. The basic idea underlying the social approach was that structural change in agriculture should be gradually implemented in a way that would cushion farmers from the shock of structural adjustment, while still maintaining pressure on farmers to adapt their holdings to market needs. Thus, as long as no alternative income and employment solutions are available to farmers, any income aid scheme for farmers should help avoid social hardship(43).

59. This new approach was received by farmers with little enthusiasm. They distrusted direct income aid as "welfare" and were afraid such income support would preclude them from having multiple activities(44). This is why, in 1992, the Mac Sharry reform limited itself to replacing price maintenance only partially by revenue support for agricultural producers(45). This essentially involved three measures: a fall in production prices, compensation for this fall in income by direct payments to producers and measures to limit output, such as the set-aside of arable land(46).

60. Agenda 2000 shifts further away from price support to income support by lowering intervention prices for cereals, beef and dairy products, and by increasing the level and scope of "area" aid and animal premiums in order to compensate for potential income losses. Two of the five major objectives of the Agenda 2000 CAP reform refer to farmers' incomes: to ensure a fair standard of living for the agricultural community and the stability of farm incomes; and to promote supplementary or alternative sources of employment and income in rural areas and thus contribute to economic cohesion within the EU(47).

61. In the mid-term review, the Commission emphasised that ensuring a fair standard of living for the agricultural community and contributing to the stability of farm incomes remained key objectives for the CAP. In this same document, it was noted that the growth of the average farm size, the specialisation in certain crops and the concentration of production in certain regions had put pressure on traditional farming methods and mixed farming. It also stated that many of these farms would need more targeted support to adapt to the opportunities offered by more open markets and consumer demand for quality products. Moreover, the Commission observed that a minority of farmers benefit from most of the direct payments and questioned whether the distribution of direct support was optimal(48).

62. A comparison of the programmes clearly shows that the requirements as regards information on the financial situation of the agricultural community have changed: whereas, at the outset, it was a matter of compensating for the loss of income via direct payments, since Agenda 2000, the Commission has also been interested in non-agricultural sources of income. Moreover, the mid-term review document also deals with the distribution of incomes and the question of knowing whether the final beneficiaries of direct payments are in a precarious position. However, these questions (concerning other sources of income and the distribution of incomes) can only be answered if the overall disposable income of the various holdings is known.

63. The Commission is also aware of the changed demands as regards information on the farmers' financial situation, in particular for the structural policies: "The integrated development of rural areas is becoming an ever greater concern of sectoral policy, including agricultural policy (...). This generates new needs in terms of statistical data, and particularly the availability of information (...). These new requirements are for more detail in terms of the territorial breakdown of data. Moreover, these requirements also extend to information on the non-farming component of the income of holdings and families, particularly activities that supplement farming activity and the community services provided by farmers"(49).

Proposed interpretation of the concepts contained in Article 33 of the EC Treaty

64. In short, the European Community has hitherto failed to interpret the very generic objective of a "fair standard of living" and subdivide it into specific, precise and measurable objectives. Moreover, there is also no definition of the "agricultural community" concept either(50). However, the "major stages" in the development of the CAP can provide us with information regarding the aspects of the standard of living of the agricultural community that interest the European Community, i.e., in particular, the existence and scale of non-agricultural earnings and the distribution of incomes.

65. In the absence of a clear definition of the concepts of "agricultural community" and "fair standard of living", the Court commissioned an expert(51) with a view to clarifying the meaning of these concepts, which are important for an evaluation of the Commission's statistical tools. The results of the expert's appraisal are summarised in the Annex.

Impact of the interpretation of Article 33 on the Commission's statistical tools

66. The interpretation of Article 33(1)(b) of the EC Treaty establishes the following criteria for the development of the Commission's statistical tools: the standard of living in the agricultural sector is not fair when all the disposable incomes of all the members of the farming household together fall below a certain threshold, which must be established, at the political level, in relation to other entrepreneurial households. In order to give an opinion on the question of the fairness of agricultural incomes, the Commission must therefore have relevant information on the overall disposable income of individual agricultural households and the households of comparable entrepreneurs. It also needs this information to be able to study the other sources of farmers' incomes or the distribution of incomes within the farming sector.

Evaluation of the pertinence of the Community's statistical tools

67. The Court examined the statistical tools applied by the Commission and in the Member States in order to verify, in particular, whether these tools allow conclusions to be drawn in respect of the fairness of the standard of living of the agricultural community.

The Farm Accountancy Data Network (FADN)

68. The FADN's strength lies in its microeconomic database. As agricultural holdings are classified, under the European typology, according to their size, region and activity, a differentiated analysis is also possible. The extensive farm returns control programme, which has been accessible on the Internet since 2002 (data for the financial year 2000) and guarantees a precise control over the data, is another positive aspect. The fact that the data have been made available on the Internet in the course of the last few years, thus giving researchers a considerable degree of access, is an additional asset.

69. The FADN is based, however, on holdings and not on households. The incomes of the members of the agricultural household other than the farmer and the incomes of other persons are therefore not taken into account. Moreover, only income from agricultural activities is entered in the FADN accounts. Other sources of income are ignored completely. Figure 4 shows that, in 12 Member States out of 15, income generated by agricultural activity as such accounts for at least half the total income of agricultural households in the narrow sense of the term. However, in nine Member States agricultural households only draw between 25 % and 60 % of their income from agricultural activity. The data available indicate that the percentage is very much lower if the population of agricultural households in the broad sense of the term is taken.

Figure 4

Farm income as a percentage of total income of farm households

>PIC FILE= "C_2004045EN.001701.TIF">

Source of narrow definition:

Eurostat, Income of the agricultural households sector - 2001 Report, Luxembourg 2002. The figures for Sweden, Spain and Portugal include income from independent agricultural and non-agricultural activities.

Source of broad definition:

Organisation for Economic Cooperation and Development, Working Party on Agricultural Policies and Markets: Synthesis report on farm household income issues in OECD countries (AGRÉÉ/CA/API(2002)11/Final), 29.10.2002, p. 14.

In order to distinguish between agricultural households and other households, the OECD uses the definitions adopted by Eurostat: in the narrow sense of the term, an agricultural household is a household where agriculture is the main source of income of the reference person - as a general rule, the head of the household. In the broad sense of the term, on the other hand, all households are considered to be agricultural households where at least one member draws some income from an independent agricultural activity (European Commission, "Income of the Agricultural Households Sector", 1999 report, Luxembourg, p. 8).

70. The FADN only covers professionally managed holdings. In practice, this means that the FADN only covers 65 % of all farmers in the Community. As the economic size threshold for the holdings is not always the same, this rate varies considerably from one Member State to another: while the FADN only covers 41,9 % of Austrian farmers, the rate covered in Ireland comes to 87,9 %(52).

71. As the audit, which was carried out in a number of Member States, revealed, in contrast to the high degree of standardisation displayed by the farm returns, discrepancies still exist between the Member States with regard to population size, the procedures for selecting representative samples and the selection of new holdings for inclusion. This limits opportunities for drawing comparisons.

72. In general, it can therefore be said that the FADN gives a largely reliable picture of the disposable incomes of individual farmers living alone and with no other income other than agriculture. If the holdings are organised on a different legal basis, if the household is made up of several professionally active persons or the farmer has sources of income other than agriculture, the FADN's "family farm income" does not give an accurate indication of disposable income.

The Economic Accounts for Agriculture (EAA)

73. As the first estimate for the EAA is published before the end of the reporting year, one of its great advantages is that it provides results that are available quickly.

74. However, the income aggregates and indicators used are not indicators of farmers' total incomes or their disposable incomes, as they may have sources of income other than agricultural production. The agricultural incomes calculated in the context of the EAA must therefore, under no circumstances, be taken as the equivalent of the total earnings of the farmers or the farm households.

75. In addition, it should be borne in mind that the figures forwarded are average values, which give no indication of the distribution of incomes. If, in addition, one bears in mind that the denominator used to calculate indicators A and B (see Figure 2), i.e. the annual work unit(53), has not been completely standardised, it is impossible to compare incomes in absolute terms between the Member States. Only changes within each Member State can be compared. No conclusions should thus be drawn concerning the absolute value of the incomes. For this reason Eurostat has refused for some years to draw such conclusions and has confined itself to a comparison of changes in incomes observed within the various Member States(54).

The Income of the Agricultural Households Sector (IAHS)

76. The income of the agricultural households sector project was designed to overcome an information deficit with regard to the real standard of living of the agricultural community. Net disposable income calculated using the IAHS offers the advantage that it is not based on agricultural products or the holding but on the household. In addition, it allows various other sources of income to be taken into consideration, in addition to income from capital. Overall, therefore, because of its structure, the IAHS appears greatly superior to the FADN and the EAA for the evaluation of living standards. Where the IAHS is determined on the basis of the macroeconomic approach, reservations are called for since it says nothing about the distribution of incomes.

77. The IAHS has never achieved reliable and comparable results, however, because of technical and financial obstacles and because the Commission and a number of Member States do not give it high priority.

CONCLUSION

78. The three Commission instruments which were examined - the FADN, the EAA and the IAHS - are all intended to determine agricultural income. When subjected to closer scrutiny, it is apparent that these three instruments have not been designed for the same purposes. The EAA is used for calculating surplus income in relation to the costs of all agricultural products. The aim of the FADN is to calculate the income earned by agricultural holdings managed on a professional basis. The FADN and the EAA do not provide any direct information on the standard of living of the agricultural community (Article 33(1)(b) of the EU Treaty). This is what the IAHS was designed to do. However, this project has never produced reliable and comparable results.

79. To conclude, at the present time the Community's statistical instruments do not provide sufficiently exhaustive information on the disposable incomes of agricultural households and do not allow an assessment of the living standard of the agricultural community to be made.

RECOMMENDATIONS

80. The Commission should evaluate the statistics available at present concerning the situation of holdings and agricultural households and then define precisely the statistical framework it requires to monitor achievement of the objectives of the CAP, in particular those set out in Article 33 of the EC Treaty.

81. The FADN measures the incomes of agricultural holdings. For the results to allow comparison between the various Member States, the Commission has devised numerous rules governing the content of the data, the controls on them and the calculation of indicators. However, the room for manoeuvre is so large that the differences between the Member States may impair the comparability of the results. The Court recommends that a multi-disciplinary group should be set up to review these differences, evaluate them and propose the necessary improvements.

82. The EEA are based on various types of sources depending on the Member States. As the origin and quality vary from one Member State to the other, the Court recommends that the transparency of the data should be improved by drawing up an inventory and introducing standardised controls. In addition, it would be very useful if the user of the EAA could form an opinion of the quality of the data with the help of precise statistical indicators.

83. With regard to the IAHS, the Commission should evaluate the project and decide under what conditions it can be pursued. The Court recommends that the Commission should opt for a uniform approach, either macroeconomic or microeconomic, on the basis of prior definition by the Commission of the statistical information which it absolutely must have if it is to monitor achievement of a fair standard of living for the agricultural community.

This report was adopted by the Court of Auditors in Luxembourg at the Court meeting of 4 December 2003.

For the Court of Auditors

Juan Manuel Fabra Vallés

President

(1) The wording of Article 33 of the EC Treaty is as follows:

"1. The objectives of the common agricultural policy shall be:

(a) to increase agricultural productivity by promoting technical progress and by ensuring the rational development of agricultural production and the optimum utilisation of the factors of production, in particular labour;

(b) thus to ensure a fair standard of living for the agricultural community, in particular by increasing the individual earnings of persons engaged in agriculture;

(d) to assure the availability of supplies;

(e) to ensure that supplies reach consumers at reasonable prices.

2. In working out the common agricultural policy and the special methods for its application, account shall be taken of:

(a) the particular nature of agricultural activity, which results from the social structure of agriculture and from structural and natural disparities between the various agricultural regions;

(b) the need to effect the appropriate adjustments by degrees;

European Union, Consolidated Treaties, Office for Official Publications of the European Communities, 1997."

(2) Communication from the Commission to the Council and Parliament, Perspectives for the common agricultural policy, Luxembourg, 15 July 1985, COM(85) 333 final, p. 20 f; Commission of the European Communities, Proposals for Council Regulations (EC) concerning the reform of the common agricultural policy, Brussels, 18 March 1998, COM(1998) 158 final, p. 5 (paragraph 2.1); Commission of the European Communities, Communication from the Commission to the Council and the European Parliament, mid-term review of the common agricultural policy, Brussels, 10 July 2002, COM(2002) 394, pp.1, 7 ff. (explanatory memorandum).

(3) Eurostat, Assessment of the Quality in Statistics, Luxembourg, 4 to 5 April 2000, p. 2.

(4) Eurostat, Assessment of the Quality in Statistics, Luxembourg, 4 to 5 April 2000, p. 2.

(5) OJ 109, 23.6.1965, pp. 1859-1865.

(6) Article 1(2) of Regulation No 79/65/EEC.

(7) Article 6(1) of Regulation No 79/65/EEC.

(8) Article 20 of Regulation No 79/65/EEC.

(9) Intermediate consumption consists of the costs of certain supplies (including inputs produced on the holding) and miscellaneous costs associated with production during the accounting year. Examples: seed, manure, cattle feed, etc.

(10) Including wages paid to members of the family under a contract of employment.

(11) The standard gross margin is the balance between the standard value of production and the standard value of certain specific costs (Article 3 of Commission Decision 85/377/EEC of 7 June 1985 establishing a Community typology for agricultural holdings (OJ L 220, 17.8.1985, p. 1)).

(12) European Commission, Community Committee for the Farm Accountancy Data Network (FADN), Revision of thresholds in FADN selection plans (RI/CC 1321), Brussels, 1 December 2000, p. 2.

(13) The CAP is applicable without restrictions in the French overseas departments. For essentially budgetary reasons, the French authorities do not, however, ask for the inclusion of these departments in the list of FADN areas annexed to the basic FADN Regulation.

(14) Manfred G. Korter, Stichprobenverfahren für Revisoren und Controller, 5.2.6. Ottokar Schreiber Verlag, 1st edition, Hamburg, 1997.

(15) http://europa.eu.int/comm/ agriculture/rica/ methodology2_en.cfm.

(16) Article 6(1)(b) of Regulation No 79/65/EEC.

(17) Commission Regulation (EEC) No 1859/82 of 12 July 1982 concerning the selection of returning holdings for the purpose of determining incomes of agricultural holdings (OJ L 205, 13.7.1982, p. 5).

(18) Article 3 paragraph 1 of Commission Regulation No 1915/83/EEC.

(19) Commission Regulation (EC) No 368/2001 of 23 February 2001 fixing the standard fee per farm return for the 2001 accounting year of the Farm Accountancy Data Network (OJ L 55, 24.2.2001, p. 40).

(20) Commission Regulation (EC) No 1251/2002 of 11 July 2002 (OJ L 183, 12.7.2002, p. 9).

(21) Commission Regulation (EC) No 1409/2003 of 7 August 2003 derogating from Regulation (EEC) No 1915/83 on certain detailed implementing rules concerning the keeping of accounts for the purpose of determining the incomes of agricultural holdings (OJ L 201, 8.8.2003, p. 8).

(22) Article 6(1)(d) of Regulation No 79/65/EEC.

(23) Article 4 of Commission Regulation (EEC) No 1915/83.

(24) Council Regulation (EEC) No 3508/92 of 27 November 1992 establishing an integrated administration and control system for certain Community aid schemes (OJ L 355, 5.12.1992, p. 1).

(25) In the case of holdings organised as corporations, the only two meaningful indicators are "farm net added value" or "farm gross added value", which are less meaningful in this context as far as the level of income is concerned.

(26) If we take into account the FADN field of observation (professional holdings), the "Structures" survey in France gave for 1997 25,4 % of holdings as being registered as companies.

(27) European Commission, Directorate-General Agriculture, Summary Report on the 141st FADN committee meeting in Midi-Pyrénées, VI. A.3/YP D(99), p. 2. Ministry of Agriculture and Fisheries, Agreste Primeur No 93-June 2001.

(28) German Federal Ministry of Consumer Protection, Food and Agriculture, summary of the report by the Federal Government on food and agricultural policy 2002, Berlin, 2002, Table 8, p. 12.

(29) Commission of the European Communities, Proposal for a Regulation of the European Parliament and of the Council on the economic accounts for agriculture in the Community, Brussels, 4 February 2003, COM(2003) 50 final, first preamble.

(30) Commission of the European Communities, Proposal for a Regulation of the European Parliament and of the Council on the economic accounts for agriculture in the Community, Brussels, 4 February 2003, COM(2003) 50 final.

(31) Commission of the European Communities, COM(85) 333 final, p. 56 ff.

(32) European Commission, Income of the Agricultural Households Sector, 1999 Report, Luxembourg, 2000, Table 1.1, p. 104.

(33) European Commission, Income of the agricultural households sector, 1999 Report, Luxembourg, 2000, p. 103 ff.

(34) European Commission, Income of the Agricultural Households Sector, 1999 Report, Luxembourg, 2000, p. 106 ff.

(35) This also explains the reference in Article 33(1)(b) of the EC Treaty ("thus") to the increase in productivity (Article 33(1)(a) of the EC Treaty).

(36) Collection of documents from the agricultural conference of the Member States of the European Economic Community held at Stresa from 3 to 12 July 1958, pp. 215 and 222.

(37) Hans-Broder Krohn and Jacques van Lierde, Proposed Criteria for Fixing the Prices of Agricultural Products in the European Economic Community, International Journal of Agrarian Affairs, Vol. 111 (1963), p. 246.

(38) Agriculture conference of the Member States of the European Economic Community from 3 to 11 July 1959, quoted from: Gottsmann, Der Gemeinsame Agrarmarkt, Nomos Verlagsgesellschaft Baden-Baden, Loseblattsammlung, Stand 1.8.1998, I A 1, p. 2 ff.

(39) At the Messina conference in 1955, the Foreign Ministers of the Member States of the ECSC set up a working party, chaired by Paul-Henri Spaak, to study the possibilities for increasing economic integration. The Spaak Report (April 1956) contained the first drafts of the Community Treaties providing for the creation of the EEC and Euratom.

(40) "Comité intergouvernemental créé par la Conférence de Messine, Rapports des chefs de délégation aux Ministres des Affaires étrangères", 21 April 1956, p. 47.

(41) European Economic Community Commission, Proposals for the Working-Out and Putting Into Effect of the Common Agricultural Policy in Application of Article 43 of the Treaty establishing the European Economic Community, Brussels, 30 June 1960, VI/COM(60) 105, II/13 and 14.

(42) Berkeley Hill, Farm incomes, wealth and agricultural policy, Ashgate Publishing Ltd, 3rd edition, Hants 2000, p. 28.

(43) Commission of the European Communities, COM(85) 333 final, p. 56 ff.

(44) Nigel Robson, The Changing Role of Part-Time Farming in the Structure of Agriculture, Journal of Agricultural Economics, Vol. XXXVIII No 2, May 1987, p. 169.

(45) Communication of the Commission to the Council - the development and future of the CAP - reflections paper of the Commission, COM(91) 100 final, p. 12.

(46) Commission of the European Communities, Communication of the Commission to the Council - the development and future of the CAP - reflections paper of the Commission, COM(91) 100 final, p. 13 and p. 14.

(47) Commission of the European Communities, Proposals for Council Regulations (EC) concerning the reform of the common agricultural policy, Brussels, 18 March 1998, COM(1998) 158 final, p. 5 (No 2.1).

(48) Commission of the European Communities, Communication from the Commission to the Council and the European Parliament, Mid-Term Review of the Common Agricultural Policy, Brussels, 10 July 2002, COM(2002) 394, pp. 2, 7 ff.

(49) Second Commission report to the European Parliament and the Council on the state of progress of the implementation of Council Decision 96/411/EC relating to the improvement of Community agricultural statistics, paragraph 5.9, p. 12.

(50) Hill, loc.cit. p. 28.

(51) Professor Stephan von Cramon of the University of Göttingen.

(52) http://europa.eu.int/comm/ agriculture/rica/ methodology2_en.cfm.

(53) An Annual Work Unit represents the labour provided by one person in full-time employment on an agricultural holding (Eurostat, Statistics in Focus, Theme 5-2/2003, p. 7).

(54) See previous situation: European Commission, Income from agricultural activity in 1998, Luxembourg, 1999, Chapter 6, p. 95 ff.

ANNEX

MAIN CONCLUSIONS OF THE EXPERT APPRAISAL CONCERNING THE INTERPRETATION OF ARTICLE 33 OF THE EC TREATY

1. First of all, it appeared necessary to look at the question of the fairness of the standard of living in terms of a comparison with a given level(1). This approach presupposes that we:

(a) define the subject of the comparison - the standard of living;

(b) identify the target group; specify whose standard of living is to be measured;

(d) establish a threshold below which the standard of living can no longer be considered "fair".

2. The considerations summarised below are based on published literature as well as the expert appraisal carried out for the Court.

The subject of the comparison: "the standard of living"

3. "Standard of living" is a concept of social well-being. The quality of life is derived from consumption, or in this case, from potential consumption. The latter comes from the aggregation of two variables - income and wealth. Potential consumption, associated with the environment and personal factors such as health, gives the individual standard of living (see Figure below)(2).

Figure

Factors which affect the standard of living

>PIC FILE= "C_2004045EN.002002.TIF">

Source:

Prof. Stephan von Cramon, Department of Agricultural Economics, Göttingen University.

4. Defined in this way, a person's "standard of living" is very difficult to measure in practice. This is why the Court followed an approach involving the measurement of disposable income only(3).

The target group: "the agricultural community"

5. Given that "standard of living" and "disposable income" can only concern natural persons or groups of persons, the holding cannot be used as a reference unit in the comparison. The choice of the household as the unit is justified by the fact that, in agricultural households, incomes are generally pooled and that the members of the household take decisions on consumption together. The reference unit for the comparison must therefore be the household and not the individual farmer(4).

6. The last stage in the definition of the target group consists in distinguishing agricultural households from non-agricultural households. In accordance with the two criteria most often used for this purpose (income and working time), an occupation is held to be agricultural where agriculture constitutes the main source of income or the main activity. This distinction implies the exclusion of certain agricultural populations, such as persons who carry out a farming activity as a secondary occupation or the owners of land that is leased and used for farming(5).

The reference group

7. The Community institutions have not specified whether the criterion for comparison should be set at the regional, national or European level or whether comparisons must be made within the agricultural sector or between this sector and other social groups(6). However, as early as the 1958 Stresa Conference, the Agriculture Ministers made it clear that they were concerned, above all, by the relatively less favourable situation of farm incomes(7), but did not specify which occupational groups should be used as a benchmark. Since the majority of farmers work on their own initiative and at their own risk, it is logical to compare them with other entrepreneurs(8).

The definition of "fairness"

8. Any farmer's income that is below a certain threshold may be considered "unfair"(9). This individualised approach disregards mean values. It refers to the distribution or allocation of incomes(10).

9. Like the so-called "poverty line", it is difficult to establish a limit below which the standard of living can no longer be considered to be fair(11). The European Communities have not, to date, solved this problem. There is therefore no criterion for comparison to enable us to appraise the fairness of agricultural incomes.

(1) Professor Stephan von Cramon, Department of Agricultural Economics, University of Göttingen, Aspects of the Commission's Management of the "Fair Standards of Living" Treaty Objective, p. 1.

(2) von Cramon, loc. cit. p. 6 ff.

(3) Disposable income: Total income-compulsory expenditure (e.g. taxes and social security contributions).

(4) von Cramon, loc. cit. p. 8.

(5) von Cramon, loc. cit. p. 10.

(6) von Cramon, loc. cit. p. 3.

(7) Agriculture conference of the Member States of the European Economic Community from 3 to 11 July 1959, quoted from; Gottsmann, Der Gemeinsame Agrarmarkt, Nomos Verlagsgesellschaft Baden-Baden, Loseblattsammlung, 1.8.1998, I A 1, p. 2 f.

(8) von Cramon, loc. cit. p. 5.

(9) The same procedure is generally followed for the definition of the concept of poverty (see, for example, von Cramon, op. cit., p. 1 and p. 38).

(10) von Cramon, loc. cit. p. 1 and p. 14.

(11) von Cramon, loc. cit. p. 1.

The Commission's replies

SUMMARY

II. (a) Improving the quality and especially the comparability of the statistical data from the FADN is a major concern of the departments responsible for these statistics and has been the subject of constant attention on the part of the Commission, the FADN Committee and the Member States. The latter have introduced a basic methodology designed to improve quality and harmonise the results from the FADN. It is true that the regional diversity of agriculture - in both socioeconomic and structural terms - has meant the adoption of rules specific to some Member States. However, the impact of the methodological differences referred to by the Court is reduced as much as possible by the efforts made regarding sampling and weighting.

(b) The economic accounts for agriculture are continuously and closely monitored for accuracy and comparability by the Working Party "Agricultural accounts and prices" and more specific controls such as the inventories. This is why the quality of the income statistics produced under the EAA has always been considered satisfactory. In addition, although the choice of sources of data is left to the Member States depending on the specific structure of their agricultural sector and their statistical system, this has little impact on the quality of the EAA data. It is in fact not the harmonisation of data sources but the harmonisation of definitions and principles of data extraction which guarantees the harmonisation of the EAA data among Member States and contributes to their quality. Finally, studies have shown that the multiplicity of data sources makes accuracy indicators very difficult to establish and complex to interpret for such statistics as national accounts and accounts for agriculture.

(c) The statistics on incomes of agricultural households do not benefit from the same degree of quality, homogeneity and updating as the FADN and the EAA. A taskforce to develop and improve these statistics was set up in 2003 in collaboration with the FAO, the OECD and UN/ECE.

III to IV. There are various different measures of income (inter alia, income by branch of activity and income of households at macroeconomic level, and income of holdings at microeconomic level). These have been extensively used and published by the Commission for the purposes of monitoring and analysing the levels of and changes in incomes in agriculture, as well as contributing to generating agricultural policy.

VI. The statistical instruments represented by FADN and EAA have made it possible from their inception to monitor, evaluate and analyse changes in the income generated by agricultural activity, both across the agricultural sector as a whole and by farm specialisation. Moreover, since the most significant part of farm households' income comes from agricultural activity (and mainly commercial holdings), these statistics also make it possible to monitor and analyse changes in farm households' income in a relatively reliable way. In fact, this is in line with paragraphs (a) and (b) of Article 33, which explicitly link the goal of ensuring a fair standard of living for the farming community to increases in income generated by agricultural activity.

VII. Each of the statistical tools provides additional information in the decision-making process. Therefore, they should be seen as complementary parts in a very complex policy field.

The statistics of the EAA and of the FADN have for many years provided very useful, reliable and relevant information on the incomes generated by agricultural activity. This information has proved very important for the monitoring, evaluation and design of the common agricultural policy. So far these figures on income from agriculture could be regarded as relatively reliable estimates of farm households' income to the extent that the latter draw the major part of their income from farming. Equally, the use of these statistics is in line with paragraphs (a) and (b) of Article 33 explicitly linking the goal of ensuring a fair standard of living for the farming community to increases in the income generated by agricultural activity.

The socioeconomic developments in the agricultural sector and recent changes in agricultural policy (in particular as regards rural development) may well require more detailed figures on farm income. Although reliable statistics on the income of agricultural households already exist in a considerable number of Member States, it would be worth making a study of the feasibility and cost of statistics that measure the standard of living of the farming population across all the Member States of the Union.

The financial situation and income levels of farming undertakings and in the agricultural sector generally will continue to be the subject of regular monitoring on the basis of FADN and EAA data, the harmonisation and consistency of which will need to be strengthened through increased collaboration among the various services concerned.

STATISTICAL TOOLS FOR THE ASSESSMENT OF INCOMES

12 to 14. The threshold for the economic size of holdings, defining the field of survey, has been adopted in accordance with the management committee procedure laid down in Article 19 of the Council Regulation. This threshold has to reflect the diversity of agriculture in the Member States, and a uniform definition would not be feasible. The FADN data have to be representative by region, type of farming and economic size class. The Commission has developed a new tool to monitor the representativeness of the sample.

12. (a) For Austria, a lower threshold would have been more appropriate at the time of accession. The issue has to be reconsidered taking into account the results of the 2003 farm structure survey.

(b) No separate sub-sample for companies exists in the Netherlands. This does not, however, prevent such holdings from existing in the sample.

13 to 14. The Commission follows the representativity of the selection plans and the results in respect of the three criteria set in Commission Regulation (EEC) No 1859/82 (region, type of farming and economic size class). Member States have been allowed to do the selection in a way that best suits to their conditions. In some Member States the FADN makes use of existing national data collection networks that take into account certain country-specific factors of their agriculture. Keeping in mind that participation is voluntary for farmers, the Commission can only recommend that sampling should be random.

15. For business analyses, continuity of the sample is needed. Therefore, neither a low nor a very high renewal rate is desirable. The Commission recommends a continuous renewal of the sample, but taking into account that participation is voluntary, has not imposed strict rules on how long a farm should stay in the FADN sample. Therefore, differences between Member States or regions occur.

17. (a) In the new selection plan for Greece, discussed in the FADN Committee in December 1999, information concerning selection rates exists.

(b) The Commission has sent several requests to the Greek authorities on this matter with a view to improving the situation. It has also financed specific actions to improve Greece's FADN statistics.

(c) Participation in the FADN is voluntary for farmers. Depending on experience with the new data delivery and control system (FADN-1), the Commission is considering whether implementation reports could be suppressed.

(d) Until 2001, the comparison of selection plans with the results had to be done by analyses. These analyses have been discussed in the FADN Management Committee. The latest discussion took place in November 2002.

(e) The requirements for selection plans are set out in Article 4 of Commission Regulation (EEC) No 1859/82. The selection plan is to include particulars of the statistical reference sources, the procedures for stratifying the field of survey in accordance with the Community typology of holdings and the procedures for determining the selection rate chosen for each stratum. The FADN Committee is consulted for the purpose of verifying that the plans are in conformity with the provisions of Article 4 of the Council Regulation. The Commission considers that, taking into account differences in farm types and size classes in the Member States, this procedure is sufficient to ensure harmonisation of the selection plans.

19. The situation until 2002 in which Member States were responsible for the delivery of duly completed farm returns, but only the Commission had access to the data control programme, led sometimes to a lengthy and cumbersome process of asking for clarifications and corrections from the Member States. Therefore, the Commission started a modernisation process concerning data delivery and control. The new FADN-1 system has been operational since mid-2002 and was first applied to 2000 data. Its purpose is to improve the quality of the results and make them available quicker.

23. In the new FADN-1 data delivery and control system, Member States either correct the data or justify the error message. The Commission is in continuous contact with the Member States and has bilateral discussions concerning the errors, how to correct them and how to improve the overall quality of the FADN data.

24. An indicator of the organisational form of the holding will be available in 2004 for the accounting year 2002. This will allow the use of farm family income only for those holdings that are family farms. The other income indicators - farm net value added and farm net value added per work unit - can be used for all farms and are in fact more widely used in analyses than Farm Family Income.

27. Analyses based on the FADN data most often use farm net value added as an income indicator. This indicator can be used in a comparable way for all farms irrespective of their legal status.

30. A Regulation of the European Parliament and of the Council concerning the economic accounts for agriculture in the Community is in the process of adoption. This Regulation will consolidate the methodological framework of the EAA and will fix a framework for data transfer. It should thus contribute to improving the quality of the EAA.

34 to 36. The economic accounts for agriculture are the subject of continuous and detailed checks on their accuracy and comparability by the Working Party on "Agricultural accounts and prices" and during more specific exercises such as the inventories. This is why the quality of income statistics produced under the EAA has always been considered satisfactory.

In addition, if the choice of the sources of data is left to the Member States according to the specific structure of their agricultural sector and of their statistical system, this has little impact on the quality of the EAA data. It is in fact not the harmonisation of data sources but the harmonisation of definitions and principles of data extraction which guarantees the harmonisation of the EAA data among Member States and contributes to their quality.

Similarly, studies have shown that the multiplicity of data sources makes the accuracy indicators very difficult to establish and complex to interpret for such statistics as national accounts and accounts for agriculture.

48. Eurostat continues to collect the IAHS data available (without recourse to contractors).

49 to 50. The statistics for the FADN, the EAA and the IAHS have been drawn up for different reasons which go well beyond the simple calculation of income in some cases (the FADN is increasingly used for economic and financial analyses). If all the costs of the FADN seem to have been taken into account, the costs given for the statistics on agricultural accounts and farm households' incomes seem to reflect their real cost only very partially. The EAA and IAHS statistics are largely taken from other statistical sources (including national accounts and the FADN), so a precise estimate of the costs of these various statistics is difficult (a large part of the costs of the EAA and IAHS being borne by national accounts and the FADN).

THE PERTINENCE OF THE STATISTICAL TOOLS USED BY THE COMMISSION

57. From the beginnings of the CAP, the Commission has used the statistical tools provided by the FADN and EAA intensively for monitoring and evaluating the common agricultural policy with regard to its income goals. Since the most significant part of farm households' income comes from agricultural activity (and mainly commercial holdings), these statistics have also made it possible to monitor and analyse changes in farm households' income in a relatively reliable way. In fact, this was in line with paragraphs (a) and (b) of Article 33, which explicitly link the goal of ensuring a fair standard of living for the farming community to increases in income generated by agricultural activity.

64 to 66. For the reasons given in paragraph 57, the monitoring and evaluation of the common agricultural policy with regard to its income goals have always been ensured in a relatively reliable and relevant way by the statistical tools represented by the FADN and EAA. For example, at the time of the adoption of the last reform of the CAP in June 2003, all the key data as regards the evolution of incomes and the potential impact on incomes in the agricultural sector were based on indicators derived from the EAA and FADN.

67. The Commission has been using these different statistical tools on the one hand to observe the past and on the other hand as a basis for the reform proposals. Each of the tools provides additional information in the decision-making process. Therefore, they should be seen as complementary parts in a very complex policy field.

69. The FADN database was created to satisfy the needs of the common agricultural policy and was therefore designed to be able to monitor, inter alia, the changes in income of those most directly concerned with the CAP.

Figure 4 confirms the analysis of the Commission that farm households draw most of their income from agricultural activity and that, consequently, the EAA and the FADN can be used as relatively reliable estimators of farm households' income. According to Figure 4, the share of income generated by agricultural activity in the total income of agricultural households is considerably greater than 50 % in 12 of the 15 Member States. This share is less than 50 % only in Germany (partly for cyclical reasons), in Finland (where farmers also engage in major forestry activity) and in Sweden (for methodological reasons, households being classified by main occupation and not by income, and because of changes in the taxation of capital).

Finally, it should be recalled that while the broad definition of farm households can provide interesting quantitative information for economic purposes in quite specific situations, such statistics cannot be compared with data obtained under the narrow definition, which is the only one which conforms with the rules applied in the system of national accounts.

70. Although the FADN covers only 65 % of holdings, the income indicators drawn from it are broadly representative of the agriculture sector because the share of holdings covered by the FADN is much larger in terms of income. Since the structure of the agricultural sector is different from one Member State to another, the threshold of economic size has been set at different levels. Extending the coverage of the FADN to small holdings has proved very difficult and expensive, and contributes only a little additional information on incomes.

71. The Commission is continuously working to improve the quality of the results. The approach and methodology of the FADN are the same for all Member States, but country-specific factors can be taken into account. The question of representativity has been discussed in the FADN Management Committee. The Commission is working together with the Member States to improve the sampling and weighting methods of the FADN. These methods have to take into account the diversity of agriculture in the Member States.

72. The FADN may include, if appropriate, any woodland management and tourism connected with a farm. In 1994, the Member States rejected a proposal to include any other sources of income. The basic observation unit in the FADN is a holding, not a farm household. For business analyses, the second main objective of the FADN, "holding" is the only possible observation unit. In the 2000 census, the share of legal persons or group holdings in the FADN field of survey in EUR-15 was 6 %, up from just 3 % in 1990.

74. The main object of the EAA is the description and analysis of the production process and income generated by agricultural activity. As such, the EAA are not designed to measure the income drawn from other activities or various transfers.

75. The EAA are accounts by branch of activity, making it possible to assess and analyse the evolution of income across the agricultural sector. This information can be supplemented by the microeconomic data from the FADN which allows, inter alia, ananalysis by technical-economic type and by distribution.

77. The IAHS project has so far encountered difficulties in the Member States due to resources and methodological problems. The recent developments under the CAP and in the structure of the agricultural sector could bring new demand for such statistics and encourage further development.

CONCLUSION

78. The FADN and the EAA provide information on the income generated by agricultural activity which is used as basic information in monitoring and evaluating the CAP. These statistics are also necessary for developing the statistics provided by the IAHS, which can give additional information on the standard of living of the farming community.

79. The statistics of the EAA and the FADN have for many years provided very useful, reliable and relevant information on the incomes generated by agricultural activity. This information has proved very important for the monitoring, evaluation and design of the common agricultural policy. So far, these figures on income from agriculture could be regarded as relatively reliable estimates of farm households' income to the extent that the latter draw the major part of their income from farming. Equally, the use of these statistics is in line with paragraphs (a) and (b) of Article 33 explicitly linking the goal of ensuring a fair standard of living for the farming community to increases in the income generated by agricultural activity.

RECOMMENDATIONS

80. With the help of the relevant committees and working parties, the Commission is constantly evaluating and trying to improve the quality and coverage of the statistics of farm incomes, which are necessary for monitoring the objectives of the common agricultural policy.

81. The approach and methodology of the FADN are the same in all Member States, but country-specific factors can be taken into account. Several studies, aiming to contribute to a further harmonisation of the methods and an improvement of the results are currently foreseen. The Commission will develop the FADN in cooperation with the Member States in the FADN Committee. A further harmonisation between FADN and the farm structure survey is also planned.

82. The EAA will continue to be monitored for quality by the Working Party on "Agricultural accounts and prices". The improvement of the EAA (by systematically following up the gaps revealed by the inventories) will be actively pursued, especially since the EAA will henceforth be based on a Regulation of the European Parliament and of the Council.

Studies have shown that the multiplicity of data sources made accuracy indicators very difficult to establish and complex to interpret for such statistics as national accounts and accounts for agriculture. In the absence of progress in measuring the accuracy of economic accounts data, it seems impossible at present to take action on this recommendation.

83. The Commission will continue to evaluate the project recently started. On this basis, it will study the feasibility and cost of statistics making it possible to measure the standard of living of the farming community in all Member States. Likewise, it will decide on the methodological approach to be taken in collaboration with the Member States or the relevant committees.

52003SA0014