**3. Review**

#### *3.1. Cross-Sectoral Considerations in EU Policy Documents by Policy Sector*

As point of departure for the in-depth qualitative analysis of policy integration, we examined policy documents for considerations of policy coherence. In a first step, explicit references to policy coherence were considered by policy sector. Policy coherence, in this context, was defined in line with its understanding in the Sustainable Development Goals as 'synergies and trade-o ffs among [ ... ] targets, between di fferent sectoral policies, and between diverse actions at the local, regional, national, and international levels' [77]. This definition served as the basis for analysis, where the specific focus was set on the cross-e ffects among the sectors. The analysis revealed that, within the policy sectors of direct relevance to water, energy and land, considerations of policy coherence are most prominent in the sector of general environmental policy. Of the directly resource-relevant sectors—water, energy, and agriculture—policy coherence is most strongly considered in water policy documents (cf. Table S1: WFD, WFD\_DecMak, WatScarcImpact, WFD\_IntegrWat, NitrDir\_Backgr, SoilWaterStudy, ResEcE ffWatDistr, WFD\_Leaflet). Given that environmental considerations and sustainability objectives are of direct and spanning relevance for all resource sectors, this result is not unexpected. Furthermore, as discussed (Section 2.1), the EPI focus of integrating environmental concerns into other sectors has long been part of EU policies, which accordingly is reflected in the documents.

In a second step, the documents were reviewed with regard to implicit considerations of policy coherence for sector-specific policy considerations. The results indicate that of the relevant policy sectors, in the energy sector, cross-sectoral considerations are least integrated into policy design. In the considered energy policy documents, no reference to water policy was found. Agricultural policy was singularly referred to with regard to the combined emission of CO2 in the case of energy crop production for bioenergy (cf. Table S1: En2030Strt). The same trend shows in the analysis of the agricultural and water policy sectors with none or only few energy policy considerations (cf. Table S1: WatScarcImpact, SoilWaterStudy, ResEcE ffWatDistr). According to the results, it can further be assumed that policy coherence, at least to some degree, has been achieved among agricultural and water policies with frequent cross-sectoral considerations, respectively.

To further di fferentiate these first, general results, a more detailed analysis of policy coherence was conducted. For this purpose, the documents were analyzed for reflections of cross-sectoral considerations. Specifically, the implementation of policy coherence as a policy objective was searched for and found primarily in general environmental policy with few instances in water and agricultural policy (cf. Table S1: IndicIntegrEnvCAP, ReviewEIA, CAP\_2013RefOverv, WatScarcImpact, ResEcE ffWatDistr). These observations again confirm that the implementation of cross-sectoral policy considerations is least advanced in the energy policy sector.

Furthermore, references to the respective other resources were searched for. One central result was that the agricultural policy sector seems to play a crucial role when analyzing the integrated policy perspective, since the respective cross-sectoral considerations with energy and water resources were repeatedly and specifically addressed (cf. Table S1 CAP\_Markets, CAP\_RuralDev, CAP\_Tow2020, CAP\_2003CrossComp, CAP\_Payments, IndicIntegrEnvCAP). Also, these considerations were found in di fferent documents and address a broad number of di fferent objectives. The analysis of energy policy documents, in contrast, revealed references to both agriculture and water policies; however, all originate from the same document, namely the Renewable Energy Directive (2009/28/EC) (cf. Table S1: RenEnDir). Within this document, general sustainability considerations play an important role, so that merely the very general objective of 'measures taken for soil, water and air protection, the restoration of degraded land, the avoidance of excessive water consumption in areas where water is scarce' is referred to [78]. The review of policy coherence finds that the agricultural sector plays a central role for integrated resources governance. In order to look beyond these indicative results, they serve as the starting point for the subsequent comparative analysis of policy integration among the sectors.

#### *3.2. In-Depth Analysis of Vertical and Horizontal Policy Integration in Nexus-Relevant European Policies*

To consolidate the above analysis, this section provides an in-depth qualitative comparative analysis and discussion of those EU energy, water, and agricultural policies that were identified to show a significant degree of implemented vertical and horizontal policy integration. Firstly, we evaluated the vertical dimension. This allows us to investigate to what extent EU sector policies are nexus-informed.

### 3.2.1. Vertically Integrated EU Water and Agricultural Policies

In case of water and agricultural policies, a high degree of vertical policy integration can be found. In Europe, the managemen<sup>t</sup> of water resources is primarily regulated via three binding directives: (a) the Water Framework Directive (Directive 2000/60/EC), (b) the Bathing Water Directive (Directive 2006/7/EC), and (c) the Marine Strategy Framework Directive (Directive 2008/56/EC). The WFD, as the EU's main instrument for water protection, requires its member states to achieve 'good status for surface and groundwater' by 2015 (Article 4 WFD). It was adopted in 2000 and in this early version highlights the need to integrate water protection and sustainable managemen<sup>t</sup> into other policy areas such as 'energy, transport, agriculture, fisheries, regional policy and tourism' [79].

One integral part of theWFD is the Nitrates Directive, which came into e ffect in 1991, and specifically addresses the prevention of ground and surface water pollution from agriculture by promoting the use of good farming practice [80]. Within the WFD, several measures have been defined with a direct link to agriculture, including, e.g., the managemen<sup>t</sup> of water demand, fertilizer emissions, as well as efficiency and reuse measures [81]. It thus shows that the negative side-e ffects of intensive agriculture production were considered in water management. To date, however, specific analyses to determine the actual e ffects of such policy frameworks on other resource sectors are rare [2]. However, the link between water and agriculture policies is bilateral, as the above results indicate.

The CAP is currently the most important EU policy framework to mandate the formal consideration of externalities among the nexus resources, especially low water quality resulting from agriculture. With a total spending of over 58 billion Euro, the CAP accounts for almost 40% of the EU budget [81] and pursues three main long-term objectives: (a) viable food production, (b) sustainable managemen<sup>t</sup> of natural resources, and (c) climate action and balanced territorial development [82]. In March 2013, the cornerstones of the post-2013 CAP were defined. 'Cross-compliance' was reinforced, directing payments for farmers to comply with rules on farming practices that account for the environment, food safety, animal and plant health, and animal welfare while maintaining 'agricultural land in good agricultural and environmental condition' [81]. The post-2013 reform of the CAP was motivated by three main challenges within the agricultural sector, which—although not explicitly referring to the WEL Nexus—very closely correlate to the nexus challenges (Table 1):


**Table 1.** Challenges in the agricultural sector [82].

In order to address these challenges, a new policy instrument has been added to the new CAP: the 'Green Direct Payment' [82]. It rewards farmers for complying with three agricultural practices: (a) maintenance of permanent grassland, (b) ecological focus areas, and (c) crop diversification [82]. Furthermore, as part of the second pillar, rural development, at least '30% of the budget of each rural development program must be reserved for voluntary measures that are beneficial for the environment and climate change' and will be implemented as part of the national (or regional) rural development programs [82]. Given the financial endowments of almost 100 billion Euro for the period from 2007 to 2013, rural development could thus contribute substantially to funding the protection of water resources [81].

The main challenge to do so consists in that EU water policy objectives are anchored in different policy areas sub-ordinate to different authorities with partially contradictory interests [81]. In order to overcome these challenges, the re-established mechanisms of cross-compliance and the European Agricultural Fund for Rural Development provide the means to more strongly encourage good farming practice in compliance with environmental legislations [82].

Integrating policies across the water protection and agriculture policy realms in the EU constitutes a valuable contribution towards an integrated policy framework, which comprehensively considers both direct and secondary policy effects on each of the nexus resources. Notwithstanding the progress made, the European Court of Auditors' analysis pointed out that the integration of EU water policy goals as part of the CAP cannot be regarded as completed yet, and that several major challenges remain, often related, e.g., to weaknesses in the definition of standards or inconsistencies among member states [81].

3.2.2. Approaches for Horizontal Policy Integration in Nexus-Relevant European Policies

Next, horizontal policy integration was investigated, i.e., the degree to which the EU has developed cross-sectoral strategies. Specifically, this analysis evaluates to what extent the EU pursuits an overall integrative approach. In addition to the sectoral approaches of policy integration, further instruments of horizontal EPI exist in the EU in the form of political initiatives, overarching strategies, plans and assessment tools that are of grea<sup>t</sup> relevance for achieving a nexus-enabled policy approach. Prominent examples of such merely horizontal approaches are EU roadmaps. As part of the Energy Roadmap 2050, the EU reaffirmed its objective to reduce greenhouse gas emission to 80–95% below 1990 levels by 2050 [83]. Furthermore, the share of renewable energies in final energy consumption is to increase to at least 20% by 2020. More specifically, in terms of the energy sector, the 'Roadmap for Moving to a Competitive Low-Carbon Economy in 2050' broadly refers to potential negative side-effects of renewable energies, e.g., biofuels, concluding that 'any negative impacts on other resources (e.g., water, soil and biodiversity) will need careful management' [84].

Thus, for the energy sector, horizontal policy approaches exist within these roadmaps, formulating overarching objectives that are to be reached by vertical processes. Similar approaches exist for water management. The 'Roadmap to a Resource E fficient Europe' defines a number of non-binding water managemen<sup>t</sup> objectives including, for example, to keep water abstraction below 20% of available renewable water resources, while minimizing the impacts of droughts and floods through adapted crops, increased water retention in soils, and e fficient irrigation mechanisms [85]. In accordance with the nexus understanding, the roadmap further calls for a comprehensive and integrative managemen<sup>t</sup> approach to achieve these objectives [85]. Such overarching strategies and plans—despite their reference to the main nexus idea—currently lack an operational implementation in day-to-day policy making [48].

In addition to those long-term strategies, which propose concrete and, in some cases, binding quantitative targets, other horizontal approaches also have a long-term background. One such approach to putting environmental objectives onto a stronger legal basis was the Cardi ff-Process in the early 2000s. It moved forward the integration of environmental and sustainability considerations into the di fferent policy sectors and stipulated learning that has translated into further nexus-relevant policy measures. It followed upon the addition of Article 6 to the EC Treaty (COM (1998) 333) in 1998, according to which 'environmental protection requirements must be integrated into the definition and implementation of the Community policies [...] in particular with a view to promoting sustainable development'. Accordingly, it requested a number of Councils to develop strategies for the integration of environmental and sustainability considerations into their respective policy fields for the European Council meeting in Goeteborg in 2001. Post-Goeteborg analyses, however, revealed huge di fferences in the quality, scope, and ambition of objectives among the submitted documents [86]. Unfried [87] identified the 'lack of support from political leaders together with—maybe also as a result—the lack of commitment among the councils of ministers [ ... ] as reasons the Cardi ff process had not achieved the intended level of integration', so that the Cardi ff process was eventually 'singled out for its unrealistic potential to lead to policy learning for EPI' [15]. Nonetheless, the Cardi ff process has instigated some important progress in the policy integration of environmental considerations. It was an innovative policy measure that stipulated, e.g., cross-sectoral councils and working groups as well as significant progress in sectoral environmental integration indicators [88]. It further revealed important challenges inherent in cross-sectoral policy integration that, in a similar form, must be overcome in order to implement an e ffective nexus policy framework. These challenges include (a) promoting the implementation of the policy measures in the member states, (b) the development of national processes for integration, (c) establishing harmonized reporting mechanism for information exchange, and (d) promoting a process of trans-national policy learning and the development of networks of experts [86].

Other measures implemented to environmentally assess operational public policies include, for example, Environmental Impact Assessment (EIA) (Directive 2011/92/EU) and Strategic Environmental Assessment (SEA) (Directive 2001/42/EC) [89] as procedures to ensure that 'the environmental implications of decisions are taken into account before the decisions are made' [90]. Within the EIA, the direct and indirect e ffects of a project on (a) human beings, fauna and flora, (b) soil, water, air, climate and landscape, (c) material assets and cultural heritage, and (d) interaction between the above listed are to be identified [91]. In contrast to the EIA, SEA applies to public plans and programs—not, however, policies [90]. In addition to identified implementation gaps, the absence of obligatory environmental standards was identified as a major drawback [92]. This gives rise also to challenges related to the fact that many projects are under the jurisdiction of di fferent member states, which in turn inheres the risk of duplications, inconsistencies, burdens (e.g., administrative) and thus potential conflicts [92]. Such ine fficiencies often result from di fferences in the national EIA procedures, e.g., with respect to di fferent stages of the project proposal process or di ffering timeframes, but could likely be reduced, if more formal consultation on transboundary impacts among neighboring countries was implemented, e.g., in the form of specified timeframes for consultation with neighboring countries or joint procedures for international projects [92].

The e ffect of these European environmental assessment approaches on actual policy making, however, has been limited [89]. According to Jordan and Lenschow [29], this hints at a weak impact of environmental policy in general, but can partly also be explained by the lack of clarity as to what these measures are to achieve: Is it to 'strengthen environmental policies, integrate the environment into other sectors, promote public participation or deliver sustainable development' [29]. Furthermore, the attempt to regulate and better inform decision makers by using these assessment tools is often diminished by their improper application or even abuse [28].

### **4. Discussion and Policy Implications**

Given the EU objective to develop a more integrated policy framework that explicitly accounts for the dynamic interplay between di fferent natural resources, the above analysis of current, relevant policy measures can serve as a first foundation. The analysis showed that including the objectives of the WFD into the CAP is currently the most prominent attempt of especially vertical policy integration on a sectoral level. With regard to the energy sector, vertical integration does not seem to play a role at all. Despite the fact that numerous studies have a ffirmed the direct links between energy production and both water resources e.g., [51,53,55,56,93,94] as well as land resources and food production [47,95], the explicit consideration of these cross-resource e ffects is not ye<sup>t</sup> formally integrated into current energy directives or action plans. Thus, overall, the degree of vertical integration among energy and water, as well as energy and land sectors, has been identified as rather weak.

One reason can be seen in the fact that, currently, authority is spread across di fferent actors with mandates for specific sectors, such as energy, agriculture, or water. Within the EU Commission, water policy falls within the responsibilities of the Directorate General (DG) Environment, which o fficially aims at 'greening' other policy areas. The energy and agriculture sectors instead form their own DGs. Such di fferent sub-units usually act based on historically grown—and thus di fferent—sets of formal and informal rules and processes, which can be inconsistent or conflicting, thus making policy integration hard to reach [96]. This fragmented responsibility is in line with our finding that these two sectors show a relatively weak level of policy coherence. Additionally, their uneven legal 'weight' impedes the development of a common and integrated 'nexus-enabled' policy understanding [32]. Whereas the CAP is completely communitized, the energy sector is a policy field of shared responsibilities between the EU and its member states. The EU is thus limited to setting overarching policy objectives in the form of directives, which leave the actual implementation in the responsibility of the member states. The CAP, however, is based on regulations that are transferred directly into national law [97]. The di fferent legal character, therefore, makes it hard to achieve a common level of integration among the nexus sectors. Furthermore, not only the di fferent legal character but also di ffering, partly contesting, underlying values and objectives impede a more integrated policy design. For example, a CAP that aims at economically e fficient and intensive production, which inevitably comes with a high demand of fertilizers as well as a high quantity of manure, undermines or outright contradicts environmental objectives such as low nitrate levels [14].

However, with certain challenges remaining, selected aspects of water managemen<sup>t</sup> have already been successfully integrated into the CAP by binding payments to farmers to conformance with the standards of cross-compliance, which now include standards explicitly related to water management. Given the often trans-national character of externalities related to the provision of water, energy or food, the WFD can serve as a reference for a policy measure that explicitly recognizes the relevance of geographical—rather than national—system boundaries (e.g., river basins) and is thus conceptualized as to provide a mandatory policy framework that leaves the development of specific, trans-national implementation strategies to local authorities. Nevertheless, as the example of Germany shows, the success of these instruments highly depends on the national implementation. In 2018, the European Court of Justice found Germany guilty of failing to reach required maximum levels of nitrate pollution of groundwater [98].

In terms of horizontal policy integration, the impact of existing European roadmaps, as well as impact assessment tools, is strongly limited by lacking enforcement measures. Nevertheless, the cross-cutting procedures of environmental impact assessment, for example, may serve as the conceptual foundation for mandating the explicit analysis of cross-sectoral e ffects among the nexus resources when proposing plans or programs within the EU. The existing variety of approaches—if systematically brought together—can provide a solid foundation for developing a coherent European nexus policy framework.

In order to address the identified remaining challenges of such a policy framework, a combined top-down and bottom-up policy approach is needed that can stipulate cross-sectoral cooperation and the consolidation of political system boundaries with natural-geographic, often trans-national requirements. Important assessment tools in this regard are cross-compliance mechanisms and support for regional development, in line with the approach taken by CAP. Additionally, however, the policy framework must carefully address the challenges of determining a reasonable degree of regulation, which—while framing the path towards its policy objectives—leaves su fficient room for competition-based market structures or locally adapted implementation strategies. Providing universal access to the limited natural resources essential for human life is of indispensable importance. Furthermore, our analysis highlighted that tools and measures are only useful if integrated and used in day-to-day policy making. Specifically, implementation could be improved by streamlining standards and procedures, and formulating clear and specific policy goals. Therefore, a revised and updated EU sustainable development strategy seems to be urgently necessary in order to coordinate policy objectives towards more sustainable pathways. This demand is backed by e.g., the German Federal Government since the last EU strategy was published in 2006 [99]. In this case, the role of member states also needs to be considered. For example, national implementation of EU regulations as well as existing national sustainable development strategies play an essential role in achieving integrated policy making [32]. A central aspect of this is also the promotion of policy learning, e.g., by cross-sectoral working groups as introduced by the Cardi ff-Process, to increase the political will to follow through on integration e fforts.

These policy implications are especially relevant for EU energy policy. The analysis revealed that the energy sector lags behind most in terms of integration. Though horizontal integration approaches exist in the form of overarching strategies, vertical approaches are almost non-existent. Cross-sectoral considerations are lacking beyond very general sustainability considerations. Here, EU energy policy has to clearly and specifically include concerns of and consequences for other sectors as well in order to enable a generally more integrated natural resources policy. Energy policy could especially profit from policy learning, since EU agricultural and water policy have considerable experience in comparison.

With regard to the applied analytical frameworks, each has its limits, but there is also notable overlap. Analysis along the frames of EPI and WEL Nexus revealed that including environmental aspects into other sector policies often remains on a rhetorical level, and that instead, changes to the governance structures are required that transcend the historically grown sectoral boundaries. Here, the Nexus concept enables a broad, comprehensive approach to integrating natural resource policies, while the EPI concept is capable of identifying and evaluating specific governance challenges, although in a sectorally defined context. However, both the EPI and the WEL Nexus concept are considered to allow for various interpretations and implementation options [29,76]. This conceptual ambiguity is di fficult to translate into jurisdiction since it does not define the degree of integration required [29], and thus, is mirrored in the incomplete realization of more integrated EU natural resource policies. Here, synthesizing policy integration and nexus research can o ffer valuable insights, such as considering cognitive factors in the analysis of nexus governance [65]. In this paper, synthesizing both integration concepts provided valuable insight into the di fferent dimensions of integrating natural resource policies. The paper thus o ffers a new method for analyzing integrated natural resources governance

and contributes to the ongoing discussion on finding 'the right level' of policy integration in the European Union.
