Bio-based and Applied Economics 13(1): 27-48, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14414 Bio-based and Applied Economics BAE Copyright: © 2024 Le Gloux, F. & Dupraz, P. Open access, article published by Firenze University Press under CC-BY-4.0 License. Firenze University Press | www.fupress.com/bae Citation: Le Gloux, F. & Dupraz, P. (2024). Upscaling environmental incentives in the Common Agricultural Policy: an assessment of the potential of trans- fers from the first to second pillar. Bio-based and Applied Economics 13(1): 27-48. doi: 10.36253/bae-14414 Received: February 23, 2023 Accepted: December 11, 2023 Published: May 20, 2024 Data Availability Statement: All rel- evant data are within the paper and its Supporting Information files. Competing Interests: The Author(s) declare(s) no conflict of interest. Guest Editors: Stefano Targetti, Andreas Niedermayr, Kati Häfner, Lena Schaller ORCID FL: 0000-0003-3645-4381 PD: 0000-0001-9910-1482 Upscaling environmental incentives in the Common Agricultural Policy: an assessment of the potential of transfers from the first to second pillar Fanny Le Gloux1,2,*, Pierre Dupraz1 1 SMART, INRAE, Rennes, France 2 Fanny Le Gloux’s affiliation moved during the peer review process to: Unit Research and Innovation, Directorate-General For Agriculture and Rural Development, European Com- mission, Brussels, Belgium. The information and views set out in this article are those of the authors and do not necessarily reflect the official opinion of the European Commission *Corresponding author. E-mail address: legloux.fanny@gmail.com Abstract. Agri-environment-climate measures and organic farming support have been the main contractual instruments promoting environment-friendly agricultural prac- tices in the European Union since the 90s. They are insufficient in reaching significant environmental improvements, partly because underfunded. Using French panel data from the farm accountancy data network, we evaluate the impact of a budget trans- fer from income support to environmental incentives on contract uptake. We apply a generalised Tobit model to estimate the adoption probability and the acceptable farm- level payment triggering this adoption and simulate a transfer from direct payments to organic farming support and agri-environment-climate measures budget. Results suggest this mechanism increases adoption. Decreasing direct payments affects partici- pation probabilities and acceptable farm-level payments, differently depending on the type of environmental contract, the type of direct payment and the farm technical ori- entation. We evaluate several transfer scenarios and provide ex-ante elements on how it could help reaching the Green Deal organic target. Keywords: common agricultural policy, Tobit model, agri-environment-climate meas- ures, organic farming support. JEL codes: Q15, Q18, Q58. 1. INTRODUCTION The agricultural sector accounted for 10% of the European Union’s (EU) greenhouse gas (GHG) emissions for the period 1990 to 2018 and is the sec- ond largest contributor after the energy sector (EEA, 2020). The continuous intensification of agricultural activities also contributed to natural habitat degradation and dramatic biodiversity decline (Dasgupta, 2021). Behind the concept of agroecological transition lies the idea of moving away from agri- https://doi.org/10.36253/bae-14414 http://www.fupress.com/bae https://doi.org/10.36253/bae-14414 https://orcid.org/0000-0003-3645-4381 https://orcid.org/0000-0001-9910-1482 mailto:legloux.fanny@gmail.com 28 Bio-based and Applied Economics 13(1): 27-48, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14414 Fanny Le Glouxa, Pierre Dupraz cultural practices harming ecosystem services, in par- ticular the systematic use of chemical inputs, towards farming systems maintaining or supporting them (Mil- lennium Ecosystem Assessment, 2005). The EU adopted ambitious environmental targets by 2030 and 2050, in particular on the development of organic farming (OF) to reach 25% of organic agricultural land by 2030. Many levers at various scales can foster this transition. An important one is better targeting agricultural support to make agroecological farming more profitable than conventional farming (FAO et al. , 2021). The Common Agricultural Policy (CAP) represented 36% of the 2019 EU’s budget (58.4 billion euros) (EC, 2019) and is the main EU policy supporting environment-friendly farm- ing practices (Coderoni, 2023). The CAP budget allo- cated to environmental commitments is low in compari- son to income support payments (direct payments of the “first” CAP pillar), the latter including little restrictions on agricultural practices (Dupraz and Guyomard, 2019; European Court of Auditors, 2017; Grethe et al., 2018; Matthews, 2013). Following the definition of the Bio- diversity Strategy for 2030 and the Farm to Fork Strat- egy for the agricultural and food sectors, rethinking the design of the CAP and its instruments is central to trig- gering the large-scale agroecological transition of farm- ing systems (EC, 2020a, 2020c). In this study, we develop a farm-based model- ling framework to assess a reorientation of the direct payments budget specifically towards environmental contracts in France. In the 2014-2020 CAP program- ming period, environmental incentives were offered in two voluntary 5-year contractual schemes of the rural development pillar (“second” pillar) of the CAP: (i) sup- port to OF, and (ii) agri-environment-climate measures (AECM). OF support are area-based payments to eligible farms undertaking a conversion towards OF, or to eligi- ble certified organic farms for maintaining their organic practices. AECM are area-based payments to eligible farms complying with a set of management require- ments targeting an environmental objective such as the maintenance of biodiversity or the improvement of water quality. OF support has proven to be effective in main- taining the relative competitiveness of OF and is a major driver of the sector development (Casolani et al., 2021; Sanders et al., 2011), while AECM are the CAP instru- ments the most targeted towards public good provision (Batáry et al., 2015; Dupraz and Guyomard, 2019; Euro- pean Court of Auditors, 2020; Matthews, 2013). In 2019, direct payments accounted for 69% of the CAP budget (40.5 billion euros), while 8.6% (3.5 billion euros) was allocated to OF support, AECM and Natura 2000 sites altogether (EC, 2019). The literature shows that after 30 years of existence, the voluntary environmental schemes of the CAP were unsatisfactory to improve the state of the environment. The lack and unbalanced funding, as well as poorly designed instruments, led to insufficient participation and effort to reach environmental thresh- olds (Dupraz et al., 2009; Dupraz and Pech, 2007; Espi- nosa-Goded et al., 2013; Targetti et al., 2022; Zavalloni et al., 2019). In 2020, only 13% of the EU’s UAA was under an AECM contract, and 6% under an OF support contract (EC, 2020b, 2020d). Rather than increasing the policy budget to raise environmental incentives, many argued in favour of rebalancing the budget allocation among the various CAP instruments (Dupraz and Guy- omard, 2019; Matthews, 2013). Since the 2014-2020 CAP programming period, Member States have the flexibility to transfer up to 15% of their direct payments budget to increase support to rural development measures, includ- ing OF support and AECM (EU, 2013). In France, 7.5% of direct payments have been redirected since 2017 (MAA, 2021). For the 2023-2027 CAP programming period, it has been decided to dedicate 25% of the direct payments budget to finance a new instrument (eco- schemes) open to all farmers and supporting the volun- tary implementation of environment-friendly measures (generally less ambitious than OF support or AECM contract requirements) (EC, 2021; Runge et al., 2022). Although the negotiations ruled out this option, dedi- cating a higher share of the CAP budget to finance more OF support and AECM was another potential (comple- mentary) lever to upscale environmental incentives and was preliminarily evaluated by (Chatellier et al., 2021). In this paper, we estimate an environmental con- tract adoption model with observed panel data from the French farm accountancy data network (FADN). We propose a generalised Tobit model estimating the adop- tion decision and the minimum farm-level payment trig- gering adoption (“acceptable” farm-level payment). We develop a simulation approach to predict the impact of a budget transfer from direct payments towards the imple- mented environmental contracts during the 2014-2020 CAP programming period: support to OF and AECM. Simulating a budget neutral transfer under ceteris pari- bus conditions, we decrease the direct payments received by farmers and increase the environmental payments to be distributed to OF support and AECM adopters. Our farm-based model estimates are used to predict a new contract uptake outcome in 2019. Our framework does not integrate the market effects of the simulated budget transfers. It means that we assume that induced farm input and output price changes are negligible. We find that the transfer of an additional 7.5% (reaching the maximum transfer rate of 15% between 29 Bio-based and Applied Economics 13(1): 27-48, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14414 Upscaling environmental incentives in the Common Agricultural Policy the two CAP pillars) of direct payments towards AECM and OF support results in an increase of participation in AECM from 11% to 23%, and in OF support from 7% to 15%. The predicted participation rate and UAA under environmental contracts increase linearly with the budget transfer rate simulated. Our model suggests that an additional transfer rate of 15.5% to reach 23% of transfer between the two pillars would allow to reach the Green Deal target of 25% of organic UAA. We observe an indirect effect on farmers’ behaviour of decreasing direct payments. In particular, the probability of partici- pating in AECM significantly increases with the amount of coupled payments for suckler cows received at the farm level (+0.1% per 1,000€). We also estimate a strong positive effect of decoupled direct payments on OF sup- port acceptable farm-level payments (+1,039€ per 100€/ ha), such that our model predicts that farms participate in OF support for lower farm-level payments after the budget transfer. We identify a differentiated impact of the budget transfer according to the type of farm, with an increased incentive for farms specialised in grazing livestock to contract AECM, and for farms specialised in cereal and field crops, permanent crops, dairy, pigs and poultry or mixed farming with field crops and grazing livestock to contract OF support. Our first contribution is an ex-ante evaluation meth- od of the transfer mechanism from direct payments to environmental contracts. In particular, we model the impact on adoption. To our knowledge, the effect of such a budget transfer has not yet been assessed at the farm level and for an allocation targeting environmental con- tracts specifically. Previous ex-ante evaluations of the reorientation of direct payments used the CAPRI (Com- mon Agricultural Policy Regionalised Impact) partial equilibrium model (Himics et al., 2020; Schroeder, 2021; Schroeder et al., 2015), or linear programming (Gian- nakis et al., 2014), to study the impact on environmen- tal indicators aggregated for farm types and EU regions. Hence, it remains unsure how effective it can be to sig- nificantly increase the voluntary adoption of environ- ment-friendly practices at the farm level, and what are the underlying microeconomic mechanisms. Adoption results from the confrontation of the supply of envi- ronmental commitments by farmers (farm and farmer characteristics, opportunity costs), and the demand from public authorities (budget, eligibility criteria, technical requirements, payment). Our model partly overcomes the absence of information on the diversity of contract characteristics and eligibility rules by controlling for many factors of farm heterogeneity. Our second contribution is to capture the effect of direct payments on both the environmental contract adoption decision and the associated acceptable farm- level payment in France under the 2014-2020 CAP framework. Beyond a direct positive effect on the par- ticipation of an increased budget available to finance environmental contracts, one can expect an indirect effect of the transfer on farmers’ response to environ- mental incentives, resulting from the decrease of direct payments (lower income support). Monetary aspects from different sources, including direct payments, are important drivers of the decision to adopt AECM and OF (Darnhofer et al., 2019; Jaime et al., 2016; Sanders et al., 2011; Van Herzele et al., 2013). Allaire et al. (2011) and Pufahl and Weiss (2009) found different effects of direct payments coupled to production on participation in AECM, with an overall positive effect in Germany, and a marginal or negative effect in France for exten- sive grassland measures. Moreover, a positive effect of the decoupling of direct payments on the adoption of OF was found in Sweden (Jaime et al., 2016). This lit- erature proved that both direct payments and environ- mental payments affect the decision to adopt environ- ment-friendly practices, showing the importance of con- sidering direct and indirect effects when evaluating the potential of a budget transfer in boosting more adop- tion. In our study, we complement previous studies by looking at the effect of direct payments on not only the adoption decision, but also the amount of payment to allocate to farms to trigger this adoption. The paper is organised as follows. Section 2 presents the data, theoretical framework and econometric model of environmental contract adoption, and the procedure to simulate a reorientation of the CAP budget. Section 3 describes the estimated econometric models and pre- sents the predicted results. Section 4 discusses the meth- odological approach and the findings. Finally, section 5 draws some conclusions. 2. MATERIALS AND METHODS Our methodological approach to simulate a change of CAP budget allocation comprises three steps: 1. Estimation of the model of voluntary contract adop- tion under the current budget allocation. 2. Prediction of new probabilities and acceptable farm- level payments with a reduction of direct payments. 3. Starting from the farm with the highest probability to participate, allocation of the initial instrument budget plus an additional amount from the direct payments budget to participants, up to their estimat- ed acceptable farm-level payment, until the budget is exhausted. 30 Bio-based and Applied Economics 13(1): 27-48, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14414 Fanny Le Glouxa, Pierre Dupraz In this section, we present how we applied this meth- odological approach using observed French data from the 2014-2020 CAP programming period with two types of environmental contracts: OF support and AECM. 2.1 Data The French Metropole FADN data for the years 2015 to 2019 were used in the study. The data represent an unbalanced panel of 36,251 farm observations and include information on the total farm-level payment (€) received for AECM contracts on the one hand, and OF support contracts on the other hand. The dataset does not include information on the surfaces enrolled in each contract type, nor on the specific measures adopted, but knowing the organic certification and organic conver- sion status of the farms allows us to identify whether a recipient of OF support has a conversion OF sup- port contract or maintenance OF support contract. The national FADN is designed to be representative of medium and large farms contributing to more than 90% of the gross production and utilised agricultural area (UAA) and covers the scope of 65% of all farms (Agreste, 2022). This data source is therefore particularly relevant for ex-ante CAP evaluations. From 2015 to 2019, a total of around 1.6 billion was allocated to the farms of our FADN sample for engag- ing in AECM and OF support (Table 1). The highest budget was for 2019, with 228 million € to 11% of sam- ple farms for AECM, 66 million € to 1.5% of sample farms for conversion OF support and 138 million € to 5% of sample farms for maintenance OF support. For that same year (2019), the French Government reported allocating a total of 244 million € for AECM, 191 mil- lion € for conversion OF support and 58 million € for maintenance OF support (DDT Ariège, 2020). In terms of participation rate, it corresponds to around 11% of metropolitan farms having contracted an AECM, 5% conversion OF support and 3% maintenance OF sup- port (DDT Ariège, 2020; INSEE, 2022). Hence, the FADN sample describes the allocation of 93% of the Table 1. Common Agricultural Policy budget and beneficiaries in 2015-20191. Year Direct payments Decoupled direct payments Coupled direct payments for suckler cows AECM OF support Conversion OF support Maintenance OF support Budget (million €)2 2015 7,288.4 6,095.6 667.5 165.3 122.5 23.6 99.0 2016 6,955.9 5,781.6 631.5 136.5 123.5 18.7 104.8 2017 7,124.9 5,880.6 651.6 159.4 140.0 19.4 120.6 2018 6,727.5 5,576.2 623.7 189.7 147.2 30.6 116.6 2019 6,676.0 5,561.1 655.0 227.9 203.3 65.7 137.6 Beneficiary farms (%) 2015 85.7 84.1 24.9 6.2 5.6 0.9 4.7 2016 85.5 84.2 25.2 6.6 5.8 0.7 5.1 2017 85.1 83.6 25.9 7.8 5.6 0.6 5.0 2018 85.6 84.4 25.8 8.9 5.3 0.8 4.5 2019 85.3 84.1 26.5 10.8 6.9 1.5 5.4 Beneficiaries’ UAA (%) 2015 97.9 97.3 34.0 8.8 4.2 0.8 3.5 2016 98.4 98.1 34.1 9.1 4.4 0.7 3.8 2017 98.4 98.0 35.3 10.3 4.9 0.5 4.3 2018 98.4 98.1 35.0 11.9 4.9 0.9 4.0 2019 98.6 98.1 35.8 14.5 6.3 1.6 4.7 AECM: agri-environment-climate measures. OF: organic farming. 1 All figures are weighted by the extrapolation coefficient of each observation. 2 To compute the total policy instrument budget for year t, we corrected for delayed payments distributed at year t+1 or t+2. Less than 0.2% of the direct payments were distributed at t+1 for 2015, 2016, 2017, 2018 and 2019, and at t+2 for 2015 and 2017. Less than 8.0% of the AECM and OF support payments were distributed at t+1 for 2018 and 2019. We could not correct for 2019 instrument budgets distributed in 2021 (data not available at the time of the study). Source: 2015-2020 French FADN data. 31 Bio-based and Applied Economics 13(1): 27-48, 2024 | e-ISSN 2280-6172 | DOI: 10.36253/bae-14414 Upscaling environmental incentives in the Common Agricultural Policy AECM budget and 82% of the OF support budget to a representative ratio of participants/non-participants in 2019. However, it does not represent well the repartition between conversion OF support and maintenance OF support and overestimates the allocation of OF support to certified farms relative to farms in conversion. Yet, we observed the ratio within the OF support eligible population (i.e. farms converting to OF or already certi- fied in 2019) is well represented in the FADN, at least when it comes to the utilised agricultural area (UAA) (see Appendix A1) (Agence bio, 2020). 2.2 Theoretical model of voluntary adoption of an environ- mental contract For a given type of environmental contract (AECM on the one hand, and OF support on the other hand), we represent the demand for environmental commit- ments from authorities during a CAP programming period by a function θ(M,B,Γ) describing a set of meas- ures M (the diversity of technical requirements belong- ing to the contract type), a total budget B, and policy parameters Γ defining exclusion rules. For OF support contracts, M includes a diversity of measures designed for specific land use, and either for maintaining organic practices (maintenance OF support) or for converting to organic practices (conversion OF support). For AECM contracts, M includes a diversity of measures designed for a specific land use and generally an environmen- tal target (water quality, biodiversity…). In France, not all farmers are eligible to AECM contracts and main- tenance OF support contracts. The exclusion rules are based on the location of the farm and described by Γ. The confrontation of demand and supply of environ- mental commitments results in an uptake equilibrium such that B=∑iPi(M,Γi,ai,ki,ei). With Pi the farm-level pay- ment allocated to farms, Γi whether the farm is eligible to the environmental contract type (location in the eli- gible area), ai the farm characteristics affecting eligibility to a subset of environmental measures of M (location, land use, organic certification status…), ki other farm and farmer characteristics (economic size, surface, age, education, technical orientation…), and ei the farm eco- nomic context (market prices, CAP support, etc). We assume the supply of environmental commit- ments by farmers is driven by the profitability of adop- tion and eligibility. In practice, the payment for an envi- ronmental contract is delivered as a payment per hectare enrolled, and for most measures, the farmer can decide to enrol all or part of his/her farmland. However, the binary adoption decision (participation vs. no partici- pation) is made at the farm level. Therefore, we assume the farmer decides based on whether the total farm-level payment received for enrolling his/her profit-maximising amount of farmland in an environmental contract is suf- ficient to make participation profitable. The decision Di* of farmer i to participate and the binary participation Di are defined as follows: (1) With mi*∈M the characteristics of the measure(s) adopted by the farm (technical requirements, payment per hectare), Φi≥0 the maximum farm-level payment the farm is eligible to for adopting mi* on all eligible surfac- es, and Pi*>0 the minimum farm-level payment trigger- ing the adoption of mi* (acceptable farm-level payment) by the farmer. mi*=mi*(M,Γi,ai,ki,ei) is the optimal con- tract uptake and the solution to the profit maximisa- tion programme of farm i. If ∀mi∈M,Φi(mi,Γi,ai,ki,ei)=0 or 0<Φi(mi,Γi,ai,ki,ei)