SUMMARY

I. Under the CAP, the European Union has two similar instruments specifically targeting children:

- the School Milk Scheme, through which grants have been available since 1977 for the sale of reduced-rate milk products in schools;

- the School Fruit Scheme, which has co-financed the distribution of fruit and vegetables in schools since the 2009/10 school year.

II. Both schemes use the same basic strategy (the provision of a product in schools) to pursue the same two objectives:

(a) to help stabilise the market (the "market" objective);

(b) to promote healthy eating (the "nutrition and health" objective).

For the Milk Scheme in particular, which was originally conceived of as a way of disposing of stocks, the Commission has gradually come to present better nutrition as the main objective.

III. Both schemes also aim to achieve a twofold impact:

(a) in the short term, to increase or maintain the consumption of these products by young people by making them available in schools;

(b) in the long term, to have an educational influence on eating habits.

IV. The Court has carried out a joint audit of the effectiveness of both schemes. Given the very recent nature of the Fruit Scheme, the audit focus was on school milk, supplemented by a comparative review of the system now being set up for fruit.

V. To a large extent, both schemes are now grounded in the hypothesis that there is a positive link between public health and consumption of these subsidised products. The audit did not set out to test that hypothesis, because the Court is not mandated to do so. However, the Court finds that the hypothesis is not universally accepted.

VI. Ten years after the Council’s decision to maintain the School Milk Scheme despite its being rated very negatively and despite a termination proposal by the Commission, the Court found that the scheme has not been significantly modified and, at best, still has very limited impact.

(a) Regarding the anticipated short-term impact (the provision of milk in schools), the Court finds that the scheme is achieving very little:

(i) Owing in particular to the low subsidy rate, the scheme continues to be relatively unattractive and, as a result, generally has no more than a deadweight effect. In most cases, the products subsidised either would have been included in canteen meals anyway or would probably have been bought by the beneficiaries even without the subsidy.

(ii) The audit showed that, while the decision by certain Member States to organise distribution free of charge has resulted in a more satisfactory impact, this form of distribution is at present covered by costly national schemes to which the Community budget makes only a marginal contribution.

(b) Regarding the anticipated long-term impact (education), the Court finds that at present the scheme takes insufficient account of the stated educational goals. In particular, distribution is not always made in a visible manner, and no other specific educational measures have been introduced.

VII. On the other hand, the Court did find that most of the weaknesses that were identified in respect of school milk had been noted and, at least in part, taken into consideration by the Commission in its planning of the Fruit Scheme. The main consequences were the adoption of a single model for distribution, which is made free of charge and outside the canteens, and the introduction of tools to facilitate the educational goals. However, these improvements come at a high cost, a significant part of which is borne by national and local budgets. While it is still too early to come to any definitive conclusions about the new scheme’s ultimate effectiveness, it does nonetheless appear considerably more likely to achieve its short- and long-term objectives. As such, it offers some indication as to how the Milk Scheme might be improved.

VIII. The Court has drawn up the following main recommendations:

(a) Given the very limited impact of the Milk Scheme, the matter of whether or not to retain it should be made conditional on the possibility of thorough reforms to remedy the weaknesses that have been identified. The decision should, in particular, take account of the product’s recognised nutritional value in the light of the public health objectives. If thorough reforms are to be undertaken, they should be based on the considerations set out below.

(b) If the Milk Scheme is to have a real impact on the volume of milk consumed in schools, the subsidy paid per kilogram should be increased very significantly to a level where milk can be distributed free of charge. However, this recommendation would not entail increased EU spending; on the contrary, it would mean focusing resources on a narrower target population, which could also effectively contribute to addressing the problem of deadweight. The population to be targeted would have to be determined in the light of a scale of nutritional needs.

(c) Where the Milk Scheme is concerned, steps should be taken to limit the specific deadweight effect associated with distribution through canteens, whilst taking care to ensure maximum visibility for the scheme.

(d) Where the educational objectives are concerned, the two schemes should adopt a more harmonised assessment of the role and importance of accompanying measures. If the importance of these measures is confirmed, the desirability of making them eligible for EU co-financing could be reconsidered.

(e) More generally, there should be greater coordination and synergy between the two schemes in order to ensure that they present a harmonised approach to nutrition and are managed efficiently.

INTRODUCTION

+++++ TIFF +++++

Source: http://ec.europa.eu/agriculture/tasty-bunch/index_en.htm

TWO SIMILAR NUTRITIONAL SCHEMES IN SCHOOLS

1. "Eat it, Drink it, Move it!" This was the slogan of the Commission’s 2009-10 publicity campaign to promote healthy eating by children in the EU. This campaign, which was funded from the EU’s agricultural budget, aimed to encourage children to adopt better eating habits. The "Tasty Bunch" campaign was launched by means of a travelling exhibition and was essentially channelled through the following two schemes for making food products available in schools:

(a) The School Milk Scheme [1] has made grants available since 1977 for the sale of reduced-rate milk products to schoolchildren.

(b) The School Fruit Scheme [2] was set up in time for the start of the 2009/10 school year. It offers co-financing towards the costs of distributing fruit and vegetables in schools.

2. The current situation is therefore that, since 2009, the Commission has had two similar instruments, one covering milk products and the other fruit and vegetables (see the comparative table of the two schemes in Annex I). They target the same beneficiaries and have broadly comparable budgets (see Graph 1).

3. Both schemes pursue the same two objectives:

(a) to help stabilise the market (the "market" objective) [3];

(b) to promote healthy eating (the "nutrition and health" objective) [4].

For the Milk Scheme in particular, which was originally intended as a measure to dispose of stocks, the Commission has gradually come to present better nutrition as the main objective (see Box 1).

GRAPH 1

EU SPENDING [5] ON THE SCHOOL MILK SCHEME AND SCHOOL FRUIT SCHEME (IN MILLION EURO)

+++++ TIFF +++++

Source: EU budget.

4. Based on this strategy, the two schemes target an impact at two levels:

(a) in the short term, to increase or maintain the consumption of these products by young people by making them available in schools;

(b) in the long term, to have an educational influence on eating habits [6].

+++++ TIFF +++++

+++++ TIFF +++++

Source: http://ec.europa.eu/agriculture/index_fr.htm

BOX 1

THE NUTRITIONAL AND EDUCATIONAL OBJECTIVES OF THE MILK SCHEME

In an internal note of 1990, the Commission said "The School Milk Scheme has not been designed as a means for the disposal of surplus milk […]. Instead it is a true educational measure that is primarily intended to maintain or promote in schools the consumption of milk and milk products. The scheme also has a nutritional aspect in that the milk products distributed could supplement an insufficient or unbalanced diet. It undeniably has the secondary effect of ensuring the consumption of a significant volume of milk products which would otherwise be added to the milk surplus and disposed of under other measures."

Source: The Commission, 1990.

"The School Milk Scheme has both nutritional and educational character and contributes to the fight against obesity and to providing essential elements for children’s growth and health."

Source: Press release by Commissioner Fischer Boel, 2008.

OVERVIEW OF THE TWO SCHEMES’ ADMINISTRATIVE ARRANGEMENTS

5. Both forms of support are paid directly to schools or to other applicants, which may be education authorities (generally a local council), suppliers or intermediary bodies set up specifically for that purpose [7]. Once approved, applicants have operational responsibility for the supply procedure (purchase, storage, distribution and administration).

MILK PRODUCTS

6. Aid comes in the form of a fixed-rate EU subsidy of 18,15 euro for 100 kg of milk [8], with no obligation for Member States to contribute.

7. Each establishment decides whether or not to participate, and the EU contribution is calculated on the sum of the eligible applications that are received. No budgetary ceiling has been defined by the regulation. However, the latter provides for a maximum subsidisable quantity of 0,25 litre per school day and per pupil in regular attendance at the establishment [9], a ceiling that is not usually reached by the applicants.

GRAPH 2

SCHOOL MILK — RELATIVE SHARE OF THE DIFFERENT MODELS IN THE MEMBER STATES VISITED

+++++ TIFF +++++

Source: Member States visited.

8. Although the scheme may take any number of different forms, three models in particular were encountered in the Member States that were visited (see Graph 2 and Box 2):

(a) subsidies for milk products included in canteen meals [10] (mainly France, Italy and Sweden);

(b) milk sold at a reduced price outside canteens (Germany, Poland (secondary schools) and the United Kingdom (children above 5 years of age));

(c) milk distributed free of charge outside canteens [11] (mainly Poland (primary) and the United Kingdom (children up to 5 years of age)).

FRUIT AND VEGETABLES

9. EU aid accounts for between 50 % and 75 % [12] of the cost of the fruit and vegetables distributed, including certain associated charges, with the remainder usually being financed by the Member States.

BOX 2

TYPICAL EXAMPLES OF MILK PRODUCTS THAT ARE SUBSIDISED IN THE MEMBER STATES VISITED

France: portions of cheese or yoghurt as a dessert (canteens)

Italy: parmesan on a serving of pasta or mozzarella in a salad (canteens)

Sweden: self-service milk fountains during meals (canteens)

United Kingdom and Poland: cartons of milk sold at a reduced rate or distributed free of charge to younger children, mainly with help from the national budget

Germany: cartons of flavoured milk sold at a reduced rate in school shops

10. The regulation sets an overall budget ceiling of 90 million euro, shared among the participating Member States. Each Member State is free to decide, through its national strategy, how to divide its national allocation among potential applicants.

11. Support for the distribution of fruit is conditional on the mandatory adoption of accompanying educational measures to ensure that the scheme is a success. Although ineligible for EU funding, these measures are considered a crucial part of the scheme.

12. All of the fruit that the auditors saw being distributed was free of charge and supplied outside mealtimes.

AUDIT SCOPE AND APPROACH

A COMBINED AUDIT OF THE EFFECTIVENESS OF BOTH SCHEMES …

13. The audit was mainly concerned with the effectiveness of the School Milk and Fruit Schemes. It was felt that the launch of the Fruit Scheme in 2009, using a model similar to that already in place for the Milk Scheme, was the right time for such an analysis. Because of this timing, the audit largely focused on the Milk Scheme, supplemented by a comparative review of the system now being set up for fruit.

14. The audit questions asked in respect of the schemes’ effectiveness were:

(a) Is participation sufficiently encouraged (assessment of any disincentives to participate) and is it enough for the objectives to be achieved?

(b) Does spending have a direct impact on the beneficiaries’ consumption of the products distributed? (Deadweight effect: would the products be consumed if no aid were available?)

(c) Are the schemes likely to meet their educational objectives and influence future eating habits?

GRAPH 3

SCHOOL MILK — 2009 PAYMENTS BY MEMBER STATE

+++++ TIFF +++++

Source: European Commission.

15. To a large extent, both schemes are now grounded in the hypothesis that there is a positive link between public health and consumption of these subsidised products [13]. The audit did not set out to test that hypothesis, because the Court is not mandated to do so. However, the Court finds that the hypothesis is not universally accepted.

16. The audit looked at the 2008/09 and 2009/10 school years. It also examined the preparations that were made for the two regulations currently in force and, for the Milk Scheme, assessed the most recent evaluation results.

17. The audit was carried out at the Commission and in Germany [14], France, Italy, Poland, Sweden and the United Kingdom. These six Member States (the last two were participating in the Milk Scheme only) accounted for 75 % of Milk Scheme payments in 2009 and 63 % of the budget estimate for the first year of the Fruit Scheme (see Graphs 3 and 4).

GRAPH 4

SCHOOL FRUIT — 2009/10 BUDGET BY MEMBER STATE

+++++ TIFF +++++

Source: European Commission.

18. The auditors examined the procedures set up by the Commission and the Member States for implementing and monitoring the schemes. They also visited 31 applicants receiving aid on behalf of over 40000 educational establishments and some 2 million children. Further visits were made to 56 establishments (attended by over 16000 children) participating in at least one of the schemes.

… AND A REGULARITY REVIEW

19. The audit also included a regularity review of key aspects that were most likely to have a direct effect on beneficiaries and policy effectiveness. Problems — amongst other things with regard to the interpretation of the rules in force — were detected in the various Member States that were visited, and were raised in a communication to the Commission and the Member States concerned. For further information, these aspects are summarised in Annex II.

+++++ TIFF +++++

Source: http://agriculture.gouv.fr/un-fruit-pour-la-recre

OBSERVATIONS

THE MILK SCHEME WAS SEVERELY CRITICISED IN A 1999 EVALUATION, YET NO REAL CHANGES HAVE BEEN MADE SINCE

20. In 1999 the School Milk Scheme was the subject at Community level of an external evaluation [15]. The evaluation was extremely critical of the scheme and it recommended termination of it because of its low effectiveness and poor cost–benefit ratio (see Box 3).

21. The Commission reacted by proposing that the scheme be wound up — a proposal that was rejected by the Council. In response to this political decision, the Commission then submitted new proposals with a view to reducing the Community contribution to the scheme. These were only partially accepted (see Box 4).

BOX 3

1999 EVALUATION — CONCLUSION AND FINAL RECOMMENDATION

"[…] Judged purely against the current, stated documented objectives of the measure (maintaining and increasing consumption of milk products), the measure has had a marginal positive impact and represents poor value for money. This suggests that the Commission should give serious consideration to withdrawing the measure. The main implications of such action would be to place the onus for continuing to provide any form of subsidised milk to schoolchildren on Member State governments. Whilst it is impossible to predict the outcome of such an action it is likely that the net effect of such action would undoubtedly lead to decreased availability of milk products and decreased consumption in schools. However, the evidence presented in this report suggests that the impact would probably be very limited.

[…] As the main recommendations given above focus on discontinuation of the measure and use of resources for alternative, more effective measures, no further recommendations relating to improving the efficiency of the existing measure are made. The authors consider that the focus of any change should seek to address the fundamental weaknesses of the measure in achieving its objectives. Whilst making recommendations for improving the efficiency of the existing measure could be offered, these would not significantly address the issues contributing to the very limited effectiveness of the measure."

22. In 2005, another evaluation was made in the UK [16], relating to both the EU support and the top-up from the national budget (see paragraph 8(b)). Its conclusions describe the measure as ineffective, expensive and characterised, in particular, by disproportionate management costs and low added value. The evaluation recommends that the UK government should "consider discussing with the European Commission the idea of ending the EU SMSS".

23. For its part, the Commission has regularly sought since 1999 to improve the scheme by reworking its implementing rules (simplifying them, amending the list of products and the aid rates by category, etc.). However, these developments have remained "marginal" — i.e. they have not affected the logic of the scheme or its main features. Although generally acknowledged to be positive, therefore, the amendments made have in no way responded to the fundamental criticisms concerning the scheme’s basic conception.

BOX 4

THE COMMISSION’S PROPOSALS IN RESPONSE TO THE 1999 EVALUATION

"Drawing on this recommendation, the Commission initially intended to propose the suppression of the scheme and presented its preliminary draft budget for 2000 based on this assumption […] all participants pleading in favour of maintaining the EU subsidy notably because of its nutritional merits […] In the light of this discussion, the Commission reconsidered its initial intention and is now proposing a co-financing of the school milk programme, on a 50/50 basis. […] In view of the tight budgetary framework and of the low cost/efficiency of the measure, the budget line for school milk cannot be maintained as its current level" […] "The school milk programme is very often defended on the basis of broader health, nutritional or even social objectives." While the relevance of this argument is fully recognised, "[…] these objectives are too ambitious for a scheme that is part of the CMO for milk and dairy products. Again, […] when Member States invoke such objectives, they should be willing to share the responsibility for expenditure under the scheme."

Source: Note from DG VI to the office of the Commissioner responsible for agriculture and rural development, March 2000.

ANTICIPATED IMPACT LIMITED BY MODEST SPENDING AND NO MECHANISM FOR TARGETING PRIORITY NEEDS

NO SATISFACTORY PERFORMANCE ASSESSMENT SYSTEMS

24. The schemes’ two objectives – "market" and "nutrition and health" – are defined very broadly and in qualitative terms (see Box 5). They have not been expressed quantitatively, still less as "SMART" objectives [17].

25. Where the Milk Scheme is concerned, in the absence of precise objectives against which to assess the policy’s effectiveness, the available indicators (quantities distributed by category, approximate number of participating pupils) are activity indicators that give no measure of policy outcomes or impact [18]. The audit also highlighted the unreliability of the information gathered, especially the figures for the approximate number of participating pupils [19].

BOX 5

THE GENERAL NATURE OF THE MILK SCHEME OBJECTIVES

"To contribute to balancing the milk market and to stabilise […] prices" (recital 43 of Council Regulation (EC) No 1234/2007)

"… play a role within broader general and health, nutritional or social policy objectives" (explanatory memorandum on the proposal for a Council regulation amending Regulation (EC) No 1255/1999)

"It also plays an educational role by supporting the development of good eating and nutritional habits that will last a lifetime" (Agriculture and Rural Development DG website)

"The School Milk Scheme has both nutritional and educational character and contributes to the fight against obesity" (press release by Commissioner Fischer Boel, 11 July 2008)

26. No indicators have been set for the Fruit Scheme. However, the scheme does introduce a number of provisions that are conducive to the monitoring of performance, from preparation of an initial strategy to the obligation for periodic evaluations in each Member State. However, it is still too early to assess the effectiveness of these measures, which are to be further developed.

VERY MODEST SPENDING ON THE SCHEMES’ OBJECTIVES

"MARKET" OBJECTIVE

27. During talks prior to setting up the Fruit Scheme, the European Parliament had proposed a budget of 500 million euro per annum. In the end, at the Commission’s suggestion, a ceiling of 90 million euro was agreed. The Commission felt in particular that the project had to prove itself on a smaller scale before a larger budget could be considered.

28. The Commission’s approach was justified in hindsight by the very hesitant start to the scheme (see Annex IV). Nonetheless, the effect of this decision, even when national co-financing is factored in [20], was that the budget ceiling set in the regulation equates to around 0,3 % of the production value of fruit and vegetables in the EU. A similar figure is obtained for the Milk Scheme, which, to judge from the available figures, helped to fund the distribution of a little less than 0,28 % of total EU production in 2008/09.

29. In both cases, even if these amounts were spent effectively, the volumes to which they correspond are not likely to have a significant direct impact [21] on market equilibrium.

30. Where the Milk Scheme is concerned, this finding was already part of the 1999 evaluation [22].

"NUTRITION AND HEALTH" OBJECTIVE

31. When divided by the number of children in the European Union, the expenditure that is currently budgeted for each scheme gives between 50 and 80 cents per child per year. On this scale, it is also hard to imagine it having a significant influence on the eating habits of the target population as a whole.

32. In practice, as spending benefits fewer children [23] the impact per beneficiary is potentially greater. However, the schemes’ overall effectiveness in terms of the nutrition and health objective depends very much on the sub-population that is "chosen" to receive support.

NO MECHANISM FOR TARGETING NUTRITIONAL PRIORITIES

33. Supplying milk to a child who ordinarily consumes far less than his peers, or than recommended amounts [24], adds more value than supplying the same quantity of milk to a child whose consumption of milk products is already above average or more than recommended amounts. For both schemes, therefore, the nutrition and health objective means prioritising distribution in areas (by region, Member State, social category, etc.) where consumption is deemed to be insufficient or below average. The key aim is not so much to increase average consumption as to enable as many children as possible to attain a satisfactory minimum level of consumption.

34. Where milk is concerned, some Member States have already shown awareness of this situation. In the United Kingdom, the 2005 evaluation found that, to a large extent, the target population already consumed sufficient quantities of milk products [25], and it recommended focusing on a number of priority categories. In France, the authorities have chosen to allocate the national budget complement to less-favoured areas, which are deemed to represent the greatest nutritional need [26]. Such targeting could take many other forms, although some of them may involve additional costs.

35. However, neither of the EU schemes provides for resources to be allocated on the basis of a nutritional-needs analysis.

36. The Fruit Regulation leaves it to the Member States’ discretion whether or not to incorporate this approach into their national strategies. However, no state has yet taken up this option [27].

37. There is no such mechanism under the Milk Scheme. An analysis by Member State not only reveals that the scheme does not target resources at areas of greatest need, it even shows that, on the whole, the Member States receiving most from the scheme are those where the consumption of milk products is already highest and, consequently, needs are less significant and less pressing (see Box 6).

BOX 6

COMPARISON BETWEEN MEMBER STATES OF THE CONSUMPTION OF MILK PRODUCTS AND AID RECEIVED IN 2009 UNDER THE MILK SCHEME

The two Member States with the highest per capita consumption of milk products are also the two main per capita beneficiaries of aid.

On average, the six Member States with the highest per capita consumption of milk products receive 3,5 times more Community aid per capita than the 10 states with the lowest consumption rates.

The six Member States receiving the most aid per capita already have an above-average per capita consumption of milk products.

The 10 Member States with the lowest per capita consumption of milk products receive just 20 % of all aid. Nine of them receive two to five times less per capita than the average or are not even participating in the scheme.

DESIGN PROBLEMS RESTRICT WHAT THE MILK SCHEME CAN ACTUALLY ACHIEVE. HOWEVER, SOLUTIONS EXIST.

THE MILK SCHEME IS RELATIVELY UNATTRACTIVE

38. The level of spending on school milk is not set directly by the Commission but is the aggregate of individual choices by applicants and potential beneficiaries. The 74 million euro spent in 2009 accounts for less than 10 % of the potential maximum for expenditure [28] under the scheme. Thus the take-up rate is low and the policy relatively unattractive (see Boxes 7 and 8), to a similar extent to what was found in the past [29].

39. The audit revealed that this unattractiveness was largely the result of the low level of aid combined with the disproportionate burden of administration.

MILK SOLD AT A REDUCED RATE OUTSIDE CANTEENS

40. At current aid levels, the EU subsidy is only sufficient to cover a small fraction of the cost of the products distributed [30]. In the absence of other, much more substantial, sources of funding (see paragraph 60 et seq.), therefore, the EU scheme does not permit distribution free of charge but extends only to the sale of milk at a reduced rate.

BOX 7

THE LOW PARTICIPATION RATE AND THE SCHEME’S UNATTRACTIVENESS

A study made in Germany shows that, of the 56 % of establishments which offer milk or milk products to their pupils, just 21 % have applied for aid.

Source: Bavarian Ministry of Agriculture.

41. In Poland, EU aid covers just one quarter of the direct cost of milk. In Germany and the UK, it covers just 10 % to 15 %. In Germany, for example, it allows cartons of flavoured milk to be sold for 35 rather than 40 cents. Reductions on this scale are generally too small to trigger a purchase decision, all the more so as studies have revealed a low price elasticity of demand [31].

42. In many cases the establishment itself chooses not to participate. Usually goodwill is needed if the scheme is to be implemented at a school. However, organising milk to be sold in schools at a reduced rate often necessitates a significant amount of work (purchasing, storage, distribution, collecting money from recipients, etc.), which can be a real obstacle to participation [32].

43. In some cases, the readiness of decision-makers to accept the workload associated with the scheme is reduced by the fact that families show little interest in purchasing subsidised milk or, on occasion, that the relevant staff are not themselves convinced of the scheme’s health benefits.

MILK PRODUCTS INCLUDED IN CANTEEN MEALS

44. Where the subsidy is paid for products used in canteen meals, management of the scheme is absorbed into regular canteen management duties, causing most of the administrative costs to disappear. What this finding means, without prejudice to any negative assessment of the scheme’s effectiveness in this form (see paragraphs 47 to 50), is that the only specific cost is the time requirement to apply for a grant. Paradoxically, even this solitary administrative burden continues to be unwelcome and may prove prohibitive.

45. What makes the administrative procedure so excessive for most applicants — especially the smallest — is not so much its complexity as the fact that so little aid is received (see Box 8). The absolute value of aid is perceived as being even more insignificant because it is compared not with the cost of milk products but with the overall budget — of which it is only a small part (under 1 % for canteens, and far less still when compared with the total budget of an independent private school or a nursery).

THE IMMEDIATE IMPACT OF SPENDING UNDER THE MILK SCHEME IS COMPROMISED BY A SIGNIFICANT DEADWEIGHT EFFECT [33]

46. Whatever form distribution takes, the milk subsidy is available for all eligible products, and there is no mechanism for restricting payment for quantities which the beneficiaries would anyway have purchased without aid. The audit found that the deadweight effect is different according to the distribution model concerned.

BOX 8

THE UNATTRACTIVENESS OF AID FOR CANTEENS

In France, some 70 % of the 11000 applicants receive on average around 300 euro per year.

Many of the persons interviewed stated that aid on this scale was not enough to get them to change their habits. For example, one school Head said: "All I get is 200 euro a year and I have no intention of wasting my time reading a circular for such a small return."

Although any canteen uses eligible milk products for which it could receive the minimum amount in aid simply by applying for it, only 60 % of establishments in France, and barely 15 % in Italy, submit an application.

MILK PRODUCTS INCLUDED IN CANTEEN MEALS

47. The audit showed that the payment of aid for products included in canteen meals systematically generates deadweight. In particular:

(a) although a typical meal often includes milk products (they may even be compulsory), there is at present no way of preventing canteens from receiving payment for products which they would have served anyway;

(b) no evidence was found that aid has any influence whatsoever on the volume of milk products included in menus;

(c) virtually all the persons responsible who were interviewed stated explicitly that aid in no way influenced the volume of milk products that they served to pupils (see Box 9).

BOX 9

STATEMENTS BY CANTEEN MANAGERS ABOUT THE MILK SCHEME

"The composition of our menus is not influenced by EU aid. […] That isn’t an opinion, it’s a fact" (Italy, manager of the catering company running the canteens in one of the country’s biggest cities)

"I have never thought of the EU Milk Scheme when setting lunch menus" (Sweden, representative of a catering company receiving aid for several schools in the same town)

"The existence of EU aid in no way influences our menus" (Italy, representative of an establishment receiving aid direct)

In France, the managers of a nursery awaiting authorisation explained that they were interested in obtaining the money to which their usual practice entitled them — namely for products they were in the habit of buying before learning of the scheme.

The person responsible for canteen menus at another applicant was not even aware that his establishment had been receiving aid for years.

48. In short, canteens are receiving aid for milk products that they would be serving anyway and the aid does not encourage them to increase their use of milk products. Even in Sweden, where the situation is apparently most positive [34], the subsidised scheme (chiefly in the form of self-service milk fountains in canteens) existed well before EU support became available. The high rate of milk consumption is a long-standing cultural habit that cannot be ascribed to the EU scheme, which does no more than pay a contribution towards practices that it has not influenced.

49. At present, therefore, there is no easily perceivable impact of aid payments to canteens. To judge from the on-the-spot visits, there is at best a tiny impact on the beneficiaries’ consumption of milk products. Despite the provisions in the regulation, moreover, the scheme’s potential impact on the price paid by families is often unverifiable (see Annex II), and in any case it would be insignificant.

50. This finding also applies to nurseries, especially as, the younger the children, the greater the probability that their normal diet will include milk products. For example, in the French system of crèches (for children up to 3 years of age) milk products are considered a compulsory component of the food served.

MILK SOLD AT A REDUCED RATE OUTSIDE CANTEENS

51. Given the current low level of aid, it appears to have very little positive impact. The small reduction in the sale price which is attributable to the EU scheme (see paragraphs 40 and 41) not only does little to encourage participation, it also mainly benefits persons who would not have chosen any differently even without aid. As the aid is insufficient to influence purchase decisions in any real way [35] (see also Box 10), most pupils benefiting from the aid would also have been most likely to buy the milk even if unsubsidised.

52. What is more, these beneficiaries are also, by definition, most likely to consume milk products outside school. Just as the analysis by Member State shows that aid tends to go where it is least needed (see paragraph 37), such small reductions on the sale price will mainly benefit children who are already likely to be consuming the greatest quantities of milk products.

BOX 10

EXCERPTS FROM THE EVALUATION REPORTS DEMONSTRATING THE LIMITED IMPACT OF AID ON THE VOLUMES CONSUMED BY BENEFICIARIES

United Kingdom evaluation, 2005: "Our survey of schoolchildren found only weak evidence that milk consumed in school was an addition to milk consumed elsewhere […]. Further, there were children who drank very little milk in both schools that did and did not participate in the SMSS. […] Our survey found weak evidence that schools’ participation in the EU School Milk Subsidy Scheme affects children’s total milk consumption. […] In both participating and non-participating schools, a minority of pupils drank very little milk. The available evidence suggests that the gains in milk consumption due to schools’ participation in the SMSS occurred particularly among children who would have consumed a considerable amount of milk had their school not participated."

"[The] experiment with providing children with high-quality food showed that many parents who were unwilling to buy such foods in supermarkets were unwilling to pay for them in schools either. The same problem applies to milk that schools make available for parents to buy under the Top-Up."

1999 evaluation: "In terms of impact of the scheme on milk product availability there is very limited evidence of impact. […] Whilst factors external to the scheme are the primary factors of influence for take up, the scheme itself contributes in a very limited way to increasing product availability."

SOLUTIONS HAVE ALREADY BEEN INTRODUCED BY THE COMMISSION (FRUIT) AND SOME MEMBER STATES (MILK)

POSSIBLE RESPONSES ALREADY BUILT INTO THE COMMISSION’S DESIGN FOR THE FRUIT SCHEME

53. When planning the Fruit Scheme, the Commission took into consideration, at least in part, the unattractiveness of the Milk Scheme and the risk of a deadweight effect.

ORGANISING THE DISTRIBUTION OF FRUIT OUTSIDE CANTEENS

54. As the Commission had identified the distribution of subsidised fruit at the same time as regular school meals as a risk to added value [36], this mode of distribution was made subject to such restrictions that in practice it was excluded. As a result, not one Member State included this option in its strategy, thus all of the fruit that was observed being distributed was supplied outside canteens.

A HIGHER LEVEL OF AID AND DISTRIBUTION FREE OF CHARGE

55. The level of EU co-financing for the Fruit Scheme varies, according to Member State, from 50 % to 75 %. It is therefore far higher than for school milk.

56. On this basis, distribution free of charge has become the only model employed for the Fruit Scheme (and the Fruit Regulation often uses the term "distribution" where the Milk Regulation speaks of "supply").

57. Thus all fruit distribution schemes are free of charge to pupils, so that there is almost universal participation by those who are offered free fruit. One effect of this is to minimise the relative deadweight, because the beneficiaries include children who would not have been prepared to purchase the product without a subsidy. Another effect is that the role of schools is enormously simplified by their no longer having to collect money from families.

58. In most Member States the distribution of fruit is also free of charge to schools, as the EU co-financing is complemented by a national contribution paid directly to suppliers. Schools receive deliveries free of charge and without having to pay cash in advance. In Italy, the practice of "all-inclusive" tendering (including the physical distribution of fruit in classrooms and the organisation of accompanying measures) has even made it possible to remove all administrative duties from schools. The only remaining limit on schools’ participation is then the size of the available budget.

59. In France, however, where there is no national contribution, the participation of schools depends on their ability to raise finance locally. As this condition has very rarely been met in practice, so far the number of applicants is extremely low (see Annex IV).

SOME MEMBER STATES NOW ALSO DISTRIBUTE MILK

60. In France, although the bulk of EU support is still paid to canteens, the authorities have opted to allocate the national contribution to free distribution schemes that do not involve canteens. This was explicitly justified by the wish for the national complement to have a real impact, which is not felt to be possible in the case of canteen subsidies.

61. In Poland and the United Kingdom, EU support is used in association with national budgets, making it possible to replace subsidised sale schemes by the distribution of milk free of charge outside canteens to an even greater number of beneficiaries (primary-school pupils in Poland, children of up to 5 years of age in the UK).

62. This model is far more attractive than that of subsidising sales, and it ensures far greater participation (see Box 11). In so doing, it considerably reduces the problem of deadweight that attaches to the sale of milk at reduced prices, as all children can be reached rather than just those who would have been prepared to pay anyway.

BOX 11

THE IMPACT OF DISTRIBUTION FREE OF CHARGE

A UK study has shown that pupil participation declines sharply as soon as distribution stops being free of charge (source: the largest applicant in the UK for primary and secondary schools). The audit came across one establishment where participation in the Nursery Milk Scheme (or NMS, a free nationwide scheme for children up to 5 years of age) was extremely high but not one pupil was prepared to take part after the cut-off age. Other places refused to run a subsidised distribution scheme and limited their participation to those age groups that were eligible for distribution free of charge.

In Poland, following the switch in 2007/08 to a system of distribution free of charge in primary schools, the number of participating schools more than doubled and the number of participating children rose more than fourfold.

A German study shows that even substantial differences in sale price (30/35 cents rather than 15 cents) have little impact on participation. It is only when distribution becomes free of charge that there is a real change in impact (see Graph 5).

GRAPH 5

RELATIONSHIP BETWEEN SALES PRICE AND PUPIL TAKE-UP IN PARTICIPATING SCHOOLS

+++++ TIFF +++++

Source: Government of North Rhine-Westphalia.

63. In both Poland and the United Kingdom, however, distribution free of charge depends on a very large national effort (see Graph 6), with the EU subsidy covering a very small fraction of the total cost [37]. Given current EU aid levels, therefore, distribution free of charge cannot be directly attributed to the EU scheme [38]. However, the situation in these two Member States demonstrates just how much funding is necessary in order to set up such distribution schemes.

GRAPH 6

NATIONAL BUDGETARY EFFORT BY THE TWO MAIN MEMBER STATES DISTRIBUTING MILK FREE OF CHARGE (IN MILLION EURO; 2008/09 ACADEMIC YEAR)

+++++ TIFF +++++

Source: MS budget data.

OTHER GOOD PRACTICES IDENTIFIED DURING THE AUDIT WHICH HELP TO IMPROVE PARTICIPATION AND EFFECTIVENESS

64. The audit also highlighted a number of local factors and decisions which made it possible to reduce administrative costs and were conducive to improved participation and effectiveness. In particular:

(a) making use of any possibility in the regulation for simplifying the completion and verification of aid applications (in particular by employing flat-rate coefficients in calculations);

(b) concentrating on applicants over a critical size (reduces the proportion of administrative costs and may even allow significant economies of scale);

(c) engaging private-sector partners to take on some or all administrative and logistical tasks [39] while gaining a vested interest in improving participation.

EDUCATIONAL GOALS TO BE CONSOLIDATED

65. Over and above the immediate impact of distribution, both schemes have the same educational goal of influencing eating habits in the long term. However, this goal is addressed rather differently in the two schemes, and it still needs to be consolidated.

EDUCATIONAL GOALS NOT WELL TAKEN UP BY THE MILK SCHEME, THUS LITTLE LIKELIHOOD OF A SIGNIFICANT LONG-TERM IMPACT

THE EDUCATIONAL IMPACT OF SUBSIDISED DISTRIBUTION DEPENDS ON VISIBILITY, WHICH IS STILL NOT GUARANTEED

66. In principle, the very existence of a subsidised distribution scheme may help to convey a message about the value of the product in question. For this to happen, however, the scheme must be visible.

67. Most of the schemes that were observed for the distribution of milk outside canteens met this criterion. It was also satisfied in Sweden by the use of milk fountains in canteens. In certain other cases, however, neither the subsidy nor the product distributed was sufficiently visible.

68. In Germany, milk for sale in school shops is not specifically promoted and suffers from the proximity of other items (fizzy drinks, crisps and sweets). It does not have the clear visibility that comes from single-product distribution.

69. In much the same way, where subsidised milk products are included in canteen meals, distribution as such is no longer the right term and the educational impact declines. Having explicitly identified this weakness, the Commission has sought to restrict the delivery of aid through canteens [40]. While this is a positive development, it is not sufficient to guarantee that the educational goals of aid are achieved in practice.

Menu including buttered pasta (Parmesan sauce) subsidised under the Milk Scheme

+++++ TIFF +++++

Source: Italian canteen.

70. Firstly, the product is not presented for what it is. Instead, by definition, it is emphasised less than when specifically distributed [41]. Secondly, beneficiaries are generally not made aware that the scheme exists [42] or, hence, that the product has been assigned a particular status. In France and Italy, there were even some school Heads who were unaware that their canteens were being subsidised.

LACK OF ACCOMPANYING EDUCATIONAL MEASURES

71. Even where the visibility criterion is met, distribution is not felt to be enough, by itself, to achieve the educational goal in a satisfactory manner.

72. The 1999 evaluation had already emphasised the educational shortcomings of a scheme which did no more than offer price reductions, to the exclusion of promotional and educational measures [43]. The 2005 UK evaluation picked up on the same point [44]. Many of those interviewed during the audit confirmed this analysis, pointing out in particular that the main reason why milk is not more widely consumed (at least as a drink) by adolescents lies in an image problem, something of which no account is taken in the way the scheme is designed.

73. In spite of these findings, Regulation (EC) No 657/2008 is concerned only with the availability of milk products and does not require or encourage the use of accompanying measures. Against this background, not one of the Member States visited had imposed an obligation of this sort, and none of the establishments visited had introduced or was taking advantage of specific accompanying measures.

74. There are undeniably promotional campaigns in addition to the School Milk Scheme. As well as the "Tasty Bunch" [45], the Commission co-finances promotion measures under Council Regulation (EC) No 3/2008 [46], a few of which relate to milk products and target children [47]. Other examples found in various Member States undeniably have a positive impact. However, as they are neither required nor encouraged by Regulation (EC) No 657/2008, measures of this sort are still largely independent of the Milk Scheme and cannot be attributed to its influence.

NO SIGN OF A LONG-TERM IMPACT ON EATING HABITS

75. In the absence of specific tools, even after several decades of the scheme there is still no evidence that it has had a long-term impact on its beneficiaries’ eating habits. To quote from the 2005 UK evaluation: "it is not clear that school milk programmes inculcate a lasting habit of milk consumption".

76. On the contrary, in those Member States where data are available it can be seen, all else being equal, that participation in such programmes drops steeply with age. Use of the scheme in primary schools does not have sufficient impact in the medium term to ensure that participation remains satisfactory in the following years. It is all the more probable, given the scheme’s inability to retain participants for a moderate time span, that it will not be likely to have a significant long-term influence on their nutritional habits.

+++++ TIFF +++++

Source: www.drinkitup.europa.eu

BETTER TAKE-UP OF EDUCATIONAL GOALS UNDER THE FRUIT SCHEME

77. While it is still too early to comment on the long-term effectiveness of the Fruit Scheme, its approach does appear to be more consistent with the educational goals.

78. Firstly, all the cases of fruit distribution that were observed during the audit were highly visible, making them all the more likely to have an educational impact. This visibility was partly due to enthusiasm about the novelty of the scheme, but it was also guaranteed by the fact that distribution was specifically organised outside canteens (see paragraph 54).

79. Secondly, unlike the Milk Scheme, from the very start the Fruit Scheme, thanks to a specific tool, incorporated the requisite educational dimension. The Fruit Regulation states that "School Fruit Schemes require accompanying measures in order to be effective" [48] and, although it does not specify funding for them, it makes aid conditional on the existence of these measures.

BOX 12

EXAMPLES OF ACCOMPANYING MEASURES EXAMINED DURING THE AUDIT

Travelling exhibition to educate pupils about fruit and vegetables

Interactive information booth in a school entrance hall

Discussions organised for parents with doctors and nutritionists

Vegetable garden on school premises

Educational support material for teachers

80. For the same reason, synergies are encouraged with promotion measures organised under Regulation (EC) No 3/2008, by offering an opportunity for additional co-financing [49] that is not available for milk.

81. Nevertheless, the audit showed that, given what it may cost to set up accompanying measures, the absence of specific finance and the difficulty of determining exactly what constitutes a satisfactory accompanying measure result in arrangements that vary in their ambitiousness (see Box 13). It therefore remains to be seen whether the educational programmes currently being set up will prove to be effective.

BOX 13

THE WIDE RANGE OF ACCOMPANYING MEASURES IN PLACE AT THE TIME OF THE AUDIT

In Italy, under an ambitious system introduced during the first year, accompanying measures are provided regionally by the producers’ organisations which are responsible for distribution.

In Germany and France, the establishments visited had organised a number of accompanying educational activities on the themes of fruit and vegetables and food in general. However, given in particular that there were no control arrangements, these activities were still dependent, at that stage, on goodwill at local level.

In Poland, there were still no accompanying educational activities, or plans for such activities, at the establishments that were visited (under the transitional provisions in the regulation, such measures are optional during the scheme’s first year, but the Polish authorities had unilaterally undertaken to implement measures by this stage).

CONCLUSIONS AND RECOMMENDATIONS

82. In 1999, despite the very negative conclusions reached by an external evaluation, the Council took a decision, contrary to the Commission’s initial recommendations, to maintain the School Milk Scheme (see paragraphs 20 to 23). Ten years on, a few improvements notwithstanding, the Court has found that no real solutions have been applied to overcome the fundamental weaknesses that were identified at the time.

83. The Milk Scheme is still largely ineffective today.

(a) Regarding the anticipated short-term impact (the provision of milk in schools), the Court finds that the scheme is achieving very little of note:

(i) Owing in particular to the low subsidy rate, the scheme continues to be relatively unattractive and, as a result, generally has no more than a deadweight effect. In most cases, the subsidised products either would have been included in canteen meals anyway or would probably have been bought by the beneficiaries even without the subsidy (paragraphs 38 to 52).

(ii) The audit showed that, while the decision by certain Member States to organise distribution free of charge has resulted in a more satisfactory impact, this form of distribution is at present covered by costly national schemes to which the Community budget makes only a marginal contribution (paragraphs 60 to 63).

(b) Regarding the anticipated long-term impact (education), the Court finds that at present the scheme takes insufficient account of the stated educational goals. In particular, distribution is not always made in a visible way, and no other specific educational tool has been introduced (paragraphs 66 to 76).

84. On the other hand, the Court does find that most of the weaknesses identified in respect of school milk were noted and, at least in part, taken into consideration by the Commission in its planning of the Fruit Scheme. In particular, the new scheme uses a single model for distribution, which is made free of charge and outside canteens (paragraphs 53 to 59), and specific tools were introduced to facilitate the educational goals (paragraphs 77 to 81). However, these improvements come at a high cost, a significant part of which is borne by national and local budgets. While it is still too early to come to any definitive conclusions about the new scheme’s ultimate effectiveness, it does nonetheless appear considerably more likely to achieve its short- and long-term objectives. As such, it offers some indication as to how the Milk Scheme might be improved.

85. However, the Court finds that current spending under both schemes allows only for a limited impact, especially as neither scheme has a mechanism for targeting priority needs (paragraphs 27 to 37). For the Fruit Scheme, an expenditure ceiling of 500 million euro had been proposed, rather than the 90 million euro finally agreed, but the Commission felt in particular that the project first had to prove itself. As the scheme has only just been launched, therefore, it is still too soon to re-examine this aspect (see Annex IV).

86. In the light of these findings, the Court’s main recommendations are as follows:

(a) Given the very limited impact of the Milk Scheme, the matter of whether or not to retain it should be made conditional on the possibility of thorough reforms to remedy the weaknesses that have been identified. The decision should, in particular, take account of the product’s recognised nutritional value in the light of the public health objectives. If thorough reforms are to be undertaken, they should be based on the considerations set out below.

(b) If the Milk Scheme is to have a real impact on the volume of milk consumed in schools, the subsidy paid per kilogram should be increased very significantly to a level where milk can be distributed free of charge [50]. However, this recommendation would not entail increased EU spending; on the contrary, it would mean focusing resources on a narrower target population, which could also effectively contribute to addressing the problem of deadweight. The population to be targeted would have to be determined in the light of a scale of nutritional needs.

(c) Where the Milk Scheme is concerned, steps should be taken to limit the specific deadweight effect associated with distribution through canteens, while taking care to ensure maximum visibility for the scheme.

(d) Where the educational objectives are concerned, the two schemes should adopt a more harmonised assessment of the role and importance of accompanying measures. If the importance of these measures is confirmed, the desirability of making them eligible for EU co-financing could be reconsidered.

(e) There should be greater coordination and synergy between the two schemes in order to ensure a globally consistent approach to nutrition and that the programmes are managed efficiently.

87. The Court also has the following recommendations:

(a) Given the lack of a satisfactory system for measuring the schemes’ performance (see paragraphs 24 to 26), the performance monitoring mechanism should be improved to better reflect the provisions laid down in the Financial Regulation.

(b) Simplification measures should continue, and good practices facilitating implementation of the schemes should be encouraged.

(c) In the light of this goal of simplification, and with particular attention to the cost-effectiveness of controls, the points raised in connection with the regularity of expenditure (see Annex II) should, as far as possible, be resolved.

This report was adopted by Chamber I, at a meeting chaired by Mr Olavi ALA-NISSILÄ, Member of the Court of Auditors, in Luxembourg on 13 July 2011.

For the Court of Auditors

+++++ TIFF +++++

Vítor Manuel da Silva Caldeira

President

[1] Commission Regulation (EC) No 657/2008 of 10 July 2008 laying down detailed rules for applying Council Regulation (EC) No 1234/2007 as regards Community aid for supplying milk and certain milk products to pupils in educational establishments (OJ L 183, 11.7.2008, p. 17).

[2] Council Regulation (EC) No 13/2009 (OJ L 5, 9.1.2009, p. 1) and Commission Regulation (EC) No 288/2009 of 7 April 2009 laying down detailed rules for applying Council Regulation (EC) No 1234/2007 as regards Community aid for supplying fruit and vegetables, processed fruit and vegetables and banana products to children in educational establishments, in the framework of a School Fruit Scheme (OJ L 94, 8.4.2009, p. 38).

[3] Recital 43 of Council Regulation (EC) No 1234/2007 of 22 October 2007 establishing a common organisation of agricultural markets and on specific provisions for certain agricultural products (Single CMO Regulation) (OJ L 299, 16.11.2007, p. 1): "To contribute to balancing the milk market and to stabilise the market prices for milk and milk products". Recital 2 of Council Regulation (EC) No 13/2009: "[…] the School Fruit Scheme would meet the objectives of the CAP, including the promotion of earnings in agriculture, the stabilisation of markets and the availability of both current and future supplies".

[4] Recital 2 of Regulation (EC) No 657/2008: "In the light of the fight against obesity, and in order to provide children with healthy dairy products […]". Recital 4 of Regulation (EC) No 13/2009: "The clear health benefits of a School Fruit Scheme […]".

[5] For 2008, 2009 and 2010, actual expenditure, for 2011, forecast expenditure in the draft 2011 general budget.For the Fruit Scheme, see Annex IV for updated forecasts.

[6] Recital 4 of Regulation (EC) No 657/2008: "[…] the educational purpose of the scheme".Recital 2 of Regulation (EC) No 13/2009: "[…] by durably increasing the share of fruit and vegetables in the diets of children at the stage when their eating habits are being formed […] bring young consumers to appreciate fruit and vegetables and thus enhance future consumption".

[7] As some Member States have opted to narrow the range of bodies that may apply, the schemes are managed in very different ways. For example, authorisation to receive milk has been granted to over 11000 schools and local councils in France, around 40 suppliers in Spain and just four producers’ organisations in Italy.

[8] This is the rate established in Regulation (EC) No 657/2008. In the case of other eligible milk products (cheese, yoghurt, etc.), the regulation provides a scale by category for conversion into kilogram equivalents of milk.

[9] Article 5 of Regulation (EC) No 657/2008.

[10] Milk products "used in the preparation of meals within the premises of the educational establishment and which do not involve heat treatment may benefit from the aid" (Article 5(4) of Regulation (EC) No 657/2008).

[11] As the EU aid only covers a fraction of costs, distribution free of charge is mainly thanks to the national budgets. See paragraphs 61 and 63.

[12] The percentage varies between Member States, given the higher rate available in convergence and outermost regions.

[13] See paragraph 3(b) and Box 1.

[14] North Rhine-Westphalia and Bavaria.

[15] "Evaluation of the School Milk Measure", finalised in February 1999 by CEAS Consultants (Wye) Limited — Centre for European Agricultural Studies, and the Institute for the Management of Dairy Companies at Technische Universität München. Available on the Commission’s website at http://ec.europa.eu/agriculture/eval/reports/schoolmilk/.

[16] "Evaluation of the National Top-Up to the EU School Milk Subsidy in England", for the Department for Environment, Food and Rural Affairs (Defra), by London Economics in cooperation with Dr Susan New. Excerpts are given in Annex III, together with a link.

[17] Article 27(3) of Council Regulation (EC, Euratom) No 1605/2002 (OJ L 248, 16.9.2002, p. 1) requires "specific, measurable, achievable, relevant and timed [SMART] objectives" to be set "for all sectors of activity covered by the budget".

[18] As early as April 1996 an internal note at the Commission said: "The key factor in assessing effectiveness is the measure’s impact on the image of milk and consumption levels, but it is impossible for the departments in DG VI to appraise these areas or assess consumer response."

[19] For example, 7,3 million children — more than one third of all participants — were counted in France. However, these were pupils enrolled at participating establishments in France, whereas usually only those attending canteens actually benefit from the scheme. The figure given for beneficiaries in France was therefore inflated by approximately 3 million.

[20] At the average EU co-financing rate of around 58 %, the current budget ceiling of 90 million euro translates into total expenditure (Commission and Member States) of some 155 million euro.

[21] Both schemes aim to go beyond the immediate impact of the distributed quantities by having a lasting educational impact on eating habits. This aspect is examined in paragraphs 66 et seq.

[22] "The volume of milk and milk products supplied under the scheme is extremely small relative to the size of the EU market (0,3 % of total milk delivered to EU dairies in 1996/97). It is also declining in relative importance. […] At best, any impact of the scheme has been very small relative to the context of the total market and the scheme’s primary objective. […] "as a measure to help expand the market for milk products" and as a "surplus disposal mechanism", the scheme has been poor value for money."

[23] The Commission’s working estimate (shown by the audit to be unreliable) is that 21 million children, or roughly 20 % of the target population, are benefiting from the Milk Scheme. For the first year of the Fruit Scheme, the Member States visited had opted to concentrate on primary schoolchildren.

[24] Although there is no consensus on an exact figure the concept of a recommended amount is used in many countries.

[25] "Children in schools that did not participate in the SMSS [School Milk Subsidy Scheme] on average drank an amount of milk sufficient to meet their recommended daily intake of calcium. This is consistent with the most recent national diet survey, which found that average calcium consumption was adequate at ages 4–10."

[26] At the same time, in France the EU aid is essentially paid to canteens, which have proportionally fewer users from less-favoured social categories. Statistically, therefore, the EU aid is addressed less to target groups which are prioritised by national governments.

[27] It should be pointed out that, although the Fruit Regulation sets a higher co-financing rate for convergence and outermost regions, this decision was not based on a nutritional-needs analysis, and, moreover, allocation of the EU contribution is still based solely on the criterion of the number of children between 6 and 10 years of age.

[28] As the regulation does not impose a ceiling on expenditure, the potential maximum (approximately 900 million euro) was calculated on the basis of 25 cl per schoolday (the maximum daily amount allowed by the regulation) for every child in attendance at an educational establishment.

[29] From the 1999 evaluation: "Scheme take-up across the EU in 1996/97 was equivalent to only 12 % of the maximum subsidy entitlement volume. This compares with 19 % in 1992/93. This suggests a relatively poor level of efficiency and effectiveness in reaching the target population […]".

[30] Especially in view of the scheme’s objective of subsidising not just a product but an entire service, which entails specific packaging and logistical costs (e.g., in the UK, the physical delivery of fresh milk every morning).

[31] From the 1999 evaluation: "Price does not appear to be a major factor influencing take-up of milk and milk products by consumers including school children. […] This suggests that the main delivery mechanism of the School Milk Measure (price subsidy) is targeted at a factor of minor influence in determining consumption of milk and milk products. Not surprisingly, this means that its net impact on consumption is likely to be very small."

[32] From the 1999 and 2005 evaluations:

- "We also find that the EU School Milk Subsidy Scheme is an inefficient scheme. The combined SMSS and Top-Up spend around 7,2 million pounds in England annually, and create an estimated 5 million pounds of administrative costs in schools plus 831000 pounds of administrative costs in LEAs [local educational authorities] and the RPA [Rural Payments Agency]. […] We recommend such programmes provide free milk rather than milk at a price, since it is likely to be much cheaper to administer programmes that do not require teachers to track parents’ payments."

- "[…] the scheme’s administrative and financial requirements are considered by some to have acted as a disincentive to scheme take up at the school level. This stems mainly from the time commitments required to implement and administer school milk."

[33] Deadweight occurs where a measure is used to support beneficiaries who would have made the same choice in the absence of aid. In such cases, the outcome cannot be attributed to the policy, and the aid paid to the beneficiary has had no impact. Thus the share of expenditure which generates deadweight is ineffective by definition, because it does not contribute to the objectives. It is necessarily very difficult to measure deadweight accurately, since this would mean evaluating a situation — non-existence of the public subsidy — that is hypothetical.

[34] Over 90 % of schools receive aid, which is worth almost 120 cl per day for each child on the school rolls.

[35] The scaling-back of subsidies in the wake of the 1999 evaluation rendered the scheme still more ineffective, as it reduced still further its bearing on the decisions taken by applicants and potential beneficiaries.

[36] Recital 2 of Regulation (EC) No 288/2009: "So as to ensure the added value of School Fruit Schemes […], Member States should explain in their strategy how they will guarantee the added value of their scheme, especially where regular school meals are consumed at the same time as products financed under their School Fruit Scheme."

[37] This is also true of free distribution schemes in France. Because, however, these are far smaller in scale the overall national budget is far more modest.

[38] In the UK, until 2008 distribution free of charge (NMS) was wholly financed by the Ministry of Health. It was only in 2008 that the UK government took the decision to have the EU reimburse some of the costs of this national policy.

[39] Outsourcing organisation to a contractor does not eliminate administrative costs but merely serves to externalise them. The UK authorities thus allow the company in question to charge twice as much as is received from the EU. Nonetheless, externalisation (i) very probably brings significant administrative economies of scale and (ii) relieves decision-makers of a task which might otherwise induce them to withdraw their schools from the scheme.

[40] Recital 4 of Regulation (EC) No 657/2008: "[…] this is not an effective way of attaining the educational purpose of the scheme. Therefore the preparation of meals should be restricted accordingly." Article 5(4) of the same regulation prohibits heat treatment so that milk products remain visible and can be recognised by pupils.

[41] The 1999 evaluation noted in this regard that, "in cases where milk is consumed as a complementary product, the effectiveness of the subsidy scheme is even further diminished".

[42] The Milk and Fruit Regulations now require a poster to be displayed in the main entrance to each participating establishment. Without denying that this is a positive step, its impact is limited.

[43] "… the measure has probably had negligible effect on improving knowledge of the nutritional qualities of milk products. This is, however, not surprising given that the School Milk Measure has no funding provision for promotional activities and is essentially a price subsidy measure only."

[44] "The School Milk Subsidy Scheme could provide useful information to parents and children. However, sending delivery trucks to schools every day and requiring teachers to track parents’ payments appears a costly method of providing information."

[45] See http://ec.europa.eu/agriculture/tasty-bunch/about/index_fr.htm, where there is a link to the dedicated Milk Scheme site "Drink it up" at http://ec.europa.eu/agriculture/drinkitup/for_adults_fr.htm.

[46] Council Regulation (EC) No 3/2008 of 17 December 2007 on information provision and promotion measures for agricultural products on the internal market and in third countries (OJ L 3, 5.1.2008, p. 1) (budget heading 05 02 10 01).

[47] We came across rare examples of explicit references to the School Milk Scheme. However exceptional, these marked progress compared with the total absence of any synergy found by the 1999 evaluation: "The main finding of the research shows that the School Milk Measure has hardly figured in any milk product promotions across the EU".

[48] Recital 5 of Regulation (EC) No 288/2009.

[49] "With regard to the promotion of fresh fruit and vegetables, particular attention shall be paid to promotion measures intended for children in schools"; "The percentage referred to in the first subparagraph shall be 60 % for measures for the promotion of fruit and vegetables intended specifically for children in schools of the Community" (Articles 5(1) and 13(2) of Regulation (EC) No 3/2008).

[50] This might mean returning to the idea of Member State or even industry co-financing.

--------------------------------------------------

ANNEX I

COMPARISON OF THE SCHOOL MILK SCHEME AND THE SCHOOL FRUIT SCHEME

| School milk | School fruit |

Eligible products | Milk and milk products listed in the regulation (milk, flavoured milk, yoghurt, cheese, etc.) | Fruit and vegetables (including processed foods without added sugar, fat, salt or sweeteners) |

Objectives | Market | To help stabilise the market /CAP general objectives |

Nutrition and health | Short-term: to include healthy products in children’s diets Long-term: educational, to promote lasting healthy eating habits |

Structure of the scheme | Financial subsidies for the sale of milk at a reduced rate / distribution of fruit in schools |

Target beneficiaries | Children in regular attendance at an educational establishment |

Recipients of aid | Educational establishments, suppliers, education authorities or intermediary bodies set up specifically for the purpose These applicants must be approved beforehand by the responsible national authorities |

Publicity | Posters at the main entrance to establishments |

Level of aid/eligible expenditure | Scale of payments by product category (reference base 18,15 euro/100 kg milk) | Co-financing (50–75 %) of the cost of distributed products (+ certain associated costs) |

Do beneficiaries contribute? | EU aid is only sufficient for a rebate on the sale price Where distribution was observed to be free of charge it was dependent on additional funding | Regulation suggests but does not require distribution free of charge No observed cases of sales at a reduced rate, but parents may indirectly finance part of the cost |

How are products made available to children? | Sold at a reduced rate in classrooms or through a school shop Classroom distribution free of charge Milk products included in canteen meals | Distribution free of charge outside mealtimes (distribution authorised in canteens only if the Member State can demonstrate added value – in practice no cases declared) |

Budget ceiling | No budget ceiling Article 5 of Regulation (EC) No 657/2008 stipulates, however, a maximum subsidisable quantity of 0,25 litre per school day and per pupil in regular attendance at the establishment; this ceiling is not usually reached by the applicants | 90 million euro |

Conditions for participation by schools | No | MS national strategy |

Accompanying measures | No | Compulsory, stated to be crucial Organisation and financing not covered by the scheme |

--------------------------------------------------

ANNEX II

SUMMARY OF OBSERVATIONS SENT TO THE COMMISSION CONCERNING THE REGULARITY OF EXPENDITURE

SCHOOL MILK

PROBLEMS INTERPRETING THE REGULATION

1. Eligibility of nurseries: owing to differences in the way the regulation was translated and is interpreted in different languages, nurseries did not become eligible in France until 2008 [1], are still ineligible in Italy and Poland but are eligible in Germany, Sweden and the United Kingdom [2].

Although nurseries are explicitly allowed by some language versions, their eligibility does not square with the regulation’s definition of beneficiaries as "pupils" in attendance at "educational establishments". Generally speaking, in the countries visited these terms are inappropriate for nurseries.

2. Approval of local authorities as "education authorities": Regulation (EC) No 657/2008 lists four categories of eligible applicants. Local authorities, which are an indispensable part of the scheme in many Member States, are approved under category (b).

Although the French and Italian versions of the previous regulation defined this category in far broader terms ("le pouvoir organisateur effectuant la demande d’aide" / "l’amministrazione responsabile"), a standardised term ("education authority") is used in Regulation (EC) No 657/2008. However, this raises the question as to whether local authorities may legally be considered, in every Member State, to be "education authorities".

3. Reflecting the level of aid in canteens: the regulation requires Member States to ensure "that the amount of the aid is duly reflected in the price paid by the beneficiary" [3]. It appeared essential to include this clause in order to guarantee that the aid actually reaches the target beneficiaries. However, where canteens and nurseries charge beneficiaries a fixed one-time fee there are in practice no specific checks on this requirement, which is very often unverifiable.

4. Organisation of on-the-spot checks by the Member States: a legal procedure has been launched between the Commission and one Member State about its interpretation of the requirements relating to on-the-spot checks (in particular the notions of ex ante and ex post checks). However, the audit found that the other Member States do not all interpret these provisions in the same way. In order to ensure that a standard approach is used, there should be a consultation process involving all parties.

Given that the two regulations are broadly similar in this respect, the argument also applies to the Fruit Scheme.

SPECIFIC PROBLEMS DETECTED IN THE MEMBER STATES THAT WERE VISITED

5. Some examples of problems highlighted in the various Member States:

(a) The audit revealed deficiencies in one Member State’s system of controls. No documents could be obtained for certain checks, in other cases the inspectors’ work was not suitable for detailed checks of the quantities declared, and in one case the quantities declared were not justified.

(b) In another Member State, the legal basis for approval of the main applicant requires confirmation, as the applicant does not directly meet the formal conditions in the regulation.

(c) In a third Member State, applications from establishments and local authorities were often drawn up on the basis of purchase statistics which were usually not verified during controls. In addition, quantities consumed by ineligible beneficiaries were often deducted in a subjective and not easily verifiable manner.

(d) In a final Member State, seven of the nine establishments visited had not taken the trouble to put up an information poster as required by the regulation.

WORKLOAD IMPOSED BY EXISTING CONTROLS

6. Under most present arrangements, both the authorities and applicants are faced with a workload that is very significant when compared with the often very small available amounts (see in particular paragraphs 42 to 45 and Box 8).

7. The problems highlighted must therefore be resolved bearing in mind the cost-effectiveness of the controls and requirements established in the regulation.

SCHOOL FRUIT

8. The first payments in Member States under the Fruit Scheme were not made until the beginning of 2010, and it is therefore too soon to assess the effectiveness of the monitoring and control system. Nonetheless, a number of specific problems were identified in the way the regulation is interpreted. These are set out in paragraphs 9 to 11 below.

9. Funding of accompanying measures that are part and parcel of a fruit delivery contract: in one Member State the scheme was implemented through a single call for tenders, which included accompanying measures. Although, under Article 5 of Regulation (EC) No 288/2009, these measures are not eligible for EU co-financing, the audit found that they accounted for a substantial share of the tender value.

10. Difficulty in defining what constitutes a satisfactory accompanying measure, both quantitatively and qualitatively, within the meaning of the regulation: not one of the three other Member States visited had yet resolved this matter or decided on the framework for a relevant system of controls and sanctions.

11. Co-financing of "costs": the regulation provides for the costs of products delivered to schools to be co-financed. This clause is directly relevant where aid is paid to the purchaser (school or local council, as in France). However, where aid goes to the supplier (as in Germany, Italy and Poland), the notion of costs, as used in the regulation, is not directly relevant because the level of aid can only be calculated on the sale price, which necessarily includes a profit margin.

[1] In the French version of the regulation, the earlier term "écoles maternelles" (Article 2(1)(a) of Commission Regulation (EC) No 2707/2000 (OJ L 311, 12.12.2000, p. 37)) has been replaced by "crèches ou autres établissements d’éducation préscolaire" (Article 2 of Regulation (EC) No 657/2008).

[2] In the other relevant language versions the term used has not been changed — "nursery (school)" (English), "scuola materna" (Italian), "przedszkola" (Polish), "Kindergarten" (German), "förskola" (Swedish). However, differences in what these terms mean have resulted in different applications/interpretations.

[3] Article 14(1) of Regulation (EC) No 657/2008.

--------------------------------------------------

ANNEX III

SUMMARY OF THE 2005 EXTERNAL EVALUATION IN THE UK OF THE NATIONAL TOP-UP TO THE EU SCHOOL MILK SUBSIDY IN ENGLAND

1. In 2005 the United Kingdom carried out an evaluation of the National Top-Up to the EU School Milk Subsidy in England. The scope of this evaluation was limited to sales of milk at a reduced price and did not include the national scheme for distributing milk free of charge to children aged under 5 (see paragraphs 8 and 61).

2. The conclusions of this study were highly critical of the measure as being both inefficient and ineffective. In particular, the evaluation highlighted the scheme’s low added value and high management costs, and the fact that the subsidised milk costs pupils more than the supermarket price.

3. The report concluded that the National Top-Up was poor value for money and recommended that the government should stop paying it:

"1) […] Because the Top-Up creates high administrative costs and has apparently small benefits for child health, the UK government should consider ending it. […].

2) In light of the costs and benefits of the EU School Milk Subsidy Scheme, and the weak rationale for a programme to make subsidised milk available to schools, the UK government should consider discussing with the European Commission the idea of ending the EU SMSS. […]

[…] Were the Top-Up and perhaps also the EU SMSS to end, the government might be interested in other efforts to promote young people’s consumption of milk. We feel that large-scale school milk programmes do not represent a sensible use of public funds, but that justification exists for narrowly-targeted programmes. We recommend such programmes provide free milk rather than milk at a price, since it is likely to be much cheaper to administer programmes that do not require teachers to track parents’ payments."

4. The full text of the evaluation, which was carried out for the Department for Environment, Food and Rural Affairs (Defra) by London Economics in cooperation with Dr Susan New, can be consulted at:

http://archive.defra.gov.uk/evidence/economics/foodfarm/evaluation/school milk/fullreport.pdf.

--------------------------------------------------

ANNEX IV

A HESITANT START TO THE FRUIT SCHEME

LAUNCH MUCH MORE HESITANT THAN EXPECTED …

1. Compared with the initial target of 90 million euro in the first school year [1], just 33 million euro was disbursed in 2009/10. Of the Member States that were chiefly concerned, only Italy was able to implement the scheme as described in its strategy (see Graph).

COMPARISON BETWEEN BUDGET ALLOCATION AND EXPENDITURE INCURRED FOR THE 2009/10 SCHOOL YEAR (IN MILLION EURO)

+++++ TIFF +++++

Source: European Commission.

2. These initial results show both that the Commission was being realistic when it recommended reducing the budget proposed by the European Parliament [2], and that its own forecasts for the launch phase were still far too optimistic.

3. Of the countries that were visited:

(a) In Poland the delays were caused by administrative hold-ups and by the fact that the national implementing rules initially made it hard to find suppliers willing to make deliveries.

(b) In Germany, the launch was delayed by talks between national government and the regions on the topic of co-financing.

(c) In France too the very slow start was due to difficulties in finding co-financing. In the absence of a national contribution, each establishment’s participation is dependent on local funding (usually from the local council). Obtaining this local complement has so far proved problematic.

… BUT INTEREST IN THE SCHEME REMAINS HIGH

4. In the United Kingdom, the main reason for non-participation is that a far more ambitious national scheme of free distribution is already operational (worth 44 million pounds per year in England alone).

5. The huge delays experienced virtually everywhere appear to have been caused by an imbalance between, on the one hand, political support and enthusiasm for a rapid project launch and, on the other, the various partners’ practical inability to organise the launch on schedule. The timetable adopted by the Commission did not allow the Member States to know what their 2008–09 budgetary allocation was until July 2009, whereas once the scheme is running smoothly the Commission decision will be taken by 31 March of the previous year.

6. Although these delays have therefore also postponed the scheme’s anticipated impact, this does not mean that there is no interest. Where the scheme is operational the first positive echoes are being received, and 14 of the participating Member States have increased their budget forecasts for 2010/11 [3].

7. In all probability, participation will soon be limited only by the availability of resources from the EU or national budgets.

[1] The 2010 budget was based on the assumption that the entire available amount would be spent (two thirds of it in the 2010 financial year).

[2] The Parliament had initially proposed a budget of 500 million euro. In the end it was the Commission’s proposal of just 90 million euro that was agreed. The Commission felt in particular that the project first had to prove itself on a smaller scale.

[3] Nonetheless, as Member States were unable to assess their needs for the first year, it is not yet possible to determine whether these new requests will prove to have been more realistic.

--------------------------------------------------