Introduction

The European Union’s common agricultural policy (CAP) is a policy which is now well established as being among those with the biggest influences on the EU’s natural environment, as well as a large source of funding for environmentally friendly agricultural practices (Vogeler 2022). It is a dynamic policy that has developed over time to include social and environmental issues in addition to its original production and income focus (EU 2021). In the current programming period (2014–2022), environmental measures can roughly be divided into compulsory ones pre-conditioning direct aids from the first pillar of the policy, which are relatively uniform across the EU, and voluntary agri-environmental measures in the second pillar. Arguably, the latter are the most effective and ambitious from an environmental perspective, as the second pillar is subject to strategic planning. However, both measures in the first pillar and agri-environmental schemes have variable ambition and success across member states (MS) and regions (Batáry et al. 2015; European Court of Auditors 2011, 2017; Le Roux 2008; Poláková et al. 2011; van Vliet et al. 2015). Namely, despite falling under a nominally common policy, there is considerable leeway for MS (and their regions) in programming the interventions, i.e. determining their needs, selecting measures and allocating funds, especially in Pillar II; in the next programming period (2023–2027), this freedom will be further strengthened due to the expansion of strategic planning to the entire policy after 2023 (EU 2021). MS and regions thus exhibit different levels of ambition in terms of how stringent and targeted the measures are and how much money is allocated to agri-environmental measures (Alliance Environnement 2021; Bartolini et al. 2021; European Commission and DG AGRI 2017; European Court of Auditors 2011; Poláková et al. 2011). Furthermore, the uptake of voluntary measures is affected, inter alia, by farmers’ (as a central CAP stakeholder) value systems with regards to environmental issues (e.g. Baur et al. 2016; Leonhardt et al. 2021; Science for Environment Policy 2017; Stupak et al. 2019).

If better performance in the much-criticised environmental parts of the CAP is to be achieved, it can be helpful to understand the reasons behind these differences (cf. Leduc et al. 2021). As perceptions and value systems affect policy implementation (e.g. Tasic 2011; Cerna 2013; Viscusi and Gayer 2015), this study aims to explain the issue of differential implementation and uptake by showing that there are significant differences between different policy levels (i.e. EU vs. MS and regions) in the construction of agriculture and the environment under the CAP. This includes the notions, priorities, and value systems regarding agriculture, the environment and the relationship between them.

Despite the increasing importance of environmental considerations under the CAP, there has been little research done on these constructions (with some exceptions, e.g. Gallet 2012; Vogeler 2022). To fill this gap, a novel theoretical framework is proposed, detailing what is seen as distinct agri-environmental discourses adopted by actors at different levels of EU policymaking and their defining characteristics. In building this framework, two established bodies of knowledge on discourses which seem to have been kept separate so far were integrated: the studies of CAP discourses (Alons and Zwaan 2016; Erjavec and Erjavec 2009; Erjavec et al. 2009; Garzon 2007; Potter and Burney 2002; Potter and Tilzey 2005; Rutz et al. 2014; Skogstad 1998; Tilzey and Potter 2007) and the broad study of environmental discourses (Dryzek 2008, 2013; Feindt and Oels 2005; Hajer 1995).

To capture the differences between the understanding of the relationship between nature and farming at different levels and consequently compare the constructs and values of the policy communities behind agricultural policy, their discursive construction in the main relevant strategic agricultural policy documents of the EU and MS, which include the key elements of EU agri-environmental policy, was analysed. While these documents were selected to be as representative as possible of the diversity of situations across the EU at different policy levels, the discourses constructed here can most likely not be considered exhaustive but as a starting point for further discussion and analysis.

Insight from an integrated analysis of the construction of agricultural and environmental issues and their interplay could be helpful to policymakers in better designing future agri-environmental policies by revealing potentially fundamental ideological and socio-economic hurdles preventing higher environmental ambition in the respective policy communities (cf. Burton et al. 2008; Plumecocq et al. 2018). Furthermore, they can provide insight into the social relations and current power structures within the policy community (Kivle and Espedal 2022) and consequently an indication of the likely success of environmental measures.

To date, the majority of discourse-analytic work done on the CAP has been on the CAP as a whole, i.e., at the EU level, while little work has been done to explore the differences between (agri-) environmental policy discourses between different MS and/or regions (Leipold et al. 2019; but see Leduc et al. 2021). Therefore, in this paper both constructions at the EU level and what are supposed to be its emanations at the MS/regional level (Rural development programmes) for the programming period 2014–2020 were analysed. The hypothesis is that there are discernible differences at different policy levels in terms of how agriculture and the environment are conceptualised.

The paper is structured as follows. In the next section, the theoretical framework is presented, which builds on the integration of CAP and environmental discourses. In the Methodology section, the approach to analysis and the texts used are presented. This is followed by the Results section, where the main conceptualisations of farming and nature in each of the analysed texts are described. In the Discussion, we aim to draw parallels between different texts and embed this in existing literature. We finally draw some conclusions.

Theoretical framework

Discourse represents a crucial resource that can be mobilised by actors for coordination and legitimisation purposes or to achieve acceptance of a change (Fairclough 2002). Policy discourse can direct perceptions of a problem, and foster acceptance of policy measures (cf. Garzon 2007). Power relations are inherent in discourses, as the value judgments, norms and perceptions that constitute discourses inevitably prioritise some interests over others (Dryzek 2008). It is important for policy-making to assert certain interests through the dominance of a particular discourse (Goldstein et al. 1993); and if the discourse is institutionally embedded, it can have an enduring impact on policy (Skogstad 1998).

Texts, as the main elements of discourse, are sites where the power differences of policy actors in asserting the dominance of their own discourse become evident (Wodak and Meyer 2009). As agricultural policy permeates a number of other policy fields (relating not only to agriculture but also to other economic as well as social and environmental issues), its texts are a complex site of interaction, struggle and attempted reconciliation between different worldviews.

In line with the sociocognitive approach (Van Dijk 2017), policy texts are seen as ‘communicative events’ whose authors ‘rely mainly upon collective frames of perceptions’, or ‘social representations’ (Moscovici 1982; Wodak and Meyer 2009), substantiating proposed policies to recipients of the analysed policy texts. Due to the fact that CAP policy documents are simultaneously the result of broad stakeholder consultation and documents intended to legitimise policy to the public and stakeholders, their content also reflects which constructions of nature and farming are or are not acceptable in a certain policy setting (cf. Leipold et al. 2019; Ruiz 2009; van Dijk 1993).

Discourses are ‘systems of meaning production rather than simple statements or language, systems that fix meaning, however temporarily, and enable us to make sense of the world’ (Sheperd 2008, p. 10). All discourses are value-based, although some more explicitly than others. Defining values as what is ‘worth having, doing and being’ (Selznick 1992, p. 60), values can be identified through discourses. Or, in other words: values can be identified as sayings and doings of subjects in different contexts that articulate and accomplish what is normatively right and wrong, good or bad, for its own sake (Gehman et al. 2013, p. 84).

Values in discourse can be studied in different ways. Thus, different analytical techniques can be used to show whether and how values circulate in a text. By analysing texts and discourse, researchers can discover how patterns in texts and argumentation either strengthen or weaken the values in question. The best-known method for identifying values is keyword analysis. In addition to analysing the keywords used, linguistic objects (e.g. exclamation points) can also be analysed to build theories about discourse. In discourse analysis, one can look for words that weaken or strengthen meaning (such as 'like', 'in a way' or 'maybe'). It is also possible to analyse the use of modal auxiliary verbs (must, can or will). In addition, one can ask what norms determine the tone and content of the text.

The next sections presents the two underlying theoretical frameworks (CAP discourses and Environmental discourses) and our own, which builds on them.

CAP discourses

It is relatively well-established in the literature on the CAP that there are three main basic discourses competing for dominance in the CAP, i.e. the exceptionalist, the neo-liberal and the multifunctional. This distinction was initially proposed in a comprehensive manner by Potter and Tilzey (2005) and has been built upon extensively by subsequent work (e.g. Dibden et al. 2009; Erjavec et al. 2009; Rutz et al. 2014; Tilzey and Potter 2007).

The discourse of ‘Agricultural exceptionalism’ (also referred to as ‘Neo-mercantilism’, ‘Productivism’ or the state-assisted discourse) presents agriculture as a sector whose main task is to produce enough (safe) food to feed the population, standing apart from other economic activities due to being exposed to the whims of nature and imperfect markets (Skogstad 1998; Wilson 2001; Muirhead and Almås 2012). Both in Europe and elsewhere, farming was long accorded a special status as part of a productivist ideology that held that farmers safeguard a country’s strategic goals, most notably food security. This constituted the basis for what is referred to as agricultural exceptionalism and was translated into the need for high levels of budgetary support at the policy level (Muirhead and Almås 2012; Burton and Wilson 2012). Due to this special position, agriculture was long excluded from international trade negotiations; this discourse is thus often also associated with protectionist tendencies.

The discourse of ‘Neo-liberalism’ (or competitive, market-liberal discourse) argues that agriculture is an economic sector like any other in which the farmer should be treated as an entrepreneur and market forces should take precedence over state intervention (Coleman 1996; Skogstad 1998; Alons 2017). The underlying theoretical premise of (economic) liberalism is that markets are more efficient in allocating resources than government interventions and that the liberalisation of markets and trade ultimately results in higher total welfare. The notion that agriculture is just another economic sector first gained traction during the Uruguay round of GATT negotiations (1986–1993), which for the first time included agriculture, stripping it of its special status after the USA started abandoning agricultural exceptionalism (Skogstad 1998).

The discourse of ‘Multifunctionality’ (or ‘Post-productivism’), first applied to agriculture during the Uruguay trade negotiations (Garzon 2007), assumes that rural areas, in addition to simply producing food (and jointly with its production), deliver a number of other public goods and services (e.g. Potter and Burney 2002; Potter and Tilzey 2007; Midgley and Renwick 2012) such as maintaining the landscape, rural vitality and farmland biodiversity. By extension, the CAP was argued to be indispensable in providing income to disadvantaged rural areas, especially to small farmers (Grady and Macmillan 1999). It was viewed by critics as a concept that EU policymakers and trade negotiators invoked to resist pressures to liberalise and thus contest the neoliberal paradigm in world food trade (Bohman et al. 1999; Almås and Campbell 2012; Midgley and Renwick 2012), but was legitimised to a large extent by the OECD (OECD 2001) and other countries with high budgetary support to agriculture that adopted the concept. The meaning attributed to the concept of multifunctionality in various debates is ambiguous, and it has been used by different actors differently according to their own agendas and interpretations (Renting et al. 2009).

Environmental discourses

Arguably, the most comprehensive categorisation of environmental discourses is John Dryzek’s (1997, 2005, 2013, 2022); in his seminal work The Politics of the Earth, he classified environmental discourses based on whether they are radical or reformist (to what extent the discourse challenges the existing social order of industrialism) and imaginative or prosaic, resulting in four major groups of discourses: Limits and survival (radical, prosaic), Environmental problem solving (reformist, prosaic), Sustainability (reformist, imaginative) and Green radicalism (radical, imaginative). Dryzek further divides these major groups of discourses into subgroups when there are distinctive contextual elements that merit such further sub-categorization.

The radical but prosaic discourse of ‘Survivalism’ (or ‘Limits to growth’), emerging from the environmental movement of the ‘60s and heavily influenced by the publication ‘Limits to Growth’ by the Club of Rome (Meadows et al. 1972), focuses on planetary boundaries and carrying capacity and challenges both perpetual economic growth and existing power relations. However, it is considered prosaic in that it proposes solutions within the remit of industrialism (e.g., more administrative control and science-based decision-making). Directly opposed to it is the historically dominant ‘no limits’ ‘Promethean’ industrialist discourse, which sees nature as a (limitless) pool of resources. In its most extreme form, this discourse denies environmental issues, limits or resource scarcity; when these are acknowledged, adherents believe in human ingenuity and technological progress as able to overcome any natural constraints to growth and development (Dryzek 2013).

Discourses in the reformist and prosaic category of ‘Environmental problem solving’, as the designation suggests, all acknowledge environmental issues and that some, though not radical, change in society should be made to address them; they are differentiated based on who is seen as the central agent to be placed in control of environmental policies: (a) experts and the bureaucracy through administrative procedures such as strategic environmental assessment and regulation (‘Administrative rationalism’): (b) citizens reconciling priorities in deliberative procedures (‘Democratic pragmatism’); or (c) the market by communicating consumers’ demand for environmental protection through price mechanisms complemented by governments intervening to compensate for imperfect market mechanisms (‘Economic rationalism’).

The two reformist and imaginative discourses of Sustainability—‘Sustainable development’ and ‘Ecological modernization’ both aim to ‘dissolve the conflicts between environmental and economic values’. They acknowledge the notion that some change must be instituted, though again it need not be fundamental—economic development and growth remain as overarching goals to be achieved. While the discourse of sustainable development is seen by Dryzek as elusive, intentionally vague and therefore appealing to most, ecological modernization (first named and described by Hajer (1995)) is more concrete. It constructs economic development and environmental protection as mutually reinforcing, with ‘green growth and jobs’ and ‘decoupling’ economic growth from environmental degradation as typical representative notions; it is also more concrete in proposing specific ways to achieve ecological modernization, the main one being through close cooperation of government, business and science.

The two radical and imaginative discourses of green radicalism—‘Green consciousness’ and ‘Green politics’—both reject the basic structure of industrial society and propose fundamentally different understandings of the environment, human–environment interactions, and society. While the first focuses more on changing the way people think, the emphasis of the latter is more on political change and social structure. Both include a number of diverse political and social movements, including social ecology, deep ecology, bioregionalism, ecofeminism, and environmental justice, generally characterised by an ecocentric (as opposed to anthropocentric or utilitarian; see Callicott 1984) ethic motivating environmental protection.

Definition of agri-environmental discourses

For the purpose of this study, the integration of elements from existing theoretical frameworks of CAP and environmental discourse analysis was necessary. To define agri-environmental discourses for the purposes of the analysis, the vocabulary (micro-textual element) and the main questions they answer (macropropositions as macro-textual elements) were pinpointed. Examples of such micro-textual elements that characterise discourses are specific wording and metaphors, such as e.g. self-sufficiency, species decline or green growth, while examples of macro-elements are overarching notions (global meanings) such as ‘the state ought to subsidize farmer incomes’ or ‘farming is responsible for environmental pollution’.

Codes were developed in several steps; in the first, codes were developed inductively belonging to any of the CAP and environmental discourses listed above. This first development was grounded both in the cited literature on CAP and environmental discourses and practical examples of usage by CAP actors (see examples in the description of discourses below). After codes were applied to the texts (see Material) in this round, it became apparent that there is significant overlap (in terms of macropropositions) between some CAP and environmental discourses. Most notably, the CAP’s neoliberal discourse and Dryzek’s Economic rationalism were merged due to their core common macroproposition, i.e. focus on economic growth and reliance on market mechanisms. Similarly, Productivism and Prometheanism were merged into a single discourse due to the overarching focus on agricultural productivity growth that ignores environmental limits. Discourses not appearing (Democratic Pragmatism) were omitted and distinct emerging discourses resulting from the insights of either existing theoretical framework (i.e. Multifunctionality, Ecological modernization and Administrative rationalism) were kept; due to the low level of presence in the documents, the radical green discourses were conflated, but kept due to their relevance to agricultural policymaking. In the second step, the texts were deductively coded anew with the refined set of codes.

Discourse analysis is necessarily an interpretive exercise (Janks 1997)—a simplified representation of what is in reality a rich interplay of conscious and unconscious utterances reflecting ideologies and cognitive frameworks, but also hard socio-economic realities. It must therefore be acknowledged that the selection of the materials for analysis and partly the development of discourses partially grounded in these texts has likely significantly affected the formulation of discourses and our list can therefore not claim to be final. An analysis of additional national/regional texts might thus yield additional or different distinct discourses.

The 6 discourses proposed here based on the analysed documents and summarised in Table 1 are ‘Productivism’, ‘Classical neoliberal’, ‘Ecological modernisation’, ‘Administrative’, ‘Multifunctionality’ and ‘Radical green’. In addition to the original theoretical framework, the table includes the main keywords and macropropositions with regard to what we propose as the discourses’ defining elements. They include notions on farming, the environment and their interrelation and relative importance of issues (cf. e.g. Feindt and Oels 2005) and indicate the interpretation and legitimation of agri-environmental policy. Furthermore, being policy discourses (pertaining to a policy with heavy budgetary disbursement), an important element is also the role that is ascribed to the state (cf. Potter and Tilzey 2005).

Table 1 Proposed agri-environmental discourses and their main elements
Full size table

The first discourse defined as a result of our reading is the industrial discourse of ‘Productivism’, a derivation of the productivist CAP discourse and Dryzek’s Promethean discourse. It conceptualises nature as an infinite source of brute matter and largely denies, ignores or dismisses environmental issues; when acknowledged, they are seen as an obstacle to be overcome in the production of food, while environmental restrictions are a constraint. Adherents (agricultural lobby groups and their representatives; see Potter and Tilzey 2005) traditionally tend to emphasise food scarcity and the growing world population while at the same time demanding subsidies (as commodity prices are too low), investment support and protection from unfair trade (e.g. Dorfmann 2018). The main goals to be achieved are higher productivity, food security, self-sufficiency and fair prices for agricultural produce. Competitiveness may also appear in this discourse, though often more in terms of reflecting protectionism than within the liberal-economic conceptualisation of competitiveness as the driver of innovation, economic growth and consequently prosperity (cf. Linsi 2020). Environmental demands are in this case seen as elements constraining competitiveness that should be compensated for through subsidies or higher commodity prices.

The neoliberal discourse actually has (at least) two different manifestations in the CAP if environmental issues are considered, with economic growth being the overarching goal of both. The first, ‘Classical neoliberal’ (a combination of the CAP neoliberal and Dryzek's ‘Economic rationalism'), focuses more on the competitiveness of agriculture as an economic activity and the need to abolish subsidies, which distort market signals (e.g. DEFRA 2013; see Anderson et al. 2013 and Sumner and Tangermann 2002, for a discussion on the distorting effects of agricultural subsidies). The environment is not very prominent in this discourse; when environmental issues are addressed in the analysed texts, they are seen through an economic lens as negative externalities or market failure, which are to be dealt with through the proper definition of property rights or, as a measure of last resort, through government intervention—penalties, taxation or compensation for the provision of public goods (welfare) that are not reimbursed through the market due to deficient pricing (e.g. Silvis and Van der Heide 2013). Nature may also be constructed as natural capital to be rationally managed, including the accounting and utilisation of ecosystem services. ‘Ecological modernisation’ (neoliberal + ‘Ecological modernisation’), by contrast, is a more positive discourse that explicitly recognises environmental issues and emphasises the opportunities offered by green growth and green consumerism—the bioeconomy thus features prominently. Rather than focussing on leaving everything to the market, it proposes that science, business and government must work together to achieve the best environmental and economic results. This discourse also places more emphasis on people’s well-being (e.g. OECD 2019).

A further discourse that can be expected to be found in any agricultural policy documents is the ‘Administrative’ (Dryzek 2013). The main characteristic of this discourse is deference to policy documents at higher levels (legislation, strategies, international accords) and expert knowledge as reflected in research, SWOT analyses and environmental assessments. Trust in administrative and legislative processes is implicit. When this discourse is employed, the issues themselves are not as important (and can vary) as the authority that is behind them; they are taken as self-evident, with the terms used technical and leaving little room for debate or reflection on the value judgements that have been made beforehand, indicating ideological hegemony by masking effects of power and inhibiting critical analysis (cf. Taylor 2009). Therefore, references to standards, indicators, targets, classifications, good ecological status and other seemingly blank (value-free) terms are common. This discourse is typically used by administrators and scientists when appealing to existing normative commitments (e.g. DG AGRI 2017).

Further, there is the discourse of ‘Multifunctionality’, described by Potter and Tilzey (2005) as a discourse explicitly substantiating public intervention in the face of market failure to provide for ‘public goods’—which is the heading under which the environment falls. The discourse borrows heavily from neoliberal terminology and stretches it to some extent: while the notion of public goods is strictly defined by neoclassical theory as a good that is non-excludable and non-rivalrous in use, (e.g. Cooper et al. 2009), it is quite malleable in practice and there is some debate regarding what qualifies as an agricultural public good (Burrell 2011; for a practical example, see e.g. NatureScot 2020). This discourse constructs farmers/agriculture/rural areas as providing public goods wanted and used by the people but not compensated for through the market and hence deserving of public support in line with the principle of ‘provider gets’ (as opposed to ‘polluter pays’) (e.g. OECD 2008). This discourse is primarily attributed to the European Commission (Garzon 2007), but has been increasingly employed and recontextualised by agricultural stakeholders (Potter and Tilzey 2005) and their representatives to substantiate existing subsidies, as well as by moderate environmentalists and scientists attempting to achieve the redistribution of CAP funds towards environmental goals.

Since the objects of analysis are policy documents, the most radical environmental discourses (such as the ‘Limits to growth’ and ‘Green consciousness’ discourses) were not expected to be found in the documents in any noteworthy extent. However, some of them are employed by environmental stakeholders (e.g. Last chance CAP 2018) and scientists (e.g. Pe’er et al. 2020) when criticising the CAP. Since environmental discourses have affected the greening of the CAP over time (cf. Clark 1997) and might still appear in traces in policy documents, we have included them in our framework, conflated into one category (‘Radical green’). While it has a number of variants, the discourse’s defining characteristic might be that it sees agriculture and the CAP as contributing (with ‘perverse subsidies’) to widespread environmental destruction, and its adherents (e.g. environmental NGOs and movements, some scientists, organic farmers) demand that environmentally harmful subsidies be discontinued; while it has not yet been described as a distinct CAP discourse, it has been gaining momentum. This discourse is often characterised by dramatic vocabulary, such as loss, destruction, degradation and threats to nature; the notion of animal welfare, if promoted in order to actually promote well-being as opposed to catering to consumer demand, may also be seen as an element of this discourse (see e.g. Arcari 2017, and references therein, for a discussion on the positioning of animals in public documents).

Materials and methods

Discourse analysis was applied to identify specific agri-environmental discourses, their elements and interplay. Since discourse is a representation of reality through language (Fairclough 2002), no analysis of material reality itself was performed.

Material

At the EU level, the European Commission’s 2010 Communication (EC 2010) was analysed. This is one in a string of Communications that precede the legislative proposals for the periodic CAP reforms, in this case the programming period 2014–2020. Communications have often been the subject of discourse analysis (Erjavec and Erjavec 2009; Rutz et al. 2014), as they provide insight into the Commission’s policy orientation, its preferred direction of travel, as well as the reasoning and justification of the content of the reform to be proposed. Furthermore, the Communications tend to constrain the discursive space of all the actors engaged in the respective CAP reform cycle (cf. Carstensen and Schmidt 2016). The Communication, a fairly brief document, was analysed in its entirety.

At the national/regional level, the Rural development programmes (RDPs) for the programming period (2014–2020) introduced by the above-mentioned document were analysed. As such, they could in a way be considered its emanations. They are particularly well-suited for analysis because they are, in principle, identically structured and address the same set of issues in each state/region, but emphasise different areas and thereby substantiate the choice of policy measures. This makes potential differences in each policy community’s values, priorities and programming approaches readily apparent. The programmes, which require the participation of stakeholders, thus also reflect the extent of inclusion and relative power of each stakeholder group (a detailed analysis of which exceeds the scope of this paper). In addition to programming measures, they contain a number of longer textual chapters amenable to discourse analysis that were analysed here, namely:

  • ‘Member state or administrative region’ (Chapter 2),

  • ‘SWOT and identification of needs’ (Chapter 4),

  • Description of the strategy’—‘A justification of the needs selected to be addressed by the RDP, and the choice of objectives’ (Chapter 5.1).

The last available English-language versions of the programmes were used for analysis (MA 2021; DEFRA 2022; MAFF 2014). With the exception of some sections of the newest version of Croatia’s RDP, which substantiate funding from the EU Recovery instrument, no notable changes over time to the analysed sections were found.

The parts of the text detailing and justifying the selection of measures and financial allocations were not selected due to being too technical or otherwise empty of meaning for meaningful discursive analysis.

Description of RDP areas

Our selection of RDPs was guided by the aspiration to reflect the high level of diversity present in the EU; therefore, three vastly different areas in terms of their familiarity with the CAP and environmental circumstances were selected:

  • Croatia, part of ex-Yugoslavia and thus a former member of the socialist bloc, is the EU’s newest MS, making the analysed RDP its first since accession; in addition, its agricultural sector, though the country is endowed with good conditions for farming, has not yet undergone large-scale intensification, and some of the country’s rural areas are still struggling to recover from the social and environmental consequences of the 1991–1999 war (Vidosavljević et al. 2013); the country is marked by a comparative technological lag in agriculture (Tomić 2020) and relatively well-conserved environment (Republic of Croatia 2014).

  • England, as part of the UK, has had experience implementing agri-environmental schemes since the ‘80s due to the consequences of agricultural intensification; the UK in general is well-known as a proponent of agricultural liberalisation, especially of the direct payments system (first Pillar under the CAP) (Alons and Zwaan 2016), while on the other hand being the first to introduce agri-environmental schemes (Dobbs and Pretty 2004, 2008). It is one of the most densely populated European countries.

  • (Mainland) Finland, which is regularly listed as a top environmental performer (Dryzek 2013), is simultaneously known for its high dependency on agricultural subsidies (Niemi and Ahlstedt 2015) and natural setting unfavourable to farming, as well as the lowest population density in the EU.

The selection was partly also pragmatic—all documents were available in the English language, allowing for easy comparability of the vocabulary applied. While this may have affected the final formulation of the proposed discourses, the countries selected are considered by the authors to be representative of conceptualisations of the role of agriculture and nature across the EU.

Analysis

The texts were imported and analysed using the programme for textual analysis QDA Miner lite (v 2.8) based on the codes developed and described above. Texts were coded on the levels of keywords and phrases (examples are provided in Table 1), but also based on word order (relative importance of concepts), coherence and local semantic moves (such as disclaimers and implications; van Dijk 1993, 1994; Wodak and Meyer 2009) at the micro-textual level. This analysis of micro-textual elements, extracted by identifying repetitive use of words that constitute the key notions in a text, served to identify the specific discourses, their combinations, contrasts and interplay in substantiating policy.

Macro-propositions at the macro-textual level were constructed from the micro-textual level as an extension of the analysis of the keywords that are the bases of micro-propositions or principal meanings (van Dijk 1980). These meanings were identified through abstraction and hermeneutic analysis (Paterson and Higgs 2005) to extract the prevalent macro-constructs, notions and value judgments with regard to agriculture, the environment and the relationship between them. Both keywords and macropropositions belonging to any of the discourses were coded in all the selected texts (see results of final coding in Annex). This was constantly accompanied by attention to recontextualization (utilisation of same or similar wording in different meanings or contexts) of competing discourses (Wodak and Meyer 2009) with regard to the socio-economic, ecological and political context based on the authors' expertise in the field and the relevant literature (Åkerman et al. 2005; Alons and Zwaan 2016; Batáry et al. 2015; Dobbs and Pretty 2004, 2008; Erjavec and Erjavec 2009; European Court of Auditors 2011; FAO n.d.; Garzon 2007; Jokinen 2000; Juntti 2002; Kelić et al. 2018; Niemi and Ahlstedt 2015; Potter and Burney 2002; Potter and Tilzey 2005; Rutz et al. 2014; Science for Environment Policy 2017; Skogstad 1998). Analysis of the logical chain of relations between macropropositions or main ideas was also carried out to reveal how they justify the key elements of the discourses. The prevalence of a discourse was assessed primarily on the basis of the order and place assigned to elements of each discourse.

Since language is not value-neutral, we used loaded (non-neutral) keywords (Kivle and Espedal 2022), emphatic intensifiers (‘truly’, ‘very’ etc.) and macro-propositions with a normative nature (especially as marked by the modal verbs ‘should’, ‘must’ or ‘ought to’) to analyse whether and how values are circulated within the analysed documents.

Results

EC communication

As far as the macropropositions or main ideas of the document regarding the CAP are concerned, the very first thing claimed is that the CAP should remain a ‘strong common policy’, which can be seen as a productivist (state-assisted) element, as the term is usually used to justify the continuation of funding (see e.g. Fazekas 2010). The primary strategic goal is entirely productivist, as well—“To preserve the food production potential” to guarantee ‘long-term food security for European citizens and to contribute to growing world food demand’. References are made to ‘increased market instability, often exacerbated by climate change […] and pressures’—typical productivist (exceptionalist) vocabulary (cf. Grochowska 2017). Arguably, the addition of the phrase ‘long-term’ might be construed as also implying producing it in an environmentally sustainable manner, although the importance of food production seems to overshadow other concerns in this statement. This line of thought is mirrored in the challenges section, whose first point (Food security) begins authoritatively with the wording ‘The primary role of agriculture is to supply food.’ The need to assert this can be seen as a response to the ‘other’ (primarily environmental, see e.g. Alons 2020) demands directed at the CAP, clearly establishing the hierarchy of priorities and communicating it to all actors.

The second and third strategic goals listed are ‘to support farming communities that provide the European citizens with quality, value and diversity of food produced sustainably, in line with our environmental, water, animal health and welfare, plant health and public health requirements’ and ‘this delivers multiple economic, social, environmental and territorial benefits.’ This substantiation invokes the numerous additional benefits (public goods) that the policy provides through support to rural areas, which are typical terms of the multifunctional discourse that help to justify the policy to the broader public (and trading partners). What is notable in both the productivist and multifunctional discursive instances cited is the mention of ‘European citizens’, as well as the lexical use (in the latter) of the pronoun ‘our’ [requirements], shifting responsibility onto ‘us’—citizens, while positioning farmers as caterers to societal demands, i.e., providers of public goods (as opposed to e.g. businesspeople (‘Neoliberal’) or polluters (‘Radical green’)).

Overall, environmental issues (regularly referred to in neutral terms such as ‘rural landscape’, ‘biodiversity’ and ‘climate change’) mostly appear in this ‘public goods’ construction, reinforcing the justification of the CAP’s existence in that if it was discontinued, the environment (inter alia) would suffer. The phrasing that, while ‘Agriculture and forestry play a key role in producing public goods’, ‘many farming practices’ also ‘have the potential to put pressure on the environment’, strongly affirms the role of farming in the production of public goods, while highlighting its potential negative effects much more carefully. This avoids framing agriculture as a polluting sector, while using the environmental argument to substantiate aid, a typical characteristic of ‘Multifunctionality’.

Once the need to maintain the CAP has been clearly established due to ‘important environmental and social consequences’ in case of ‘any significant cut back in European farming activity’, and thus, implicitly, in CAP funding, the text takes a more neoliberal turn, claiming that further reform of the CAP is needed to ‘promote greater competitiveness, efficient use of taxpayer resources and effective public policy returns European citizens expect, with regard to food security, the environment, climate change and social and territorial balance […] to build more sustainable, smarter and more inclusive growth for rural Europe.’ This passage provides an example of mixing different discourses, combining neoliberal elements (competitiveness, policy returns, growth) with productivist (food security) and multifunctional (environment, climate change, social and territorial balance) ones into an all-encompassing, nicely sounding potpourri.

In addition to the prevailing construction of nature as a public good or a collection of natural resources to be sustainably managed, in both cases providing a case for policy intervention, there is also reference to ‘sustainable growth’, ‘green growth’ and opportunities provided by ‘supporting new patterns of demand’ and the ‘emerging bioeconomy’, terms and concepts belonging to the discourse of ‘Ecological modernisation’. This potentially indicates a shift to a more neoliberal discourse, shifting the responsibility to provide adequate incomes to farming while preserving the environment onto consumers. However, abandoning the public goods discourse entirely seems unlikely, as this would effectively leave the policy without justification to the public and trading partners.

In conclusion, the Communication’s discourse can generally be seen as a hybrid (cf. Erjavec and Erjavec 2020). When looking at the environment, however, it seems that Productivism is strongest due to the primacy given to food production. The environment mainly appears as a source of justification both to taxpayers and to trading partners for continued budgetary payments in a number of sections where the multifunctional discourse prevails. However, there are also a number of neoliberally-minded passages (though not in regard to the environment), lengthy administrative passages on the specifics of measures and occasional notions that can be ascribed to ‘Ecological modernisation’. The mix of different discourses is likely employed to legitimise the continuation of the CAP to as wide a public as possible.

RDP for mainland Finland

The main macroproposition that can be abstracted from the Finnish RDP is that Finland’s plentiful and well-preserved natural resources should be combined with environmental entrepreneurship, knowledge and cooperation in order to achieve rural development: ‘The foundation for development efforts is laid by a tradition of entrepreneurial spirit, independence and initiative combined with natural resources expertise and the associated know-how.’ This is a typical notion of the Ecological modernist discourse. Within this overarching notion, farming and forestry (which features very strongly, especially in relation to the potential of the bioeconomy) are termed ‘industries’.

Nature is overall constructed as a source of biomass, a substrate, which, through ‘the opportunities offered by bio-economy and innovative, local and decentralised energy solutions will create possibilities for earning an income in rural areas.’ It is to be exploited, both as a physical resource and as a source of intangible benefits such as recreation. Increasing consumer interest and responsibility, with implications for production methods, are among the central topics of the document, seen as providing a competitive edge. Thus, environmentally oriented entrepreneurship is presented as a direction of agricultural development: ‘Environmental entrepreneurship may promote both employment and environmental management.’ Similarly, animal welfare is seen as ‘a competitive factor’, though there are instances where animal welfare and ‘natural behaviour’ seem to be promoted for their own sake (in a ‘Radical green’ manner).

There is a notable lack of justification for environmental action—the need to protect the environment is rarely explicitly substantiated and therefore appears quite internalised and uncontroversial. It is self-evident that biodiversity should be conserved and that pollution should be prevented. This may be related to the dependence of the Finnish national economy on natural resources. The environment is thus generally described in the ‘Administrative’ discourse—in technical terms, with numerous references to indicators, plans and targets, as well as competent (often public–private) institutions. The document is generally written using very technical language, with frequent mention of legislative and strategic goals and passages written in technical agronomic, environmental, economic etc. terminology; for the most part, there appears to be little explicit value-judgment. ‘Administrative’ discourse therefore prevails overall, indicating that environmental protection has been internalised as a value and reflecting a relatively high level of conformity with the original discursive space offered by the Commission.

Another overarching notion emerging from the analysis is that self-sufficiency seems to be secondary to the profitability and competitiveness of farming; similarly, unfavourable natural conditions seem to be a problem more due to low profitability as a result of lower yields than due to low self-sufficiency: ‘In the extremely unfavourable climatic conditions due to Finland’s northern location, farming is not competitive compared to farming in more favourable conditions. A reduction in the number of farms and the weak interest shown by new entrepreneurs in farming highlights the fact that farming is not a profitable industry in any part of the country. Without natural constraint payments and the development of the agricultural structure, farming in Finland will be doomed. […] Low yield levels and crop rotation increase the arable land area required and the need to develop production.’ There are thus some productivist (state-assisted) notions present in the document, though their utterance is tempered by an overall market-liberal orientation.

In summary, agriculture (and forestry) and farming are not framed in opposition to each other but in synergy as tools to achieve overall wellbeing of rural inhabitants, placed in an open and environmentally conscious market economy. Though there are some elements belonging to the discourse of ‘Productivism’, the strongest substantive discourse is thus ‘Ecological modernisation’.

RDP: England

At the macro-level, farming is generally constructed as a business, while both farming and the environment are seen through a monetary lens. The environment is mentioned at the very beginning of the document as a source of benefits to society, and it frequently mentions ‘sound management’ to preserve the provided ‘ecosystem services’. ‘Environmental assets’ and ‘natural capital’ are to be managed rationally in the spirit of good stewardship and with the help of administrative standards; there is frequent reference to monitorings. Monetary estimates as to how much the consequences of losing ecosystem services due to unsound management cost the economy exist for several issues. The document states in plain terms: ‘Evidence from the National Ecosystem Assessment [10] indicates that England is failing to conserve and invest in its natural capital assets.’

The document is very consistent in constructing farms as businesses and farmers as entrepreneurial ‘land managers’; ‘degradation of environmental assets’ is seen as irrational, as it could affect farming in the long run. Improving ‘resource efficiency’ and balancing ‘food security and environmental security’ are seen as competitive advantages (so-called ‘future-proofing’). Far from going to lengths to justify budgetary payments in a productivist manner, the document even explicitly states that ‘CAP subsidies support underperforming farms to remain in business, and limit structural consolidation.’ There is thus little sentimental attachment to small and family farms, which are often considered to be an important provider of environmental and other public goods in other settings (e.g. Borychowski et al. 2020; Cooper et al. 2009; Czyżewski et al. 2021). However, environmental care does provide a ‘case for intervention’ due to ‘market failure’. Markets for ecosystem services should be developed so that these assets may be better managed and the environmental impact reduced. This combination of elements—a strong reliance on the market, conceptualisation of farmers as businessmen and environmental protection as sound management of natural assets—places the English RSP firmly in the ‘Classical neoliberal’ discourse.

Increasing recognition of the value of landscape and historic environmental assets is also an opportunity for rural areas—‘Landscape character and quality is a key public good produced by agriculture. But the market does not reward land managers for improving the landscape, as it is a non-traded good, so intervention is required.’ The framing, though a little inconsistent in simultaneously providing substantiation for intervention using the public goods argument and invoking a growing market demand for ‘nature-based, sustainable tourism’, is much closer to the multifunctional substantiation of public intervention than the ecological-modernistic trust in responsible consumers and the bioeconomy as vehicles of environmental preservation.

On the other hand, there are a few instances of more ‘Radical green’ thinking, where the relationship with nature is not defined in purely utilitarian terms, e.g. when talking about protecting ‘wildlife’, ensuring ‘animal welfare’ (although mostly this is a goal to reduce disease and/or improve competitiveness; generally, the term used to denote animals is ‘livestock’), ‘harming nature’, ‘threatening sensitive habitats’ or ‘harming the natural environment’. These phrases are often included in lengthier passages that generally appear as expert phrasing, perhaps indicating that in this document elements of (radical) green discourse have already achieved a certain level of dominance, e.g.: ‘More efficient production and land use and changes in housing have restricted the living space of natural species, and the species in agricultural environments have become impoverished and endangered on a wide front. [emphasis added].’

In summary, it can be said that the English RDP contains a distinctively and consistently neoliberal discourse (‘Classical neoliberal’), though at the same time it presents the ideal farmer as a responsible land manager and environmental steward who must be reimbursed for his provision of public goods to society—preferably through the market, and in the case of market failure, from the public purse. The document is thus significantly more market-oriented than the Communication, as reflected in the critique of subsidies that impede structural consolidation; in the environmental domain, conserving (their) natural assets is generally seen as farmers’ own responsibility, though in certain cases it is also seen as falling in the public domain.

RDP: Croatia

The central macroproposition of the Croatian RDP is that agriculture and rural areas in general are very vulnerable and in need of modernisation: ‘Due to depopulation, the impact of the war and the long-term economic crisis, there has been negligence in the maintenance of infrastructure […], social and cultural facilities […], resulting in a poor availability of basic services for the local rural population.’ Furthermore, ‘further technical and technological modernisation of agricultural holdings […]’ are seen as ‘essential’ to ‘improve competitiveness and economic viability of agricultural holdings’. There is a great deal of concern regarding self-sufficiency—the fact that in a number of sectors ‘production does not cover the needs of the population, food processing industry and tourism’, is clearly highlighted as an issue. In view of this lag in technological development and ‘low investment capacity’, it is largely the role of the state to provide for lacks: ‘Agricultural holdings still often use pre-1990 technology and […] need support to enhance competitive viability’.

In line with the overwhelming concern for modernisation, self-sufficiency and competitiveness (against external competition, not as a stimulant to overall economic progress), nature does not feature very prominently. The environmental pressures of this intensification are largely ignored and requirements are seen as externally imposed ‘Community requirements regarding animal health and welfare and environmental protection’. As reflected in the listing of ‘High-quality natural resources and cultural heritage’ and ‘Wealth of biological and biogeographic diversity’ under strengths and in the statement that more water must be used for irrigation, it seems that the authors feel that there is significant leeway to intensify without serious environmental repercussions. Nature is thus something to be exploited, while ‘wildlife’ is generally a nuisance—something that causes damage. The overwhelming concern with increasing productivity and self-sufficiency with the help of the state while conceptualising nature as a limitless pool of resources are typical elements of the ‘Productivism’ discourse.

The description of the ‘Agricultural environment situation’ is brief and administrative, mainly referring to species and hectares under various designations, listing emissions from agriculture in the sections on water, consumption of fertilizers and greenhouse gas emissions with reference to administratively set standards. The motivation for environmental conservation is mostly either administrative compliance, as reflected in the usage of blank terms such ‘environmentally sound’, ‘environmentally efficient’. This application of ‘Administrative’ discourse may indicate either agreement with or submission to the discursive space encircled by the Commission.

On the other hand, climate change and ‘Environmental degradation’ are proposed as a threat in the SWOT (‘degradation in key environmental assets could significantly and negatively affect the farming sector in the longer term’), and ‘increasing environmental awareness in the farming community’ is a strength, meaning that environmental issues are (at least nominally) acknowledged. A reason to conserve biodiversity, which, according to the document, Croatia is ‘exceptionally rich’ in, is its value—it is possible that intrinsic value is meant, but the need to conserve biodiversity is also substantiated with native species’ ‘genetic potential’ in helping agriculture to adapt to climate change and serve as a basis for new and improved agricultural plants and breeds. This indicates that the value meant is instrumental rather than intrinsic. Similarly, reducing ‘the negative pressure of agriculture on the environment’ (one of the rare instances that this pressure is explicitly acknowledged) is seen as ‘a long-term return, through quality agricultural products as well as improving the quality of life of rural communities’. Furthermore, animal welfare, while ‘breeders have insufficient interest’ in it due to lack of experience in implementing standards, is seen as an economic opportunity to ‘increase demand among consumers for products that originate from animal-friendly breeding’. This line of thinking—acknowledging environmental issues but reconciling them with production and market forces—is characteristic of’Ecological modernisation’.

While the prevalent discourse in the document is thus mostly ‘Productivism’, it is possible to discern some elements of other substantive discourses, notably some ecological-modernist notions, which see environmental protection as a sound business opportunity and element of wellbeing (Table 2).

Table 2 Key discursive elements found in the analysed documents
Full size table

Discussion

Integrating the traditional discursive analyses of the CAP with the study of environmental discourses to enable a better understanding of the different constructions of agriculture and environment in the CAP policy process at different levels, the study filled the research gap on agri-environmental discourses and constructed six separate discourses that were applied to strategic CAP documents at the EU and national/regional levels. The study of integrated agri-environmental discourses has proved useful in gaining insights into the terms found under the CAP in relation to the environment, which neither literature can provide separately. While—by their very nature—the discourses identified still remain markedly CAP discourses, meaning that much attention goes to justifying public intervention, the combination of the two frameworks sheds new light on the variable success of environmental interventions.

The first insight that offers itself is a result of the inclusion of Dryzek’s (2013) ‘Administrative’ discourse, which was useful in explaining the analysed documents. While it is not really a substantive discourse in itself, the normalisation of some issues to the extent that they are portrayed as self-evident can be indicative of hegemony (Wodak and Meyer 2009). The European Commission, by the very act of publishing a Communication that subsequently results in legislative proposals and, if successful, legislative acts, acts from a position of power and constrains the discursive space in which other political actors operate. While this is actually less evident in the Commission text, which contains fewer references to hierarchically superior documents, scientific expertise and technical vernacular, it is starkly apparent in the Rural development programmes, which follow the Commission template. The very fact that EU MS and regions are constrained in preparing these documents with the form and content that have already been decided at higher political levels, as well as Commission oversight, significantly tightens their own political space. It also helps to diminish national/regional administrations’ responsibility with regard to domestic stakeholders (cf. Hobolt and Tilley 2014) and removes the need to substantiate funding to them, while leaving enough ‘wiggle room’ for domestically prevalent values and priorities to manifest. With the latest CAP reform (2022–2027), which allows for significantly more manoeuvring space for MS in both CAP pillars, the Commission has essentially relaxed this space somewhat.

In contrast to the ‘Administrative’ discourse, the significantly weaker presence of ‘Multifunctionality’ (understood as the substantiation of CAP payments with farmers’ provision of public goods) at the local level in comparison with the EU level is arguably an indication that the latter is the level where these payments are most strongly contested. The demands of EU citizens, including environmental ones, are employed towards this end. It would seem that, though the wording is very similar as the Commission’s, the public goods substantiation used in the English document is more explicitly tied to market failure for each individual good, placing it in a more consistent ‘Classical neoliberal’ discourse. That said, what constitutes public goods and market failure is always to some degree a matter of value judgment (see e.g. Alons 2020, for a discussion on the discursive malleability of the term ‘public goods’ and its different usage by different actors). This indicates the possibility that the authors of the English RDP were simply more diligent in invoking market failure than the Commission was, concealing some underlying agricultural exceptionalism. Considering the UK’s habitual stance in negotiations regarding the CAP, i.e. abolish direct payments and tie public money to public goods (e.g. Rutz et al. 2014), this interpretation seems less likely, however.

While the environment serves at the EU level as a source of justification for continued payments (‘Multifunctionality’) and is, in fact, to some extent normalised and included into standards (‘Administrative’), it is still secondary in importance, as reflected in the Commission’s assertion that the primary role of agriculture is to provide food. As the environment acts here as one of the demands constraining the quantity of production, it seems that agriculture and the environment are viewed in overall opposition to each other, which is generally the stance adopted by agricultural lobbies and their representatives. This situation is similar to that of Croatia (and likely some other New MS; see Chaisty and Whitefield 2015), where environmental values are secondary. In the Croatian RDP, there is little direct opposition to the general idea of environmental protection—an example of the local level being constrained by EU-level discourse; rather, it is pushed into the background. The externally imposed demand to address it is satisfied through neutral terms such as environmentally sound production, while the severity of the infrastructural and productivity-related deficit is awarded primacy and the lack of self-sufficiency in certain foods is considered an important issue. This coincides with some of the findings regarding the ‘mental landscapes’ in European agriculture, which were shown by Wilson (2001) to still be embedded in productivist thinking. The author found a lack of evidence of a shift towards post-productivist thinking among both EU and national policy officials, let alone farmers (cf. Burton et al. 2008; Almås and Campbell 2012; Howley et al. 2015). Certain authors actually argue that both thinking and policy have taken a turn back towards productivism or neo-productivism due to world food price hikes (2008 and 2011), as well as the financial crisis starting in 2008 (e.g. Almås and Campbell 2012; Burton and Wilson 2012; Rønningen et al. 2012). More recently, the COVID-19-induced fear of food scarcity has revived the push for self-sufficiency in some places due to disturbances in supply chains (cf. Elleby et al. 2020; Euractiv 2020), while the war in Ukraine has resulted in some environmental requirements being relaxed for food production (Euractiv 2022).

The Finnish RDP, by contrast, seemingly embraces the emerging demand from environmentally conscious consumers and sees it as a competitive edge rather than as a hurdle (‘Ecological modernisation’ discourse). There are traces of this kind of thinking in all the documents analysed, but it features most strongly here as manifested through notions like environmental entrepreneurship and bioeconomy. This is similar, but not identical, to the rationality ascribed to environmental stewardship in the English neoliberal variant; while in the ecological modernist construction, the market is actually a vehicle of environmental conservation, this is less so in the classical neoliberal, where environmental conservation is a matter of prudence and still often seen as going against market forces, constraining competitiveness. ‘Ecological modernisation’ is an integrative discourse in a similar way as ‘Multifunctionality’, the main difference being the underlying premise of whose responsibility it is to bear greater costs of food production: in ‘Ecological modernisation’, it is the conscious (and affluent) consumer; in ‘Multifunctionality’, it is the public purse. Given the increasing attention to the bioeconomy and green growth (e.g. the European green deal and Bioeconomy strategy), especially in affluent societies, the ‘Ecological modernisation’ discourse is likely to become the strongest one both in the CAP and outside it; barring crises such as the current one related to the war in Ukraine, which tend to push productivism back to the forefront and environmental issues into the background, the two discourses competing for dominance in future CAP reforms may very well be ‘Ecological modernisation’ and ‘Multifunctionality’—since it is becoming increasingly difficult even for the most hard-nosed Productivists to ignore environmental issues completely.

Rather unsurprisingly, there is little ‘Radical green’ discourse in any of the documents. While it is employed strongly by actors outside the traditional policy community (cf. Greer 2017) of the CAP, it can be said that the discourse has had the effect of including environmental elements into the policy over time, but it is unlikely to feature very strongly in policy documents. Furthermore, what is arguably the most ‘radical’ element of the value system of the discourse’s adherents, i.e. intrinsic value of nature, appears only in traces.

The fact that the CAP and the texts produced within its remit are sites of ideational contestation (e.g. Coleman et al. 1996; Daugbjerg and Feindt 2017; Greer 2017) is already apparent at the EU level, where the Commission communication switches between different discourses to navigate all the different interests tugging at the CAP boat (cf. Erjavec and Erjavec 2009). The discourses employed at the local level, being the product of a larger number of people with different levels of oversight, are in a way even more schizophrenic and inconsistent at first glance. However, it is possible to discern in each of the lower-level (national/regional) documents a distinct overarching value system with regard to nature and agriculture of the local policy community as a whole. This value system becomes readily apparent when comparing documents intended to address the same general set of issues—despite all the documents being conglomerates, each of them still carries a clear undertone. One possible explanation for the differences between countries, as well as different social groups within them, may be the prevalence of survival values over values of self-expression (including environmental) related to lower existential security (Inglehart 2005).

Social constructs do not come into existence out of nothing: they have a physical context (or materiality, which is highly salient when discussing environmental discourses; Barnaud et al. 2021), a socio-economic context and a historical background in discourses previously employed to deal with similar situations (Hajer 1993). While this paper is by no means the first to assert that there are differences between countries in terms of their value systems with regard to the environment (Gelissen 2007; Havasi 2012; Sargisson et al. 2020), socio-economic situations and CAP implementation styles due to different priorities (Jongeneel et al. 2018; Juntti 2002; Vesterager et al. 2016; Zhelyazkova et al. 2016), it is relatively rare that this has been analysed using discourse analysis. Given the direction of travel of the CAP, under which a reform has just been agreed granting MS more freedom to formulate the policy with regard to their own priorities (Erjavec et al. 2018; Jongeneel et al. 2019; Lovec et al. 2020; Rac et al. 2020), it seems that the Commission has acknowledged that differences are so great that they are causing too much political friction, making an entirely ‘Common’ policy untenable. This may mean that in the future Strategic plans, which are to replace RDPs and cover the entire policy, it will be possible to discern even greater differences in the values and priorities, discourses, and consequently policy measures and budgetary allocations—with potentially important implications for the environment.

Turning to limitations, there are interactions between discourses that serve to confound analysis. There are numerous instances where words and notions from one discourse are recontextualised in another, the most obvious instances being ‘public goods’ and ‘competitiveness’; in the case of the discourse that we have named ‘Administrative’, ideas become normalised and subsumed to the extent that they enter the realm of standards and indicators, becoming self-evident and unquestioned, such as ‘economic growth’, ‘climate change adaptation’ or ‘livestock units’.

Another limitation of the analysis is the effect of the language barrier. The level of familiarity with the English language of the regional/national documents’ authors may mean that certain terms are used mistakenly, inadvertently or through copying of other English-language texts. However, this is not believed to be an overall limitation for the utility of the framework itself. Rather, its application both to other national/regional RDPs and other kinds of texts (such as media appearances, policy statements or other official policy documents) can yield further interesting insights into the conceptualisations of nature and farming across the EU.

To further substantiate findings in terms of the environmental ambition present in different regions and MS, it would be useful to compare the amount of funding allocated to environmentally beneficial CAP measures. Even more, since these allocations still allow for a significant leeway with regard to actual environmental ambition, a complex interdisciplinary analysis assessing the quality and results of individual measures in both pillars would be necessary; this kind of analysis vastly exceeds the scope of this paper, but would probably represent the most reliable measure of environmental ambition. Agri-environmental measures are especially difficult to assess due to their high diversity (Zimmerman et al. 2016), as well as the difficulty of effectively monitoring and evaluating policy measures by linking specific measures to environmental effects (Piorr and Viaggi 2015; ECA 2020). Examples of such research, though partial, can be found in e.g. Juana et al. 1999; Kleijn et al. 2003; Keenleyside et al. 2011; Vesterager et al. 2016; Zimmerman et al. 2016; Jongeneel et al. 2018.

Finally, the analysis at the national regional level was only conducted on second-pillar documents, while the EU-level document pertains to the entire policy. This decision was entirely practical: MS and regions were severely constrained in implementing the first pillar and produced no texts amenable to discourse analysis. Thus, at the national level, these two pillars may actually differ in terms of the departments responsible and thus in emphasis (cf. Jongeneel et al. 2018); this is a potential source of bias that must be acknowledged. However, since the second pillar is subject to strategic planning and broader consultation including environmental stakeholders, the policy community is likely similar to that at the EU level, with the possible exception of international EU trading partners. This may actually help to explain the notable absence of the discourse of ‘Multifunctionality’ at the local level, as the need to justify potentially trade-distorting income support is weaker.

On the other hand, the strategic plans to be implemented in the post-2023 period will pertain to both pillars. This means that these singular documents, theoretically prepared with a whole-policy focus, should be even more appropriate for analysis such as the one conducted in this paper. The same can of course be said of analysing their antecedent, the 2017 Communication, and potentially the interactions between the two levels. It might be expected that, given the new freedom given to national policymakers, the lower-level discourse could actually be even clearer—perhaps better reflecting the local policy community’s values both in the discourse applied and in the allocation of funds and environmental ambition. This would imply that, in order to enforce a common environmental standard, EU-level legislation would have to implement more stringent non-voluntary rules—which is exactly what was proposed and subsequently adopted in the legislative acts (EC 2017; EU 2021; Erjavec et al. 2018).

Finally, as stated above, the partial development of discourses based on the selected texts has undoubtedly affected the final formulation of the discourses; different texts might yield a different number and/or content of discourses. This bias is hopefully at least partly offset by the fact that the developed discourses are also grounded in theory and the authors’ practical experience. Further discussion and analysis to test the overall robustness of the proposed framework is welcomed, as well as refinements that are likely necessary for analytical purposes.

Conclusions

The synthesis of the results of the CAP discourse analysis and the environmental discourse analysis has shown that there are 6 different agri-environmental discourses that appear in the analysed strategic policy documents at the EU and national/regional level, as a combination of discourses to appease different groups of stakeholders and to accommodate different value systems.

The discourse analysis has revealed that there are indeed differences in how farming and the environment are conceptualised at the different levels of CAP decision-making within the Union. At the EU level, farming is primarily understood as a sector whose main task is to produce food (discourse of ‘Productivism’), to which everything else is subordinated, and the environment is generally used as a justification for CAP payments (discourse of ‘Multifunctionality’). At the national/regional level, RDPs, even if they follow in their broad outlines the template provided by the Commission, seem to reflect different value systems present in the policy communities: in England, environmental protection is seen as sound management of natural capital (‘Classical neoliberal’ discourse), in Finland, it represents a benefit for producers and conscious consumers to be mediated by public–private partnerships (‘Ecological modernisation’ discourse), while in Croatia it seems to be a necessary evil limiting improved productivity (discourse of ‘Productivism’) and imposed by an external authority (‘Administrative’ discourse).

This diversity in constructions shows that, despite the fact that the Commission sets a specific discursive space and even significantly constrains MS’s/regions’ freedom by providing a template for RDPs, the latter’s differences can still visibly manifest in the documents’ contents, providing a basis for a better understanding also of the differential implementation and success of environmentally beneficial measures across the Union. We hope that the proposed theoretical framework can provide a useful tool for understanding the differences in perceptions of the interplay of agricultural and environmental issues under the CAP, and consequently in designing policy interventions that take such differences into account as well as providing early indications of potentially problematic countries/regions. This does not necessarily mean e.g. lowering environmental standards in localities where environmental constraints for farming are seen as an obstacle to economic development; rather, it implies closer, more inclusive dialogue with the policy community based in sound data and transparent, locally tailored decision-making informed by an understanding of local values, needs and priorities that may be discerned from the employed discourses.