The Political Economy of the CAP Reform in Italy

Abstract

This chapter analyses the ex-ante socio-economic impact of the CAP reform on Italian agriculture and the whole economy using a micro-funded general equilibrium model which differentiates the impact at the household level. The political economy analysis of the consequences of the reform has clearly revealed the positions of farmers and agro-food industries, consumers, and farming unions concerning the issue of a total or partial implementation of decoupling. The policy analysis permits both an understanding of the possible social conflicts arising from the implementation of the reform and a unique ranking of the policy alternatives.

1 Introduction

This chapter investigates the ex-ante impact of the Mid Term Review of the Common Agricultural Policy (CAP) on Italian agriculture and describes the political economy aspects associated with the execution of the reform. Although the reduction of domestic farm supports may lead to a net gain in national economic welfare, some sectors and households can be adversely affected. Tracking the aggregate impacts down at the micro level is then crucial to understand the sources of political frictions that may hinder the process leading to the implementation of reforms. The ex-ante analysis of the possible causes of social conflicts, the identification of potential losers and the quantification of their losses may help designing accompanying policy actions making the reform politically feasible and enforcing the bargaining position of the institutions supporting the policy change. This motivation has markedly shaped the present research.

The effects of the CAP reform, in terms of producers, markets and levels of well-being of agricultural, rural and urban households, are first evaluated by using an applied general equilibrium model that permits to implement the CAP instruments by modeling the associated market failures, price rigidities and non-linearities. The general equilibrium results obtained with the MEG ISMEA model¹ are further elaborated in order to carry out the political economy analysis of the different scenarios.

The paper first illustrates the three policy alternative scenarios delineated by the Mid-Term Review of the CAP, one with full decoupling of aids and two with different options of partial decoupling. We then present the results of the simulations and their political economy interpretation aiming at ranking the policy scenarios accounting for the producers and consumers’ point of view and society’s changes in welfare.

2 The Mid Term Review and the Policy Scenarios

The present work analyses the impacts of the policy scenarios delineated in the Mid Term Review of the CAP as approved at the end of June 2003 in Luxembourg.² As it is well known, the aim of the reform is to substitute payments “coupled to specific farm activities”³ with a lump-sum payment which has no distortive effects in the markets and farmers’ allocation decisions. In essence, a price subsidy and/or an income subsidy coupled to a specific production is substituted with a decoupled income subsidy which in fact transfers support from the products to the producers. Farmers can thus optimize the activity portfolio according to the allocative information conveyed through the market, ensuring Pareto efficiency. The objective to decouple payments from specific farming activities is achieved while safeguarding agricultural incomes, by ensuring an income support as a single farm payment (SFP) representing a certain financial flow, which should help to keep farmers in business and to sustain the rural households’ livelihoods (De Filippis 2004).

The main objective of favoring greater orientation towards more and better markets is accompanied by other important objectives such as: favoring greater sustainability of agriculture; assuring more attention towards issues of food security and animal welfare, by asking farmers to sign contracts of environmental cross-compliance in exchange for public support; rising equity in distributing the support with respect to coupled payments which are mainly benefited by large producers; realizing more integration and synergy with rural development; obtaining administrative simplification.

The reform can be summarized in three main pillars (European Commission 2003a, b):

• 1. modifications of the market policies through variations of the intervention prices and/or variations of the existing premia or introduction of new premia for some products;

• 2. decoupling of the premia: decoupling introduces a single payment per farm starting from year 2005, whose amount equals the mean of the total direct payments received by the farm during the years 2000–2002, for some productions (cereals, protein crops, oil seeds, rice, livestock, sheep and goats and, from 2008, milk as well). The payment corresponding to the set-aside area in possess during the reference period is attributed separately;

• 3. modulation of the premia: all direct payments given to farmers (the single decoupled payment and specific coupled payments for durum wheat, protein crops, rice, fruits in shell, olive oil, tobacco) will be reduced in the period 2005–2012 in the proportion of 3% in 2005, 4% in 2006, and 5% from 2007 to 2012. Premia below EUR 5,000 are exempted.

The objective of the modulation, which is mandatory, is to transfer an amount of aids from the first pillar (market support) to the second pillar of the CAP (rural development). The single farm payment (SFP) is the mean of the payments received by the farm during the reference period 2000–02 for cereals; protein crops; oilseeds; rice; dried fodder; bovine meat; sheep and goats and, from 2008, milk. Permanent crops are not eligible. Further, there is a specific payment for the set-aside area. The SFP does not account for: (a) the quality premium for durum wheat, (b) the special premium for protein crops, and (c) part of the rice premium (EUR 453/ha). The eligible land has to be kept in good agronomic and environmental condition and is constrained not to produce fruit, vegetables and potatoes. If the reform is implemented regionally by giving a uniform rate per hectare, then farmers are free to produce any good with the exception of permanent crops.

The SFP that will be received in the future by the farm is composed by the per hectare SFP multiplied for the number of eligible hectares. Because of the link with eligible land, if a farmer reduces in one year the number of hectares, either by selling or renting some land in or out, the SFP will be proportionally reduced. It is then impossible to exert the right to the Single Payment without being in possess of at least one hectare of land.⁴, ⁵ On the other extreme, it is possible to produce nothing on the eligible land, if the land is maintained in “good agronomic and environmental conditions”.

The estimation of the transfers generated by the reform takes into consideration two impacts:

• the effects of the variations in the levels of intervention prices and premia in the involved Common Market Organizations (CMO), which modify the comparative advantage across agricultural activities and the absolute level of the premium;

• the non distortive effects of the decoupled lump-sum transfer which determines market-based reallocations of the activity portfolio of the farms.

The effects are microsimulated using farm level data to generate a base scenario depicting the situation of Agenda 2000 (scenario A).⁶ Because for some commodities the changes due to the reform are introduced gradually, the impacts of the scenarios are simulated referring to an abstract situation where the reform is fully implemented at year 2008.

The implementation of the Luxembourg agreement requires adjustments to the common market organization mainly for certain arable crops (cereals, oilseeds, protein crops), and dairy products.⁷ The adjustments to CMOs are summarized in Table 1. The adjustments in the CMOs, for the products which have been considered in the micro-simulation and in the MEG ISMEA model, are as follows:

Table 1 A summary of the mid term review policy changes—base situation 2001–2002 and full implementation

• 1. Cereals, oil seeds and set-aside: The direct payment of 63 EUR/ton is the same as for the base scenario.

• 2. Durum wheat: The base premium remains the same as for the base scenario; a reduction has been decided of the supplementary premium for the traditional areas from EUR 344.5/ha to EUR 285/ha in 2006, along with the elimination of the premium for normal areas; a quality premium of EUR 40/ha has been introduced in traditional production zones to farmers who are using certified seed of selected varieties within the limits of current Maximum Guaranteed Areas. We make the hypothesis that all farmers in traditional areas are eligible and access the quality incentive.

• 3. Protein crops: The base premium remains the same as for the base scenario, but the current special payment of EUR 9.5/ton is converted into a crop specific area payment of EUR 55.57/ha; with respect to the base scenario, considering the average historical yields in Italy the premium results to be slightly higher.

• 4. Rice: The intervention price for rice is reduced by 50% to EUR 150/ton and 88% compensation is provided through higher payments. The final compensation increases existing direct payments from EUR 52.65/ton to EUR 177/ton. Hence, for an average productivity of 6.04 tons/ha in Italy the premium is about 1070 EUR/ha, which is about three times the premium of the base scenario.

• 5. Milk and Butter: Dairy quotas are extended until the 2014/15 season. The intervention price of butter is reduced by 25%. The skimmed milk price is cut by 15%. As a compensation, it is introduced a premium of EUR 24.49/ton. Keeping also into account a uniform distribution of the national envelope on a per quota basis, the Italian premium is EUR 34.87/ton.

The policy microsimulation was performed using the farm budgets of the ISMEA socio-economic survey (ISMEA 2005) in the following steps:

• 1. Determination of the level of premia received by each farm of the ISMEA microdata considering the Agenda 2000 package and the 2001 situation in order to reconstruct the historical yields to compute the premia and the number of animals which effectively received a premium in the bovine meat CMO.

• 2. The changes in direct payments and variations in prices described above are then reported to the universe using the 2001 Census of Italian Agriculture and used to compute the SFPs on the basis of the estimated eligible land. The modulation is considered, that is the reduction by 5% of all premia when the reform reaches its full implementation regime.

The non-behavioral microsimulation exercise generates the sector-level impacts that are summarized in Table 2 according to the commodity disaggregation adopted in the MEG ISMEA model. The table also presents the changes in intervention and import prices introduced at the macro equilibrium level as a result of the adoption of the reform. For all other products, we do not consider changes in prices and premia.

Table 2 Level of micro-simulated direct payments and changes in intervention and import prices for the commodity disaggregation simulated in the general equilibrium model

Another important feature of the reform is the possibility given to the Member States (MSs) to partially adopt the decoupling regime. This possibility concerns only the application of the arable crops, cattle and sheep and goats regimes. In detail, for arable crops the MSs can choose to couple up to 25% of the base premium for or, alternatively, up to 40% of the supplementary premium for durum wheat.

For livestock, the MSs can choose to couple up to 50% of the actual premia for sheep and goats and up to 100% of the slaughter premium for calves and, further, one of the following alternatives: up to 100% of the suckler cows premium and up to 40% of the slaughter premium for adult bovines, or up to 100% of the slaughter premium for adult bovines, or up to 75% of the special male premium.

The Reform foresees also the constitution of national reserves by means of further percentage reductions of the premia, in addition to that coming from the modulation. The reserves are intended to permit the access to the activity to new farms, which are excluded by the SFP as they were not active in the reference period 2000–2. The decision about the premia cut for the constitution of the reserve is left to national governments. Other decisions for national governments concern the opportunity to cut part (up to 10%) of the crops, bovine meat and dairy premia and to use this amount of aids for special quality programs or to give incentive to specific productions in the same sectors. As no indication is available regarding the Italian decisions on these subjects at the moment of the simulations, we have not considered these options.

3 The General Equilibrium Model and Simulations’ Design

The MEG ISMEA model, which is described in detail in Finizia et al. (2005), is a static multisectoral computable general equilibrium model of the Italian economy with two different trade areas, the European Union (EU) and the rest of the world (RoW). The aim of this distinction is to take into account that the Italian agricultural policy is a European policy (OECD 1988; Gohin et al. 1999, 2002; Gohin 2002). Table 3 reports a summary description of the main features of the MEG ISMEA model.

Table 3 The structure of the MEG ISMEA general equilibrium model

The MEG ISMEA represents a Walrasian economy where all markets are perfectly competitive, firms maximize their profits, households maximize their utility and the production factors are remunerated on the basis of their marginal productivity. In this “ideal” economic environment some rigidities are introduced, in the goods and in the factors markets, in order to reproduce the main features of the Common Agricultural Policy (CAP) (Weyerbrock 1998; Hertel 1999; De Muro and Salvatici 2001; FAPRI 2003; ISMEA 2004).

The MEG ISMEA model includes 41 sectors and places particular emphasis on the agricultural and agri-food sectors. As shown in Table 4, agriculture is disaggregated into 23 agricultural sectors, food industry in 9 sectors, other industries in 7 sectors, and services in 2 sectors. Each sector produces a single output, using intermediate goods and primary factors: self-employed farm labor, hired labor, land (distinguished in three types), agricultural capital, and animals (distinguished in four types). The other sectors use two production factors: non-agricultural capital and labor.

Table 4 Sector definitions

MEG ISMEA considers 11 household types: 7 farm-household types describing the agricultural sector, 1 rural household type, and 3 urban low-middle-high income classes. The classification of the 7 farm-household types has been derived from the cluster analysis of the ISMEA 1995 Survey about the Socio-Economic Conditions of Italian Agriculture (ISMEA 2005). The social accounting matrix is further articulated into a rural class, and three urban classes graduated in terms of income levels. The information is derived from the Bank of Italy Income Survey. This classification permits an accurate distributional and welfare analysis of the impact of agricultural policies upon policy relevant farm-household types (ISMEA 2005).

The MEG ISMEA model builds on a Social Accounting Matrix (SAM) describing the economic relations between the structure of production and the income distribution across household classes. The SAM is based on the 1995 input-output table of the agri-food sector (ISMEA 1997). The input-output table is based on the data gathered in 1996 through two ad hoc surveys, the Survey on the Socio-economic Conditions of the Italian Agriculture and the Survey on the Economic Conditions of the Italian Food Industry.

We use the MEG ISMEA model to simulate the general equilibrium effects of the following policy scenarios:

• Scenario A (the base scenario): our ex-ante situation refers to the premia established by Agenda 2000 in its full implementation (period 2001–2). This information has been constructed using a non-behavioral model that has been updated from the 1995 to the 2001 situation and incorporated in our Agenda 2000 situation, as implemented in Italy in the years 2001 or 2002, depending on the products. This is the benchmark against which we evaluate the effects due to the introduction of the reform as described in the following scenarios.

• Scenario D1 (total decoupling, with modulation): we consider both decoupling and modulation. Modulation has been implemented by assuming that all the direct premia, both those which are part of the SFP and those coupled, are cut by 5% with the exemption of the first EUR 5,000.

• Scenarios D2A and D2B (partial decoupling): these scenarios are based on the possible options for the countries to maintain a proportion of payments “coupled” to specific commodities.⁸ In the scenarios names, the letters A and B refer respectively to:

  • i. the option of leaving 25% of the base payment of cereals coupled (scenario D2A);

  • ii. the option of leaving 40% of the supplementary payment per hectare of durum wheat coupled (scenario D2B).

The two scenarios should be considered “maximum” coupling options, where the minimum alternative is represented by the scenario D1 of complete application of the decoupling regime. In fact, each combination considers the maximum percentage of partial coupling admitted by the Regulation. Table 5 reports the percentages of decoupled premia generated by the reform for each product of the model, in the five scenarios. The political economy analysis of the different scenarios, presented in the next section, is carried out using the general equilibrium results obtained with the MEG ISMEA.

Table 5 The partial decoupling scenarios: percentages of decoupled premia in detail

4 The Political Economy of the CAP Reform in Italy

The political economy question aims at reconciling the different views of the actors involved in the decision making process into a unique social outcome. With this objective in mind, the scenarios have been ranked according to (a) the agricultural producers’ point of view as affected by the impact on value added and interested in production protection, (b) the general point of view of the agricultural and food industry, which includes other aspects besides valued added in agricultural production, and (c) the consumers and society’s point of view based on the impact of the reform on the consumer price index of the basket of food goods and on social welfare level.

We use the Borda voting rule to aggregate the individual or sector-specific preferences. We also measure the impact of the Mid Term Review on the distribution of incomes among the socio-economic groups of interest and the related effects on society’s welfare. The changes in welfare levels of each household class also influences the classes’ preference orderings with respect to the policy alternatives, and the equity—efficiency trade-off implied by them, and the prospect that political coalitions are formed thus affecting the distribution of political power and the policy ranking. We examine these issues in sequence.

According to the Borda voting rule each person reports his preference relation. Suppose that there are N alternatives. The highest ranked alternative is assigned a fixed point k _i . The alternative in the second preference place is assigned a smaller fixed point k _i −l _i , 0 < l _i < k _i for i = 1, …, N, a third place is assigned a yet smaller fixed point and so on to the last choice which is assigned l point. The sum of the weights gives the social preference ordering and the single best alternative. We assume that the Borda social decision function is incentive compatible, that is there are no incentives for strategic behavior by declaring false preferences, because in the present scheme there is only one voter.

The Borda aggregation method gives a rational collective preference but the outcome is not independent of irrelevant alternatives. As a consequence, the choice over the number of scenarios/candidates and the number of election outcomes to be aggregated, that is the control of the “agenda”, is of critical importance for determining the final collective preference. The voting mechanism is designed for one voter in the vests of a benevolent social observer. It runs in two rounds. In the first round of the elections, the benevolent social observer is asked to vote for J elections by ranking the N alternatives forming the set of alternatives A ₁ = {D1, D2A, D3B} as if each production sector were a separate industry in the economy in terms of the value added contribution of each sector.

The voting rule ranking the alternative reform scenarios assigns a higher vote to the highest positive percentage change and the lowest percentage change in value added. This is intended to reflect the producers’ interest in maximizing profits from agriculture and protecting agricultural production in general. The least preferred gets 0 points, then the sequence increases by equal increments of 1 until N. The weights need not to be equally spaced. The ex aequo outcome is attributed when the differences are within the range [−0.3, 0.3] and receives the lowest vote. In other words, ties are not counted in the sum. In the second round, the social observer is asked to produce a social rule based on an objective weighting scheme reflecting the relative “importance” of the scenarios based on the value added share contributed by each sector. The weighting scheme changes the equal spacing rule of the votes cast in the first round. The intersectoral aggregation is the weighted sum of the Borda votes, which gives a unique voting outcome revealing the most preferred scenario from the producers’ point of view.

The voting procedure can be summarized as follows:

The election

1 voter being a benevolent social observer

i = 1, …, N alternatives with N = 3

j = 1, …, J elections (one for each agricultural sector, J = 23)

The voting mechanism

Round 1—Vote for the best scenario per each sector

Round 2—Aggregate each vote using objective weights

4.1 The Producers’ Interests

The voting outcome of both the first and second round of elections is presented in Table 6. The simulated changes determined with the MEG ISMEA in the production levels for each agricultural sector are presented in the first three columns of Table 6. The general equilibrium results show that the reform induces marked productive reallocations from cereal crops to fodder. The effect is particularly unfavorable for soft and durum wheat (respectively −27.64% and −36.11%), soy-bean (−80.67%) and other industrial crops (−20.68%). Soft wheat is also less competitive. Vice versa, livestock production is slightly encouraged from the cost reduction, given the higher availability of forage (and consequent cost reduction), with the exception of sheep and goats which are typically raised on extensive agricultural areas.

Table 6 The agricultural producers point of view

The outcome of the first round of the voting where all sectors have the same importance weight gives the total decoupling scheme D1 as the winner. The Borda score is 23 as compared to 13 for the partially decoupled scheme D2A and 8 for the D2B scheme. The results of the second round of elections can be read in the last row of the last three columns. The weights used to account for the different contribution of each sector to the agricultural value added are shown in the seventh column. The aggregation rule incorporating the weighting scheme preserves the same preference ordering.

4.2 The General Interest of the Agricultural and Food Industry

Table 7 proposes a more enlarged view, which includes the general interest of the agricultural and food industry and of other sectors related to agriculture such as the chemical sector. From this wider perspective, it is not just the performance of the single sectors that is important but other factors such as the size of the trade deficit of agri-food products, the impact on land prices, the changes in both farm and non-farm labor employment and the impact on income levels are of primary importance.

Table 7 The agricultural and food industry point of view

It is interesting to note that while the outcome of the voting procedure for the producers’ point of view depends only upon the choice of the voting rule, the outcome of the more general interest at the industry level depends also on the choice of the weights, that are now subjective, and the “agenda setting” which selects the number and type of elections. While the agenda setting is less of a problem regarding the agricultural producers’ view because the number of elections corresponds to the number of sectors included in the model, in the more enlarged view incorporating also the preferences of the agricultural and food industry, the selection of the number of elections is critical.

The subjective weights are assigned according to the following “conformity rule” based on the degree of proximity of a sector outcome to the objectives of the reform. We summarize the reform objectives as follows: (a) greater market orientation and efficiency, (b) income maintenance and employment, (c) low factor use where, in general, extensive choices are preferred to intensive choices, (d) low environmental impact, (e) sustainability of agriculture and incentives for rural development, (f) fairness in the distribution of the level of support. Based on the subjective evaluation of the social observer we assign a conformity score on the basis of a low (0 score), medium (0.5 score) and high (1 score) level of conformity.

The weighted outcome is presented in the right corner of Table 7, which reports the subjective scores assigned to each item of the agenda in the first column. The order of preference ranks total decoupling (D1) first and the partial decoupling scheme D2A as more preferred to the D2B scheme. Interestingly, the conformity weights change the preference ordering of the non-weighted count.

4.3 The Consumers’ Interests

In general, consumers are worried about price instability and the impact of policy changes on the level of the consumption price index. Recently, agricultural and food products have been often blamed to be the main responsible for inflationary pressures. This situation justifies the growing public concern for the impact of the Mid Term Review on both the level and variability of the primary commodities composing the food basket and the associated impact of the consumer price index for food products on the overall level of the consumer price.

Table 8 reports the composition of the food basket as derived from the ISTAT Consumer Expenditure Survey for the base year 2001 for the household classes included in the general equilibrium model. The food budget shares are the weights used to compute the change in consumer price index and its variability. As it is apparent by inspecting the overall results, the reform has an impact that may have an economic interest only in the milk sector but the overall impact on both the levels of the consumer index and its variability is negligible. It follows that post-reform pressures on the consumption price index should not be imputed to the agricultural reform.

Table 8 The consumers point of view

Also for consumers, the order of preference ranks total decoupling (D1) first and the partial decoupling scheme D2A as more preferred to the D2B scheme.

4.4 Social Welfare, Income Distribution and the Equity-Efficiency Trade-Off

The outcomes of the different policy scenarios affect the distribution of income among socio-economic groups and the level of social welfare. As expected due to the surgical nature of the reform that limits most of the changes to the agricultural sector and the related industries, changes in income are restricted mainly to farm-households (Table 9). In fact, the incidence of the effects varies among farm-household types. The groups experiencing the highest rise in real income are the professional medium-size, large and very large farm households.

Table 9 Social welfare rankings

The change in relative net output and input prices affects the distribution of value added between sectors and, within sectors, the distribution of value added between wages and rents. These changes, along with changes in the cost of living and lump-sum transfers in the form of SFPs associated with the reform, are responsible for the distribution of income among the household types. The magnitude of farm income changes depend on the size of the elasticity of substitution between labor, capital and land and the intensity of the factor uses due to the post-reform changes in output and factor prices and the size of the lump-sum transfer which is associated with the distribution of rights at the reference situation.

The impact of the different reform scenarios on the distribution of income of the overall society has been measured using Gini coefficients. As shown in Table 9, at the society level, where about 96% of the households are non-agricultural, the differential impact of the reform scenarios on the income distribution is indiscernible. The Gini index of 0.674 is not affected by the reform. The level of inequality for the agricultural society is much lower (0.371) as it is reasonable to expect for a relatively more homogeneous segment of society. As before, it does not vary across scenarios.

The effect on inequality is not the sole dimension of interest in ranking income distributions. It is in general of interest to combine the evaluation with considerations about efficiency as described, in the present context, by changes in society’s average level of income. The social evaluation function that we choose to rank any pair of income distributions is in fact a function that aggregates both a concern for efficiency, as represented by the mean of the income distribution, and a concern for equity, as described by an index of inequality or dispersion of the income distribution:

$$ W(x) = V(\mu ,I) = \mu^{ - G} $$

where $\mu (x) = \sum\nolimits_{i = 1}^{N} {x_{i} } /N$ with N being the number of household classes, $I = I(x)$ is an index of inequality of the distribution of income x such as the Gini coefficient, and V is a function increasing in its first argument but decreasing in the second argument. Lambert (1989) terms this social evaluation function as the abbreviated social welfare function. As the last row of Table 9 shows, the welfare level of the Italian society, incorporating both a concern for equity and efficiency, is not affected by the reform.

4.5 Social Conflicts and the Distribution of Political Power

Different agricultural reform schemes have a significant impact on the distribution of welfare levels especially, as it is rational to expect, within the farming sector. This affects the distribution of political power among the interest groups representing the different farm-household types. Are there conflicts among society? How does political power affect decisions? How will the political bargaining weight of the groups of gainers and losers affect the final policy outcome?

To investigate these questions, we use the Pareto criterion to rank the policy scenarios according to the preferences of each household class, and then inquire whether there are common interests across household classes that can be grouped. This process may identify the existence of possible class conflicts among coalitions representing the interests of the groups.

Pareto optimality ranks possible outcomes (economic states) by constructing a preference ordering among the elements of the choice set using the binary relation xRy stating that “welfare at state x is at least as high as welfare at state y”. The ordinal preference relation R is complete and transitive and says nothing about the intensity, or cardinality, of the preferences. A strict preference is indicated as xPy; an indifference situation is indicated as xIy. According to the Pareto principle, the economic state x is Pareto superior to state y if xR _j y for all agent j and xP _j y for at least one agent j. In the context of the present social experiment, every household class is at least satisfied with the outcome of policy scenario y and x and at least one household class is strictly better off with x.

Inspection of Table 10, reporting the changes in welfare levels with respect to the base scenario per each policy alternative under consideration, reveals that it is not possible to establish a unique ranking across scenarios because there is at least one class that is worse off with respect to one of the binary comparisons of interest. However, some classes of households show a consistent preference ranking across scenarios. The limited resources, retirement, residential lifestyle, small farms rank D2 B.P.D2A.P.D1 as shown in the no-shadow area in Table 10. The medium size, large and very large farm-households consistently rank D1.P.D2 B.P.D2A. The urban and rural households, the dark shadow area in Table 10, do not consider the agricultural reform as a political issue of interest as a consequence of their revealed indifference D2B.I.D2A.I.D1 to the different policy scenarios.

Table 10 Pareto rankings

We then assume that the less professional classes of farm-households (limited resources, retirement, residential lifestyle, small farms) form a coalition kept together by the common interests of preferring the D2B partial decoupling scenario to the D2A scheme and total decoupling D1 that we term the “small farm coalition.” On the other hand, the professional agriculture (the medium size, large and very large farm-households) coalesce to form the “large farm coalition.”

Under a political economy perspective, it is interesting to inquire whether the different interest groups have same bargaining power, as reflected by different political weights, to the point that one of the coalitions dominates the policy arena. Does the choice of political weights affect the Pareto ranking?

We assume that these interest groups know the level of gain or losses that is going to occur, that it has perfect knowledge of Table 10 gains and losses with respect to the base scenario, and that the coalitions are self-interested groups. Further, the intensity with which any group cares about a given policy change is proportional to the relative difference in welfare levels between the three alternative policy scenarios. We also define two weights describing the likely political importance of the two coalitions. The population share weight is based on the number of the farm-households entering each coalition in line with the one person, one vote paradigm; the value added weight is defined in terms of the value added contribution of each farm-household class.

Table 11 shows that the preference rankings are not affected by the different bargaining power of the two coalitions as captured by the population and value added weight. Comparing the differences in weighted welfare levels at the coalition level, it is reasonable to expect a more intense political action capable to dominate the policy arena from the coalition of the professional farmers who would enjoy a much larger gain in welfare by pursuing the total decoupling scheme (0.3 or 0.5 % change depending on the political weight) as compared to the less professional farmers who are expected to have a weaker motivation to pursue their own interest due to the small expected welfare gains from the reform.

Table 11 Ranking and political power

5 Conclusions

The analysis of the impact of the CAP reform on Italian agriculture and the whole economy has been carried out within a micro-funded general equilibrium model capable of differentiating the impact by household type of policy concern. The political economy analysis of the ex ante impact of the reform on the interests of the society’s groups has revealed the following positions about the issue of a total or partial implementation of decoupling:

• the producers and agro-food industry’s interests: both producers, which give each agricultural sector a different importance based on the value added, and the agro-food industry, which weights the industry activities on the basis of the conformity of the impact with the goals of the reform, rank the total decoupling scheme as the most preferred;

• the consumers’ interests: the overall impact of the reform on the consumer price index for food products is negligible. As a consequence, potential post-reform pressures on the consumption price index should not be imputed to the agricultural reform;

• the society’s interests—the level of inequality and social welfare of the Italian society, incorporating both a concern for equity and efficiency, is not significantly affected by the reform;

• the farming unions’ interests: based on the impact of the reform on the welfare levels of the Italian farm-household types, the small less professional farms prefer partial to total decoupling. Professional farm-household types invert the ranking. The urban and rural households are indifferent with respect to the marginal impact of the reform on their levels of well-being. In general, it is reasonable to expect a more intense political action from the coalition representing the interests of the professional farmers who would enjoy a much larger gain in welfare by pursuing the total decoupling scheme.

The implementation of a totally decoupled reform gives back to the market both the allocative and the redistributive function thus favoring greater efficiency in the use of resources in activities and areas of greater comparative advantage. Income levels of farming households are maintained by granting a non distortive lump-sum corresponding to the amount of premia received in the reference situation of year 2001–2002. In general, a totally decoupled scheme would mitigate the problem of distributive justice associated with coupled payments which, by design, benefit mainly the large producers.

The adjustment process induced by the reform may encourage farmers to adopt least cost practices and activities with the objective of minimizing the use of labor and other inputs in agriculture. The increase in pasture production at the expenses of durum wheat in the Italian south is an example of such a change. This modification of the activity portfolio does not lead to an exit from the agricultural industry, but induces the rational adoption of cost-efficient activities and the abandonment of activities, such as durum wheat, that, without the coupled premium, do not cover operating costs in the less efficient farms. This type of change, that we term “disactivation,” releases resources which can be employed more efficiently in other sectors of the economy. Agricultural surplus labor may give rise to unemployment, especially in the south, where employment opportunities lack. Lower demand for agricultural inputs coupled with higher costs of the chemical industry has a positive impact on the environment. Higher land prices are expected to curb transactions of land properties but may activate the rental market for land. The land market may also suffer from legal conflicts due to the unclear definition of property and rental rights in the reference situation leading to higher transaction costs.

The adoption of a partially decoupled scheme would reveal greater society’s aversion to inequality in recognition of the fact that most of the benefits would accrue to non professional farmers. The evaluation of the pros and cons associated with the adoption of a partially decoupled scheme shows that the benefits would not be sufficient to mitigate the marked structural adjustments associated with the totally decoupled choice, especially in the cereal and sheep and goat production and as a consequence of the “disactivation” process, and would cause a loss of efficiency for the entire sector. Further, a partially decoupled solution has no significant effects in the livestock industry whose productivity is sensitive to market conditions and to the opportunity to gain from the reduced costs of feeding as it can be expected as a consequence of the greater availability of fodder.

In general, an obstacle to reforms is represented by the real or presumed costs of the adjustment imposed on farmers. The New Zealand experience, where in the past decade a market oriented reform without income compensations has been undertaken, teaches that farmers’ incomes and the agricultural industry in general, recovered promptly from the initial shocks of de-regulation also thanks to other reforms in connected markets and outside of agriculture and the related general equilibrium effects (Rae et al. 2003).

The reform forces a change in the professional farmers’ mentality who, despite the larger financial possibilities generated by the single farm payment, have to make production decisions without counting on the previously guaranteed returns stemming from each single activity. As a consequence, the post-reform marketing strategies have to take into account the changed competitive environment, the characteristics of the demand for their products, their competitive advantages and the special strengths of each farm organization.

The reform also imposes a “cultural” change in the quality of the Italian agricultural policy product towards greater market orientation which would foster a restructuring process in favor of better products, more efficient and competitive industries and a more effective integration between agricultural and rural policy. The push towards greater exclusion of the farms already at the margins of agriculture, especially in the South, is not so strong thanks to the single farm payment. It should be remarked, however, that these “less professional” farms are not the main object of interest of agricultural policies, but, more properly, of rural policies, which, curiously not enough, can be financed by the modulation of agricultural policy. What is relevant is then the “coupling” of agricultural with rural policies.

The reform will then be an opportunity rather than a problem, if State and Regions will be using in a modern way market policies that activate (a) the land market in order to favor the consolidation of those farms going out of market, (b) the insurance market and (c) the financial market. This action concerns mainly agricultural policies. If central and regional governments will also intervene by targeting non-professional farms, which do not fully benefit from the reform, by adopting effective rural policies, then, equity cum efficiency is a concrete objective.

Appendix

See Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 11.