The impact of the widened milk quota on the sales revenues of dairy farmers of the EU-15

(1)

The impact of the widened milk quota on the sales revenues

of dairy farmers of the EU-15

Master Thesis

Group NL2 Student Hannah van Vorselen Student number 6146783 E-mail address h.vanvorselen@hotmail.com Date 28-2-2016 Supervisor drs. N.J. Leefmans Second reader dr. D.J.M. Veestraeten Number of words 15101 Master Economics Specialization: International Economics and Globalization

(2)

2

Statement of Originality

This document is written by Hannah van Vorselen, who declares full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

The milk quota regime, which the European Union has used to restrict its raw cow’s milk production for 31 years, was gradually phased out between 2008 and 2014, until it was entirely abolished on 1 April 2015. This study has investigated the impact of this on the total sales revenues of the EU-15 and the individual member states up to 2014. The approach chosen comprises a literature review combined with a descriptive data analysis. In line with what was expected, the analysis has shown that the raw milk production of the EU-15 as a whole has gradually increased from 2008 to 2014. This increase was brought about by Austria, the Netherlands, Luxembourg, Denmark, Germany, Belgium, Spain, Italy, France and Ireland. It is safe to conclude that this would not have been possible without the expanded quota. At the same time, the production of Portugal, the UK, Greece, Finland and Sweden has decreased during the quota expansion period. Contrary to expectation, this was not caused by a price-depressing effect from the increased production of the other member states. Any price-depressing effect was more than offset by positive exogenous factors, such as a strong demand for dairy products. What has motivated the farmers of these countries to decrease their production requires further research. Therefore, it was only possible to draw a conclusion regarding the member states which increased their production. Since they were able to generate higher revenues through their increased production, the impact of the quota expansion on the revenues of these member states has proved to be entirely positive.

(4)

4

List of Figures

Figure 2.1 Dairy Supply Chain ... 9

Figure 2.2 Dairy Supply Chain - Including Policy Measures ... 11

Figure 2.3 Development of the Overall EU-9 Milk Quota ... 16

Figure 2.4 Theoretical Impact on the Raw Milk Market ... 17

Figure 4.1 Aggregate EU-15 Production Collected By Dairies ... 30

Figure 4.2 Weighted Average monthly EU-15 Raw Milk Price from 1977 to 2014 ... 34

Figure 4.3 Total EU-15 Raw Milk Sales Revenues ... 37

List of Tables

Table 2.1 Historical Data of the Milk Quota (in 1000 tonnes) ... 13

Table 2.2 Annual Adjustments to the Milk Quota ... 14

Table 2.3 Simulated Price and Production Effects (Relative To a Baseline Scenario) ... 20

Table 3.1 Included (White Area) and Excluded (Grey Area) Member States ... 25

Table 4.1 Percentage Quota Overrun or Underuse ... 28

Table 4.2 Percentage Change in Average Raw Milk Production ... 31

Table 4.3 Percentage Change in Average Raw Milk Price ... 33

Table 4.4 Percentage Change in Average Raw Milk Sales Revenues ... 38

(6)

6

List of Abbreviations

Abbreviation Explanation AT Austria BE Belgium CAP Common Agricultural Policy DE Germany DK Denmark EC European Community EEC European Economic Community EL Greece ES Spain EU European Union

EU-15 European Union, 15 member states: EU-9 + Greece, Spain, Portugal, Finland, Austria, Sweden

EU-28 European Union, 28 member states: EU-15 + Bulgaria, Czech Republic, Cyprus, Estonia, Hungary, Malta, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia, Croatia

EU-9 European Union, 9 member states: Germany, France, Italy, Luxemburg, the Netherlands, Belgium, Denmark, Ireland, the United Kingdom FI Finland FR France IE Ireland IT Italy LUX Luxemburg NL the Netherlands PT Portugal SQ Sub-question SV Sweden UK the United Kingdom

(7)

7

1. Introduction

The milk quota regime, with which the European Union (EU) has restricted its raw cow’s milk production for 31 years, came to an end on 1 April 2015. To prevent a sudden price drop, the European Commission decided on a so-called ‘soft landing’ policy instead of an abrupt quota removal: from 1 April 2008 to 1 April 2015 the quota were gradually increased by one or two percent per year. The quota were originally introduced in 1984 to curb the EU’s structural overproduction of raw milk (products) during the late 1970s and early 1980s, that had led to the infamous milk lakes and butter mountains (European Commission, 2015). The primary reason for abolishing the quota was that, driven by population and income growth, there had been a considerable increase of worldwide consumption of dairy products, which was projected to continue. In view of this growing demand, concerns were raised that continuing production restrictions would lead to a loss of the EU’s world market share (European Commission, 2012).

The announcement of the gradual quota abolition in 2008 led to much discontent. Small farmers in particular, who have little expansion potential, worried that they would be unable to survive without the quota regime. Despite the gradual abolition, they feared that lifting the quota would lead to a sharp raw milk price fall as a result of expanded EU raw milk production (Matthews, 2010). They argued that lifting the quota would only help the largest farmers, whose economies of scale allows them to operate with slim margins. These concerns led to ongoing protests all across Europe in the run-up to the abolition. Nevertheless, the European Commission remained convinced that - unlike the situation in 1984 - the EU now had an internationally competitive exporting dairy sector, that was ready for the free market. Since major change in the EU’s dairy sector was expected due to the gradual quota abolition, several economists, often commissioned by the European Commission, analysed the potential impacts for the various member states and the EU as a whole in advance. They expected an increase of aggregate raw milk production, a redistribution of raw milk production across member states, downward pressure on the raw milk prices and, depending on the country-specific price and production effect, a heterogeneous impact on the sales revenues across member states (Colman, 2002; Chantreuil et al., 2008; Réquillart, Bouamra-Mechemache, & Jongeneel, 2008; IPTS, 2009; Witzke & Tonini, 2009). Based on an assessment of the expansion potential of the various member states, it was moreover expected that the impact of the gradual quota abolition would occur predominantly during the quota expansion period itself and not so much afterwards (IPTS, 2009). Enough time has now elapsed to allow for an ex post, short term, impact analysis of the widening of the quota. This study attempts to do just this, by examining what has been the actual impact of the widening of the milk quota on the raw milk sales revenues of the EU-15 as a whole and of the individual

(8)

8 member states up to 2014. The approach chosen comprises a literature review combined with a descriptive data analysis. To be able to substantiate the net impact of the quota expansion on the sales revenues, the impact on the production and price will first be analysed separately. To avoid attributing a production and price change to the quota expansion, while they were in fact caused by exogenous factors, it will moreover be determined if, and to what extent, the quota of the various member states were still binding during the quota expansion period. The idea is that if the quota were not binding for certain member states, they had room to increase their production regardless of the quota expansion. Subsequently, the production, price and sales revenues of both the EU-15 as a whole and of the individual member states will be analysed, explained and compared with the previous decade. This study is broadly structured as follows. Chapter 2 presents the literature review, which starts with a brief explanation of the dairy supply chain and certain terms that go with it. Then, it provides a chronological overview of the history of the EU’s dairy policy, to clarify why the milk quota were introduced in the first place, and to illustrate that the quota were part of a wide range of policy instruments. Subsequently, attention will be paid to the development of the height of the quota from the moment they were introduced in 1984 until they were abolished in 2015. Then, the focus will shift from the policy change itself to its theoretical impact. Finally, the expectations of previous studies regarding the magnitude of the impact will be discussed. After the literature review, the methodology is presented in chapter 3, where the research question, sub-questions, research method, geographical scope and used data, are all accounted for. The results of the analysis are presented and commented upon in chapter 4, followed by the conclusion in chapter 5.

2. Literature review

2.1. The dairy supply chain

In favour of the readability of the rest of the study, the literature review starts with a brief explanation of the dairy supply chain and certain terms that go with it. The chain starts with the production of raw milk by dairy farmers (1st_{stage). The supply of raw milk enters the market via two routes. A small} percentage is processed on the farm for own use, such as for direct sales to final consumers, for the use in cattle feed, or for own consumption, but the vast majority is delivered to dairy factories, so-called ‘dairies’, like for example Campina (2nd_{stage). Dairies process the raw milk into dairy products,} such as, cheese, yoghurt, drinking milk, butter and milk powder. These dairy products are subsequently sold to EU retailers, exported to non-EU retailers, or stored (3th_{stage). In the last stage, retailers sell} their dairy products to final consumers (4th_{stage). For clarity, the various stages of the dairy supply}

(9)

9 Kavallari, 2011, p. 109). Figure 2.1 Dairy Supply Chain Source: Constructed based on Figure 4.5 from ‘Evaluation of CAP measures applied to the dairy sector’, by Jongeneel, Burrel and Kavallari, 2011, Wageningen University & Research Centre, p. 109.

2.2. The history of the EU’s dairy policy

To clarify why the milk quota were introduced in the first place, and to illustrate that the quota were part of a wide range of policy instruments, this section will provide a chronological overview of the history of the EU’s dairy policy.

The dairy policy finds its origin in 1957, when Belgium, France, Italy, Luxembourg, the Netherlands, and West-Germany signed the Treaty of Rome. This created the European Economic Community (EEC), a precursor of the European Community (EC) and today’s EU. The aim of the EEC was to preserve peace and liberty among its member states (Treaty of Rome, 1957). One of the means by which they tried to achieve this, was by promoting balanced economic growth. To that end, a customs union with a common external tariff was created and several common policies were developed, including the much discussed Common Agricultural Policy (CAP) (European Commission, 2010). The CAP came into force in 1963 (Delayen, 2007). Following post-war shortages, its key objective was to ensure that the EEC would become self-sufficient in terms of food (Delayen, 2007). To help accomplish this goal of self-sufficiency in the field of dairy products, the CAP

(10)

10 introduced two protectionists policy measures in 1968: import tariffs for dairy products1 and a so-called ‘intervention purchasing system’ for butter, skimmed milk powder and cheese2_{(see Regulation} (EEC) No 804/68). The import tariffs supported the EC’s price of dairy products by protecting them against foreign competition, while the intervention purchasing system stimulated the supply of the EC’s dairy products by guaranteeing to buy up all supplies of butter, skimmed milk powder and cheese for a predetermined 'intervention price', regardless of whether there was a real market for it (Colman, 2002). Given that dairy products are made of raw milk, these policies kept the raw milk price artificially high and gave a strong impetus to raw milk supply. As a result, by 1974, the EC had changed from a net importer into a net exporter of dairy products.

Despite the fact that the point of self-sufficiency had been reached, the EC maintained its system of import tariffs and intervention prices and, since the intervention price also applied to exports, this led to the use export subsidies from 1974 onwards (U.S. International Trade Commission, 2004, p. 49). Since the EC kept buying up dairy products for which there was in fact no demand, surpluses emerged, which were either stored, or sold on the world market against lower world market prices. Ultimately, due to high storage costs, the surpluses were simply spread out someplace, which resulted in the undesirable phenomenon of milk lakes and butter mountains (Watkins & Braun, 2003). Given the high cost of the dairy policy and the food waste that accompanied it, the policy became politically unsupportable in 1983 (Colman, 2002)3_{. There was debate in then ten member states about} whether cutting the intervention prices might be a good idea, but this possibility was outvoted by the farm lobby, who rather curbed production via quota so as to keep prices high (Economist, 2008). Following the demands of the farm lobby, the European Commission indeed introduced the milk quota regime in 1984. The quota covered only the production of raw cow’s milk, not sheep’s milk or goat’s milk. Based on historical production volumes, the overall EU quota was split into national quota per member state, which were in turn split into individual quota per dairy farmer. The quota were made effective by charging a substantial tax, known as the ‘super-levy’, on deliveries of raw milk above each national quota (Baldock, et al., 2008). The levy was collected from individual producers who over-delivered, but only when the national quota was exceeded (IPTS, 2009). While the milk quota regime indeed limited further growth of overproduction of raw milk and dairy products, the overproduction remained, because the quota were so high that aggregate production still overran the amount necessary for self-sufficiency. This required the continued use of export subsidies (Jongeneel, Burrell, & Kavallari, 2011). To erode the magnitude of these costly export 1 _{The dairy products for which an import tariff existed are defined Article 1 (a) to (g) in Regulation (EEC) No 804/68.} 2_{The cheeses for which an intervention price existed were Grana padano and Parmigiano Reggiano.} 3_{The budget costs of the quota regime were stuck at over 30% of the total CAP’s market support budget.}

(11)

11 subsidies, from 1984 to 1992, the EC reduced the quota limits, made the super-levy more dissuasive, reduced the intervention prices and export subsidies, and stimulated internal demand for butter by authorizing member states to provide consumption subsidies4_{(Jongeneel, Burrell, & Kavallari, 2011).} With milk supply at last under control, the dairy policy remained broadly the same from 1992 to 2003. To clarify to which particular part of the dairy supply chain the policy measures applied, the EU’s dairy policy measures that were used up to 2003 are displayed in Figure 2.2 (adapted from Figure 4.6 from Jongeneel, Burrell, & Kavallari, 2011, p. 110). While import tariffs were part of this package, they could not be shown in the figure, because they applied to imported products from abroad. Figure 2.2 Dairy Supply Chain - Including Policy Measures Source: Constructed based on Figure 4.6 from ‘Evaluation of CAP measures applied to the dairy sector’, by Jongeneel, Burrell and Kavallari, 2011, Wageningen University & Research Centre, p. 110.

After years of unchanged dairy policy, the CAP Mid-Term-Review in 2003 can be classified as the starting point of a liberalisation of the dairy market. With the aim to improve the market orientation and competitiveness of the sector, three substantial reforms were implemented from 2004 to 2008: the intervention prices and consumption subsidies were lowered and a direct income subsidy5_was 4_{Member states were allowed grant a consumption subsidy to whatever measure they considered to be appropriate, as long} as it benefited the final price paid by the private consumer and did not exceed 50 ECU4 per 100 kilograms of butter (see Regulation (EEC) No 1307/85 (3)). 5_{This direct income subsidy was called the Dairy Premium, which was later integrated in the Single Farm Payment.}

(12)

12 introduced. However, more startling was the announcement that the milk quota regime would come to an end in 2015. Other long-standing elements of the EU’s dairy policy, such as the import tariffs and the export subsidies, were maintained (European Commission, 2012). Whether the quota would be abolished gradually or in one go was left open until the CAP Health Check in 2008. There, it was finally announced that, in order to prevent a sudden price drop, the quota would be gradually phased out between 2008 and 2014, until they would be entirely abolished on 1 April 2015. The impacts of the policy reforms from 2004 to 2008 on the revenue of dairy farmers have already been evaluated in an extensive report of Jongeneel, Burrell and Kavallari (2011) 6_. Based on a comparison of the levels and trends in monthly raw milk prices before 2004 with those from 2004 to March 2007, they concluded that the policy reforms from 2004 to 2008 had led to a small decrease of the average raw milk price of the EU-15 as a whole between 2004 and 2007. Since the total EU-15 production had remained constant throughout these years, the total revenues of the EU-15 had also decreased. The total income of dairy farmers, on the other hand, had slightly increased. This counterintuitive result was obtained by the fact that the decline in the raw milk price was more than compensated by the introduction of the direct income subsidy. The impact of the widening of the quota on the sales revenues of the dairy farmers of the EU-15 member states from 2008 to 2014, will be investigated in this study.

2.3. Phasing out the milk quota

Now that the context of the introduction and abolition of the quota regime is clear, this section will proceed with a more detailed discussion of the development of the milk quota of the EU-15 as a whole and of the individual member states. Moreover, it will shed light on the motivation underlying specific reforms. To be able to see the recent quota developments from a broad perspective, attention will be paid to the evolution of the quota from the moment they were introduced in 1984 until they were abolished in 2015. Table 2.1 shows the quota over the years for EU-15 as a whole and for the individual member states (Special Report No 6/2001; Milk Market Observatory, 2015). The heights of the quota differ considerably across states, because the initial level of the quota in 1984 was determined based on historical production. For example, judging from the height of the quota, France was the largest milk producer in 1984 and Luxembourg the smallest. By comparing the initial quota with those of thirty years later, it can be seen that, on the whole little has changed. Nevertheless, there have been some 6_{Following the implementing rules of the Financial Regulation applicable to the general budget of the EU, the impacts and} effectiveness of the CAP have to be evaluated every six years by the Commission’s Directorate-General for Agriculture and Rural Development (see Council Regulation (EC) No 1605/2002).

(13)

13 adjustments. To be able to better identify these adjustments, Table 2.2 shows the annual adjustments to the quota of the various member states relative to and expressed as a percentage of their quota the previous year. In both tables, time is depicted in a somewhat unusual way, reflecting the fact that the milk marketing year does not start on 1 January, but on 1 April and ends on 31 March the following year. Table 2.1 Historical Data of the Milk Quota (in 1000 tonnes)7 Note: A dash (-) indicates that information on the quota amount is missing and a colon (:) that the country had no quota because it was not yet an EU member state in that year. Therefore, data regarding the annual quota of the EU-15 could only be presented from 1995 onwards. DE = Germany; FR=France; IT=Italy; LU=Luxembourg; NL= Netherlands; BE=Belgium; IE=Ireland; UK= United Kingdom; EL= Greece; ES=Spain; PT=Portugal; FI=Finland; AT=Austria; SV=Sweden. Source: Constructed based on data from Annex III of ‘Milk quota 2014/2015’, by Milk Market Observatory, 2015, European Commission, p. 6, retrieved from http://ec.europa.eu/agriculture/milk-market-observatory/pdf/eu-milk-quota-figures_en.pdf; Annex 2 of ‘Special Report No 6/2001 on milk quota, together with the Commission’s replies’, 2001, Official Journal of the European Communities, p. 21, retrieved from http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52001SA0006&from=EN. 7_{The milk quota actually contained two separate quota: one for raw milk delivered to dairies and (a much smaller) one for} raw milk produced for own use, such as for direct sales to final consumers, for the use in cattle feed, or for own consumption. Nevertheless, Table 2.1 only displays the sum of both quota, because the separate quota were not publicly available for all years and all member states.

(14)

14 Table 2.2 Annual Adjustments to the Milk Quota Note: The table presents the annual adjustments to the height of the quota relative to and expressed as a percentage of their quota the preceding year. An empty box indicates that the quota was unchanged. A dash (-) indicates that information on the quota amount is missing. A colon (:) indicates that the country had no quota because it was not yet an EU member state in that year. Therefore, data regarding the annual quota increase of the EU-15 could only be presented from 1996 onwards. DE = Germany; FR=France; IT=Italy; LU=Luxembourg; NL= Netherlands; BE=Belgium; IE=Ireland; UK= United Kingdom; EL= Greece; ES=Spain; PT=Portugal; FI=Finland; AT=Austria; SV=Sweden. Source: Constructed based on calculations with the data from Table 2.1. From Table 2.2 it can be seen that the quota were gradually reduced with around 9% from 1984/85 to 1992/93. This was in line with the policy objective of that time, which was to erode the magnitude of the costly export subsidies (see section 2.2). In 1993/94, the quota of all member states were adjusted upwards, because in particular Italy, Greece and Spain argued that their initial quota were set to low. The quota of Italy, Greece and Spain were indeed increased substantially more than the quota of the other member states. Thereafter, from 1994/95 to 1999/00, as a result of the entrance of new member states, the overall EU quota increased, but no further adjustments were made to the national quota of the existing member states. In 1999/00, as part of CAP reform titled the ‘Agenda 2000’, for specific local reasons, quota

(15)

15 increases of around 2,5%, that could be spread over a two-year period, were granted to four8_member states. This resulted in a relatively small overall quota increase of 1,2% in two years. During the Agenda 2000 negotiations, Denmark, Italy, Sweden and the UK furthermore unsuccessfully campaigned for a dismantling of the quota regime arguing that it inhibited the EU’s access to growing export markets. However, the majority of EU member states supported the continuation of the milk quota regime (Benjamin, Gohin, & Guyomard, 1999). In the CAP reform of 2003, titled the ‘Mid-Term-Review’, it was announced that the milk quota regime would come to an end in 2015. In the meantime, again for specific local reasons, Greece and Portugal were granted one-off quota increases of 17,1% and 2,7% respectively (Greece in 2004/05 and Portugal in 2005/06). Moreover, quota increases of 1,5% that could be spread over a three-year period starting from 2006/07, were distributed amongst those EU-15 member states who had not already received a quota increase in 1999/00. Although the period from 2000/01 to 2007/08, in which the overall quota increased with around 2,1%, can be characterised as one in which the quota became less restrictive, substantial quota increases were not scheduled for implementation until 2008/09. In March 2008, in advance of the most recent CAP reform titled the ‘Health Check’, the European Council approved a 2% increase of the quota of all member states for 2008/09 (on top of the 1,5% that could be spread over a three-year period starting from 2006/07). This extension was meant to meet the growing worldwide demand for dairy products on the one hand and to curb the at that time sharply rising EU dairy prices on the other (see Council Regulation No 248/2008 (5)). In November 2008, during the ‘Health Check’, the EU agricultural ministers made the definite decision to abolish the milk quota on 1 April 2015 and moreover decided upon a gradual phasing-out of the quota, starting immediately. The decision to proceed to a gradual widening of the quota to the end of the quota era, was made in response to studies which showed that this would lead to a smaller and more gradual price decline than an abrupt quota removal and hence would reduce the need to intervene after the quota had ended (Witzke & Tonini, 2009; Réquillart, Bouamra-Mechemache, & Jongeneel, 2008). To allow for this so-called ‘soft landing’ of the milk sector towards the end of quota, quota were increased with 1% each year from 2009/10 to 2013/149_{. In addition, the fat correction} factor10 _{was adjusted, which de facto resulted in a further quota increase of 1% (see Council Regulation} 8_{Italy, Spain, Greece and Ireland.} 9_{For Italy, the stepwise quota increases were implemented as a one off increase of 5% in 2009/10 to reduce the likelihood} of the surplus levy being incurred. 10_{Next to the individual quota, a reference fat content was assigned to each individual producer. Any difference between the} real fat content and the reference fat content led, at the end of the quota year, to an adjustment in the volume of each individual delivery (the quantity produced for dairies per farm). Before the CAP Health Check, the so-called butterfat adjusted volume was calculated as follows: the amount of raw milk delivered was multiplied by 0,18% per 0,1g milk fat/kg in excess of

(16)

16 72/2009 (14)). To ensure that the quota increases would lead to a controlled and smooth transition, it was decided that farmers who exceeded their quota by more than 6% in 2009/10 and 2010/11 had to pay a levy 50% higher than the normal penalty (European Commission, 2009). In 2014/15, the last year of their existence, the quota remained the same. The development of the overall EU quota from 1984 to 2015 is summarized in Figure 2.3. It is easy to observe that the quota were substantially reduced from 1984/85 to 1992/93 (in total by around 9%), followed by seven years of stable quota from 1993/94 to 1999/00. In 2000/01, some member states were granted specific quota increases for local reasons but most member states only received a small quota expansion in 2006/07. This resulted in a small quota increase of 2,1% over an eight-year period from 1999/00 to 2007/08. From 2008/09 to 2014/15, with a view to meet the growing worldwide demand for dairy products and to allow for a soft-landing to the end of the quota regime, the quota were increased more substantially until they fully expired 1 April 2015. Over the seven-year period from 2008/09 to 2014/15 they were increased in total by around 7,6% and on average by 6,6% per year. The analysis will focus on the impact of this last, substantial, quota expansion which runs from 1 April 2008 to 31 March 2015, from now on referred to as the quota expansion period. Figure 2.3 Development of the Overall EU-9 Milk Quota Note: To see the development of the quota over the entire period clearly, the bars represent the sum of the quota of the following 9 member states: Belgium, Denmark, Ireland, Greece, Spain, France, Italy, Luxembourg and the Netherlands. With the exception of Germany, these are the member states that were already a member in 1984. Including Germany would have given a misrepresentation of reality, since the quota of Germany was increased by 26% in 1990/91 in the context of the reunification of East and West Germany. Source: Constructed based on calculations with the data from Table 2.1. the reference fat level or reduced if the fat was less than the reference level. In the context of the CAP Health Check, this rate was reduced to 0,09%, leading to the de facto increase of the quota of 1%.

(17)

17

2.4. The theoretical impact of the quota expansion

For the purpose of better understanding the possible impact of the quota expansion, it is useful to explain the consequences of quota expansion from a theoretical perspective. To develop the right intuition, use will be made of Figure 2.4, which displays a theoretical framework, provided by Réquillart, Bouamra-Mechemache, & Jongeneel (2008, p.27). Their framework provides insight into the impacts of abolishing the quota in one go, but the underlying mechanisms remain the same in case of a quota expansion. The difference is that the consequences of a quota expansion will be more gradual and less pronounced than the consequences of an abrupt quota removal (Réquillart, Bouamra-Mechemache, & Jongeneel, 2008). For completeness, and because it will facilitate the explanation, both the consequences of quota introduction and quota removal will be discussed. Figure 2.4 Theoretical Impact on the Raw Milk Market

Source: Reprinted from ‘Economic analysis of the effects of the expiry of the EU milk quota system’, by V. Réquillart, Z. Bouamra-Mechemache, R. Jongeneel, 2008, Institut D'economie industrielle, p. 27. The left panel of Figure 2.4 shows the aggregate raw milk demand and supply (or marginal costs) curve, with and without the quota in place. In this context, aggregate can refer to a specific region or member state or to the EU as a whole. To concentrate solely on the impact of quota introduction and removal, other factors that could potentially influence the supply and demand curves are kept constant in this model. Furthermore, it is assumed that the super-levy is large enough to preclude any raw milk

(18)

18 production in excess of the milk quota. This is represented in the figure by the kinked, partly inelastic, supply curve. Note that the supply curve is kinked at a lower quantity than the free market equilibrium quantity, indicating that throughout this analysis, the overall quota is assumed to be binding (restricts aggregate production). As usual, the intersection of the linear supply and demand curve determines the perfect-competition market outcome, denoted by quantity Q1 and price P1. This is the market

outcome when there is no (government) intervention and profit-maximizing agents equate their marginal cost to their marginal revenue, i.e. if all farmers produce up to the point where their marginal costs are equal to the price of raw milk (Tonini & Domínguez, 2008).

The introduction of quota creates a departure from the perfect competition market outcome. By limiting total production to the quota quantity Q0, competition between buyers allows the quota

price P0 to rise above its free market equilibrium price P1. Phrased differently, the quota ensures that

the price is being kept artificially high, which creates winners and losers. Consumers lose, since they have to pay a higher price for their dairy products. Producers, on the other hand, gain, as they are allowed to produce less than without the quota, but receive a higher price for the raw milk they do produce. In this manner, the quota allows producers to generate a so-called ‘quota rent’, defined as the difference between the (relatively high) raw milk price under quota and their marginal costs of production. As long as the quota is binding, the quota rent is positive, which implies that even the last and therefore most costly unit produced is sold for more than its marginal costs. Figure 2.4 shows the ‘aggregate quota rent’, which is the difference between the raw milk price under quota P0 and the marginal costs of the marginal producer MC0. Note in Figure 2.4 that the aggregate quota rent keeps the producer surplus, defined as the area between the dotted line at the height of the free market price P1 and the supply curve, artificially high. The left panel of Figure 2.4 shows that by removing the quota - illustrated by the kinked supply curve becoming linear again - producers that were previously limited by the quota, increase production. This increases overall production from Q0 to Q1 and puts downward pressure on the price.

As a result, the price decreases from P0 to the free market price P1. As the quota is abolished, aggregate

quota rents disappear.

The right panel of Figure 2.4 shows the possibility that by removing the quota, raw milk production will not only increase, but that the raw milk production sector will also be subject to a restructuring process - illustrated by a downward shift of the supply curve. Similar to what happens in the left panel, production (that was previously limited by the quota) increases, which puts downward pressure on the price. However, at some point, this lower price can no longer cover the marginal costs of some relatively inefficient producers (who had low initial quota rents, or high marginal costs). Hence, the lower price forces them to decrease production. Relatively efficient producers on the other

(19)

19 hand (who had high initial quota rents, or low marginal costs) more than compensate for this lower production because they substantially increase production. Compared to the situation in which there is no restructuring of the sector, this redistribution of production from producers with high-costs to producers with low-costs results in a greater increase of production (from Q0 to Q2 instead of to Q1)

and a larger price decrease (from P0 to P2 instead of to P1). Note that this restructuring process can take

place at regional, member state or EU levels and depends on the difference between the initial quota rents of the various groups of producers. In sum, according to theory, three main results can be expected as a consequence of quota abolition: (1) a net increase of aggregate raw milk production, (2) a lower average raw milk price and (3) a redistribution of raw milk production from relatively inefficient producers to efficient producers.

2.5. The simulated impact of the quota expansion

Since 2003, the year in which it was announced that the quota would be abolished in 2015, several studies, often commissioned by the European Commission, started to investigate in advance how this would influence the raw milk price, raw milk production and the sales revenues of the dairy farmers of the EU as a whole and of the individual member states. These studies were all predictive in nature and hence forced to assume various ‘normalized’ conditions. Incidental fluctuations both on the supply and demand side of raw milk production were thus kept constant throughout these studies, whilst such exogenous factors probably have played an important role in determining the actual market outcome (Chantreuil et al., 2008). Nevertheless, this section discusses the spirit and main conclusions of these former studies.

Most approaches were based on partial equilibrium models, which covered either the whole agricultural sector (Binfield, Donnellan, Hanrahan, & Westhoff, 2007; Witzke & Tonini, 2009) or focussed exclusively on the dairy sector (Chantreuil F. , et al., 2008)11_{. The partial equilibrium models} relied on a multi-level approach taking into account the whole dairy chain. Nearly all models drew their information from the Eurostat database, often supplemented by national statistics and in order to provide an in-depth analysis, most models focussed not only on the market effects of policy changes at the EU level, but also zoomed in on particular regions or countries (Lips & Rieder, 2005; Colman et al., 2002; IPTS, 2009; Réquillart, Bouamra-Mechemache, & Jongeneel, 2008). The models differed from each other in terms of product and geographical coverage and the 11_{The main difference between partial and general equilibrium models is, that partial equilibrium models focus only on one} sector, in this case the milk sector, and do not take into account interactions with other commodity sectors. Amongst other things, partial equilibrium models are able to incorporate greater amounts of detail about production and policy instruments than their general equilibrium counterparts.

(20)

20

time span. However, more importantly, they differed in the way in which the policy change was defined, because at the time these studies were performed it was not yet known whether the quota were to be gradually expanded or removed in one go. As such, the consequences of the quota removal were investigated by taking into account different phasing-out scenarios. In general, the impacts of a ‘soft landing’ scenario, in which quota were increased by a few percent per year between 2009/10 and 2014/15 until they would fully expire in 2015, were compared with the impacts of a ‘hard landing’ scenario, in which quota were left untouched until 2014/15 after which they would be removed on 1 April 2015. The outcomes of these scenarios were subsequently compared to a baseline (or reference) scenario, which contained the status quo policy outlook for the agricultural sector over the next ten or fifteen years, without the quota abolition. As a result, rather than absolute price changes, price changes were generated relative to a baseline scenario. The comparison of the outcomes of the soft and hard landing scenarios with the baseline scenario was meant to isolate the (possible) effects of the quota abolition from other exogenous factors that influence the market outcome as well (IPTS, 2009; Réquillart, Bouamra-Mechemache, & Jongeneel, 2008).

The definition of none of the previously examined scenarios fully matches the ultimately implemented policy (as described in section 2.2). Nevertheless, to get an idea of the magnitude of the expected impacts, and to find out what determined the magnitude, it is still interesting to discuss the simulated outcomes of the examined scenarios that come closest to the actually implemented policy change. The results of five studies are listed in Table 2.3 and demonstrate the projected change in the aggregate EU-15 production of raw milk (△Q) and the average EU-15 price of raw milk (△P) as a result of a gradual quota abolition. Table 2.3 Simulated Price and Production Effects (Relative To a Baseline Scenario) Note: Depending of the intention of the study (oriented towards the short or medium term), the simulation results refer to either 2014 {Binfield, Donnellan, Hanrahan, & Westhoff (2007); Witzke & Tonini (2009)}, 2015 {Réquillart, Bouamra-Mechemache, & Jongeneel, (2008)}, or 2020 {IPTS, 2009; Chantreuil, Donnellan, Leeuwen, Salamon, Tabeau, & Bartova (2008)}. Source: Composed based on the results of the five studies of which the authors are shown in the table.

(21)

21 The direction of the price and production effects that are shown in Table 2.3 correspond to what was expected in theory: aggregate EU-15 raw milk production was projected to increase which would put downward pressure on the EU-15 raw milk prices. The magnitude of the projected production effect varied from 3,7% to 4,8% and the price effect from -3,8% to -10,3%. To interpret these outcomes correctly, recall that these percentages represent the change relative to what happens in the baseline scenario in which a continuation of the quota regime was assumed. For example, Réquillart, Bouamra-Mechemache and Jongeneel (2008) showed that the absolute price change would only be -3,8% compared to the actual price of 2008 (instead of the relative price change of -10,3% when compared to the price of the baseline scenario projected to 2015).

With regards to the magnitude of the change in production, all five studies projected that aggregate production would increase by less than allowed for by the quota expansion, because not all member states would fully utilize their quota expansion. In fact, the United Kingdom, Sweden and Finland were even projected to decrease their production in response to the lower average raw milk price (IPTS, 2009). As such, the theoretically predicted redistribution of production from less efficient to more efficient member states was confirmed by the model outcomes. In addition, the model of IPTS (2009) was equipped to project the impact of the gradual quota abolition on the sales revenues of dairy farmers of the EU-27. Because sales revenues are calculated by multiplying the raw milk production with the raw milk price, this impact depended on the direction and magnitude of the country-specific price and production effects. The projected impact on the sales revenues of the EU-27 was on the whole negative, but varied considerably across member states. In the model, highly competitive member states were able to increase their sales revenues by compensating the negative price effect with a positive production effect, while less competitive member states lost revenues both from price and quantity sides. The heterogeneous effect across member states helps to illuminate why the European Commission’s proposals to increase the milk quota was being firmly resisted by a few member states, who might have feared that (some of) their farmers would not be competitive enough to survive without the quota (Economist, 2008)12_.

The studies discussed above were all predictive in nature and therefore based on various assumptions. With regards to the importance of these assumptions, their sensitivity analyses showed that a small change in the assumed price elasticity of supply or the assumed height of the quota rents would substantially change the magnitude of the simulated production and price effect. The assumed price elasticity determined, in the model, the extent to which production was adjusted in reaction to a price change and the estimated quota rents per member state (defined as the difference between the price under quota and marginal costs of production) determined whether or not the quota were 12_{The reform met stiff resistance from France, Germany, Austria and Finland.}

(22)

22

still binding (limiting the production). This, in turn, determined the stimulus for farmers to increase production in response to expanded quota (IPTS, 2009). The sensitivity analysis of Réquillart, Bouamra-Mechemache and Jongeneel (2008) showed that by assuming 25% higher quota rents, or lower marginal costs, aggregate raw milk production after quota removal would be 9% higher than in the baseline scenario (versus +4,8% with the standard marginal cost assumption) and the average raw milk price would be 17% lower than in the baseline scenario (versus -10,3% with the standard marginal cost assumption). This illustrates that, if the quota rents per member state, which can be interpreted as a measure of the competitiveness of the farmers of that member state, were underestimated in the models, then the magnitude of the projected impacts presented in Table 2.3 will turn out to be an underestimation of the actual production and price effect. Recapitulative, it can be stated that the results of former studies are in line with what was expected in theory. Combining the theoretical expectations with the insights of these former studies, four main results can be expected as a consequence of the gradual quota abolition for the EU-15: (1) an increase of aggregate raw milk production, but with less than the overall quota increase allowed, (2) downward pressure on the raw milk prices, (3) a redistribution of raw milk production across member states, (4) a heterogeneous impact on the sales revenues across member states.

3. Methodology

3.1. Research question, sub-questions and method

While former studies have predicted the impact of the gradual quota abolition, this study aims to determine the actual impact so far. Since data relating to the production of 2015 has not yet been published, conclusions can only be drawn regarding the impact of the quota expansion up to 2014. Nevertheless, much of the ultimate impact will probably be captured, as it was expected that the impacts would occur predominantly during the quota expansion period itself and not so much afterwards (IPTS, 2009). The analysis will focus on the impact of the quota expansion on the sales revenues of dairy farmers of the EU-15. More specifically, the research question is as follows: What has been the impact up to 2014 of the widening of the milk quota on the raw milk sales revenues of the EU-15 as a whole and the individual member states? The approach chosen to answer this question comprises a descriptive data analysis. Data relating to the production and price of raw milk was collected from the website of Eurostat and the Milk Market Observatory (see section 3.3). With this data, the sales revenues of each EU-15 member state and of the EU-15 as a whole will be calculated, by multiplying for each member state the annual raw milk

(23)

23 production collected by dairies with their average annual raw milk price. The data with respect to the production, price and sales revenues will subsequently be used to construct figures and tables that will facilitate the analysis. To be able to substantiate the impact of the quota expansion on the sales revenues, the impact on the production and price will first be analysed separately. The challenge will be to separate out the impacts generated by the change in the quota regime from the impacts of exogenous factors, such as the economic crisis of 2008, the Russian boycott, changes in the input costs, weather conditions and so on. To reduce the risk that a change in the production of raw milk will be wrongly assessed as the result of the quota expansion, it will first be determined if, and to what extent, the quota were still binding during the quota expansion period for the various member states. Member states for which the quota was not binding, would have also been able to increase their production without the quota expansion. For these member states, the cause of any increased production cannot be attributed to the quota expansion. That is different for member states for which the quota was binding. Although there might have been exogenous factors that incentivized the farmers of these countries to increase their production, their increased production can be attributed to the quota expansion, simply because it would not have been possible without it. When the restrictiveness of the quota has been established, the production, price and sales revenues of the EU-15 as a whole will be analysed and explained on a yearly basis in order to capture the cause of certain fluctuations. Moreover, the average raw milk production, price and sales revenues of the EU-15 as a whole and the individual member states during the quota expansion period will be compared with those of the preceding period. This will allow for the identification of changes in trends rather than levels of variables. The quota expansion period has already been defined in section 2.3 as the period which runs from 1 April 2008 to 31 March 2015. The so-called ‘preceding period’ will here be defined as the period which runs from 1998 to 2007. A relatively long benchmark period is chosen to take account of the fact that during the preceding period, the price already contains a downward trend as a result of the policy changes from 2004 to 2008 (Jongeneel, Burrell, & Kavallari, 2011). The analysis will be carried out on the basis of the following five sub-questions (SQ): Production effect SQ 1 To what extent were the milk quota still binding during the expansion quota period for the EU-15 as a whole and for the individual member states?

SQ 2 To what extent has the widening of the milk quota influenced the aggregate raw milk production of the EU-15 as a whole?

(24)

24

SQ 3 To what extent has the widening of the milk quota influenced the aggregate raw milk production of the individual member states? Is there evidence of a redistribution of production over the various member states caused by the quota expansion? Price effect SQ 4 To what extent has the widening of the quota influenced the average raw milk price of the EU-15 and of the individual member states? Sales revenues effect SQ 5 To what extent has the widening of the quota influenced the sales revenues of dairy farmers of the EU-15 and of the individual member states? In section 4, each of these sub-questions will be answered in turn.

3.2. Geographical scope

The geographical scope of this study concerns 15 of the present 28 EU member states, or simply, the EU-15. The countries that will be examined are listed in the white area of Table 3.1. By omitting 13 EU member states from the analysis, only 20% of all EU raw milk producers are excluded, since 80% of the joint EU raw milk production comes from the EU-15 (Eurostat, http://appsso.eurostat.ec.europa.eu-/nui/setupDownloads.do). The reason for omitting these member states is that, during the period studied, their agricultural sectors might have been subject to additional or differentiated transition measures as agreed in their accession terms, because they joined the EU only recently (Jongeneel, Burrell, & Kavallari, 2011). Since it is difficult to separate out the impacts of the quota expansion from the more powerful trends set in motion by these accession processes, analysing the outcomes of these countries over the years could easily contain ‘noise’ from factors other than the change in the quota regime. Accessing countries are moreover excluded for practical reasons: some data, such as historical raw milk prices and quantities produced are not available for these member states.

(25)

25 Table 3.1 Included (White Area) and Excluded (Grey Area) Member States Note: Member states that will (not) be studied are depicted in the white (grey) area. Source: Constructed based on http://europa.eu/about-eu/countries/member-countries/.

3.3. Data collection

For the calculation of the sales revenues of the EU-15 member states, it was necessary to collect data regarding historical prices and production. The sources consulted for this purpose and the choices that have been made throughout the collection process are accounted for in this section. The historical monthly raw milk prices, from January 1977 to August 2015, of all EU-15 member states, have been downloaded from the website of the European Milk Market Observatory at http://ec.europa.eu/agriculture/milk-market-observatory/index_en.htm. The prices used throughout this study were last updated on 8 October 2015. The historical production per calendar year have been downloaded from the Eurostat database (http://ec.europa.eu/eurostat/data/database). Eurostat tracks the historical production of raw milk in two files, differing in the way in which raw milk production is defined. The file named “production and utilization of milk on the farm - annual data - (apro_mk_farm)”, represents total milk production, i.e. the sum of both raw milk produced for

(26)

26

dairies and raw milk produced for own use, such as for direct sales to final consumers, for the use in cattle feed, or for own consumption (http://appsso.eurostat.ec.europa.eu/nui/show.do). The file named “cow’s milk collection and products obtained - annual data - [apro_mk_cola]13_{”, on the other} hand, exclusively includes raw milk production that has been collected by dairies and thus covers only part of the total raw milk that is produced (http://appsso.eurostat.ec.europa.eu/nui/show.do). When analysing the development of raw milk production itself, data relating to the narrower definition of raw milk production was used, that is data about only the raw milk that was collected by dairies. This choice was made to accurately calculate sales revenues. It would have been incorrect to multiply production intended for own use with the prevailing market price of raw milk ‘as if’ it was sold to dairies, because raw milk indented for own use is not necessarily worth as much as raw milk sold to dairies. An exception was made when the quota were compared with production (to illustrate for which member states the quota was still binding). In that case, data regarding total raw milk production was used, because it would have been meaningless to compare the quota that applied to total production with only that part of total production that was collected by dairies. Missing data In the file named “production and utilization of milk on the farm - annual data - (apro_mk_farm)”, the production of Denmark in 2005 and Spain in 2014 was missing. Given the relevance of this data for the analyses of aggregate EU-15 production relative to the overall quota, and given the fact that is was also not available anywhere else, this data had to be estimated. Fortunately, the part of total raw milk production in these countries which was collected by dairies, was recorded for these years. Given that the difference between total raw milk production and raw milk collected by dairies (i.e. raw milk produced for personal use) was a fairly stable percentage of total raw milk production over the years, data relating to the amount of raw milk collected by dairies could be used to estimate total production. To calculate the estimated total production of Denmark in 2005 the following approach has been

taken. First the percentage of raw milk intended for personal use in the ten-year period before 2005, from 1995 to 2004, was calculated, which was on average 3,2%. Second, the production collected by dairies, 4451,4 tonnes, was multiplied by 1,032 (103,2%), resulting in an estimated total production of 4593,9 tonnes. Similarly, for Spain, the percentage of raw milk intended for personal use from 2004 to 2013 was calculated to be 8,8% on average, the production collected by dairies in 2014 was 6679,3 tonnes. Multiplying that by 1,088 resulted in an estimated total production of 72674 tonnes. 13_{Settings in both files: GEOàall EU-15 countries; TIMEàselect all; MILK ITEMàproducts obtained (1000 t);} MILK PRODUCTSàcows’ milk-total

(27)

27

4. Results and discussion

This section will determine the impact of the widening of the milk quota on the raw milk sales revenues of the EU-15 as a whole and the individual member states up to 2014 on the basis of five sub-questions. After each sub-question, the result expected based on the literature review, is briefly recounted. The sub-questions are structured in such a way that first the production will be analysed in section 4.1, then the price in section 4.2 and finally the sales revenues in section 4.3.

4.1. Production

SQ 1 To what extent were the milk quota still binding during the expansion quota period for the EU-15 as a whole and for the individual member states? While the (increased) quota were expected not to be binding anymore for the EU-15 as a whole, they were expected to be still binding for the member states that contained relatively competitive producers, in theory referred to as producers with positive quota rents or low marginal costs. To reduce the risk that a change in the production of raw milk will be wrongly assessed as the result of the quota expansion, the analysis will start with exploring to what extent the quota were still binding during the quota expansion period for the various member states. The restrictiveness of the quota will be determined on the basis of Table 4.1, which presents the annual percentage quota overrun or underuse per member state and for the EU-15 as a whole. A positive percentage means that the quota was overrun and hence binding (total production was equal to or more than the quota allowed) while a negative percentage, marked yellow, means that the quota was underused and hence not binding (total production was less than the quota allowed).

The percentages have been calculated as follows. First, the amount of quota underuse or overrun per member state was calculated by subtracting the total annual raw milk production of each member state from the height of their quota of that year14_{. Second, the amount of quota underuse or} overrun per member state was divided by the height of their quota and modified in such a way that it 14_{To construct Table 4.1, the production per calendar year (which runs from January to December) has been compared with} the quota that applied to the production per milk marketing year year (which runs from April to March). Although this is not optimal, it was the only option, because data relating to production was only recorded per calendar year whilst data relating to the quota was only recorded per milk marketing year. As a result, for example for the calendar year 2008, this means that the quantity of raw milk produced in 2008 has been compared with the quota that applied to 2008/09, while in fact, from January to March 2008 production still fell under the quota of 2007/08.

(28)

28 finally represented the annual quota underuse or overrun per member state expressed as a percentage of their national quota of that year. The same procedure has been followed for the calculation of the quota overrun or underuse of the EU-15 as a whole (presented in the last column of Table 4.1). Table 4.1 Percentage Quota Overrun or Underuse Note: The table shows the annual quota overrun or underuse for each member state (and for the EU-15 as a whole) expressed as a percentage of the national (or in case of the EU-15 aggregate) quota of that year. A yellow box indicates an unfilled quota (the country in question produced less than its national quota of that year permitted). A dash (-) indicates that information on the quota amount is missing. A colon (:) indicates that the country had no quota because it was not yet an EU member state in that year. DE = Germany; FR=France; IT=Italy; LU=Luxembourg; NL= Netherlands; BE=Belgium; IE=Ireland; UK= United Kingdom; EL= Greece; ES=Spain; PT=Portugal; FI=Finland; AT=Austria; SV=Sweden. Source: Constructed based on calculations with the data (relating to the quota) from Table 2.1 and data (relating to production) from the Eurostat database file named “production and utilization of milk on the farm - annual data - (apro_mk_farm) retrieved from http://appsso.eurostat.ec.europa.eu/nui/show.do. Total production has been used, as opposed to production collected by dairies, because it would have been meaningless to compare the quota, that applied to total production, with only that part of total production that was collected by dairies. It can be broadly observed from the table that in the first seven years after the quota were introduced there was no quota underuse, that is, the quota were binding for all member states. From 1991 to 2003, the quota were still binding for most member states. However, from 2004 to 2014, when it was known that the quota would be expanded and removed, an increasing tendency towards quota

(29)

29

underuse appeared. As a result, the overall EU-15 quota was no longer binding from 2007 to 2013. Looking at the relevance of the quota for individual member states, it can be seen that the quota of the Netherlands, Austria, Luxembourg, Denmark, Germany, Spain and Italy were binding throughout the entire quota expansion period. For Belgium and Ireland, they were not binding at the beginning of the quota expansion period, but they were at the end. This means that any increase of the raw milk production of these nine member states was only feasible thanks to the quota expansion. For France, Portugal, England, Greece, Finland and Sweden, however, the quota was not binding. The extent to which the quota were not binding differed considerably across countries. In the final quota year, Sweden, for example, was 18,4% under quota, while France was only 2,2% under quota. Hence, to a greater or lesser extent, these six member states had room to increase their production regardless of the quota expansion. Any increase of the production of these countries, can therefore not solely be attributed to the quota expansion15_.

SQ 2 To what extent has the widening of the milk quota influenced the aggregate raw milk production of the EU-15 as a whole? It was expected that aggregate EU-15 production would increase during the quota expansion period, but with less that the quota expansion allowed, because production was only expected to increase in the competitive member states for which the quota was still binding. To help determine the impact of the widening of the quota on raw milk production of the EU-15, Figure 4.1 shows the evolution of the sum of the raw milk production collected by dairies of all EU-15 member states together over the past twenty years. The percentages displayed above the bars specify the annual change of aggregate production relative to, and expressed as, a percentage of aggregate production of the year before. The development of aggregate EU-15 production during the quota expansion period will be analysed by comparing it with the level of production during the preceding period. 15_{Note that the conclusions that are drawn here with respect to the restrictiveness of the quota for the various member} states, does not necessarily hold for the individual farmers of these member states as well, since this depends on the national arrangements for the transfer of quota and for offsetting quota overrun with quota underuse between individual farms.

(30)

30 Figure 4.1 Aggregate EU-15 Production Collected By Dairies Note: The bars represent the sum of the production collected by dairies of the EU-15 member states. The percentages displayed above the bars specify the annual increase or decrease of aggregate production relative to and expressed as a percentage of aggregate production of the preceding year. The average increase depicted in the red box represents the percentage increase of average aggregate production during the quota expansion period relative to the preceding period. Source: Constructed based on calculations with data from the Eurostat database file named “cow’s milk collection and products obtained - annual data - [apro_mk_cola]”, retrieved from http://appsso.eurostat.ec.europa.eu/nui/show.do. The figure shows that from 1998 to 2007, aggregate production fluctuated at around 114.5 million tonnes. Thereafter, following the widened quota from 2008 to 2014, aggregate production increased consistently, except in 2009 and 2012. In those years, production growth stagnated temporary in response to relatively unfavourable raw milk prices (European Commission-EU FADN, 2014). Section 4.2 will provide more details with regards to the source of these price declines. What is striking is that in 2014, in anticipation of the full quota abolition on 1 April 2015, aggregate production increased by 4,6% in one year, up to about 127 million tonnes. Since the production of some member states expanded more than their 2014 quota allowed, either their farmers or the governments of these member states had to pay the substantial super-levy (ZuivelNL, 2015). Altogether, the impact is in line with the one anticipated. Aggregate production was on average 4,7% higher during the quota expansion period compared to the preceding period and increased by less than the increased quota had allowed for, because not all member states (fully) took advantage of their increased quota (see Table 4.1). Whether this production increase can be attributed to the quota expansion must become apparent from the analysis of the evolution of production at the member state level.

(31)

31

SQ 3 To what extent has the widening of the milk quota influenced the aggregate raw milk production of the individual member states? Is there evidence of a redistribution of production over the various member states caused by the quota expansion? While the aggregate production was expected to increase, at the same time, the production of the least competitive member states was expected to decrease in response to a lower raw milk price, that was expected to occur as a result of the increased aggregate production. These two effects were expected to result in a redistribution of production over the various member states. To help determine whether a redistribution of production over the various member states took place due to the quota expansion, Table 4.2 reveals the percentages with which the raw milk production collected by dairies per member state on average changed during the quota expansion period relative to the preceding period16_{. Before interpreting these results, recall that the quota of the Netherlands,} Austria, Luxembourg, Denmark, Germany, Spain, Italy, Belgium and Ireland, were still binding at the end of the quota expansion period, while the quota of France, Portugal, England, Greece, Finland and Sweden, were not. Table 4.2 Percentage Change in Average Raw Milk Production

Note: The last column shows the percentage change in the average raw milk

16_{Note that Table 4.1 concerns the development of total production relative to the total quota whilst Table 4.2 displays results} that concern the production collected by dairies only.

(32)

32

production collected by dairies from 2008 to 2014 compared with that from 1998 to 2007. Source: Constructed based on calculations with data from the Eurostat database file named “cow’s milk collection and products obtained - annual data - [apro_mk_cola]”, retrieved from http://appsso.eurostat.ec.europa.eu/nui/show.do.

The table shows that the raw milk production of the Netherlands, Austria, Luxembourg, Denmark, Germany, Belgium, Spain, Italy, France and Ireland was on average higher during the quota expansion period compared to the preceding period, whilst the production of the UK, Greece, Sweden and Finland was lower17_{. The production of Portugal remained more or less the same. In line with what was}

expected, these results prove that there has been a redistribution of production over the various member states during the quota expansion period. The question remains if this can be attributed to the quota expansion, which can be answered by taking into account whether the quota of the member states whose production increased were still binding at the end of the quota expansion period. By doing this, it appears that, with the exception of France, all member states who increased their production faced binding quota at the end of the quota expansion period and would thus not have been able to increase their production otherwise. Since the production of France was only 2,2% lower than its quota allowed in 2014, also the production of France would not have been able to increase by this much if there had been no quota expansion. Therefore, on the whole, it is plausible to point to the widening of the milk quota as the reason for the on average 4,7% higher aggregate EU-15 production during the quota expansion period. Whether the decreased production of the United Kingdom, Greece, Sweden and Finland can also be attributed to the quota expansion - because the corresponding increased aggregate production has exerted downward pressure on the price - must become apparent in section 4.2, in which the development of the raw milk price will be analysed.

4.2. Prices

SQ 4 To what extent has the widening of the quota influenced the average raw milk price of the EU-15 and of the individual member states? It was expected that the widening of the quota would cause production to increase, which would put downward pressure on the raw milk prices of the EU-15.

17_{That some member states have decreased their production during the quota expansion period, makes it even more} understandable that the production of the EU-15 as a whole increased by less than the increased quota allowed.

The impact of the widened milk quota on the sales revenues of dairy farmers of the EU-15