Do agri-environmental schemes really cut farm emissions?
What evidence from Slovenia tells us about climate policy design
Imre Fertő, Štefan Bojnec
Across Europe, agriculture sits uneasily at the centre of climate policy. Farmers are increasingly expected to reduce greenhouse gas (GHG) emissions, while continuing to deliver food security, rural livelihoods, and environmental stewardship. The European Union’s Common Agricultural Policy (CAP) has responded to this challenge largely through agri-environmental schemes (AES), which reward farmers for adopting environmentally friendly practices.
But a crucial question remains surprisingly under-examined: do these schemes actually reduce greenhouse gas emissions at farm level?
A new study using detailed farm-level data from Slovenia offers a sobering answer. While AES may deliver important environmental co-benefits, their short-term impact on agricultural emissions appears limited. This finding matters not only for climate policy, but also for how we evaluate large, complex policy programmes that combine multiple objectives .
Climate ambition meets policy reality
Agriculture accounts for around 10% of total EU greenhouse gas emissions, primarily through methane from livestock and nitrous oxide from fertiliser use. Unlike emissions from energy or industry, agricultural emissions have remained stubbornly resistant to decline.
AES have long been the CAP’s flagship environmental instrument. Yet historically, these schemes were designed with biodiversity, soil protection, and water quality in mind. Explicit climate mitigation only became a central objective relatively recently, particularly with the post-2023 CAP reforms and the introduction of eco-schemes.
This raises an important concern: are we expecting climate results from policy instruments that were never designed to deliver them?
What this study does differently
Most existing evaluations of AES rely on static comparisons or average effects, which can obscure how emissions evolve over time and whether observed differences are truly caused by policy participation. This study instead applies modern dynamic difference-in-differences methods to farm-level panel data from Slovenia’s Farm Accountancy Data Network (FADN) covering 2014–2021.
Why Slovenia? Despite its small size, Slovenia is a revealing case. AES participation rates are high, farming systems are diverse, and public spending on agri-environmental measures has been substantial. Yet climate-specific targeting has remained relatively weak.
Crucially, farms enter AES at different points in time. The analysis therefore tracks emissions before and after adoption, comparing participating farms with similar non-participants, while explicitly testing whether emission trends were already diverging prior to participation.
Higher emitters self-select into AES
One of the study’s most striking findings is that farms that eventually join AES are already more emission-intensive before adoption. In the two to three years prior to entering a scheme, these farms emitted between 15% and 30% more greenhouse gases per hectare than non-participants.
From a policy perspective, this is not necessarily bad news. It suggests a degree of targeted uptake, with higher-emission farms recognising greater scope—or pressure—for environmental improvement. But it also complicates evaluation: without careful methods, post-adoption changes might simply reflect regression to the mean rather than genuine policy effects.
Little evidence of short-run emission reductions
After accounting for these pre-existing differences, the core result is clear: AES participation is not associated with statistically significant reductions in farm-level emission intensity in the first one to two years after adoption.
Across multiple model specifications, alternative control groups, and robustness checks, estimated reductions are small—typically below 3%—and indistinguishable from zero in statistical terms. This pattern holds across different farm types, including dairy-intensive and livestock-heavy systems.
In short, broad, practice-based AES do not appear to deliver measurable short-term climate mitigation at farm level.
Why null results matter
Null results are often uncomfortable, particularly in policy-relevant research. But they are also highly informative.
First, the findings are consistent with the design of many AES. Measures such as extensive grassland management or organic farming may affect emissions only indirectly, or over longer time horizons. Expecting immediate reductions may simply be unrealistic.
Second, the results highlight a misalignment between policy objectives and policy instruments. When schemes are multi-objective and weakly linked to measurable outcomes, climate benefits may be diluted or obscured.
Third, the study underscores the importance of methodological rigour in policy evaluation. Conventional two-way fixed-effects models—still widely used in applied research—can produce misleading estimates when treatment timing varies. Dynamic designs that explicitly test pre-trends should be the norm rather than the exception.
Implications for CAP reform
What does this mean for the future of agri-environmental policy?
The findings do not imply that AES are ineffective or should be abandoned. Rather, they suggest that climate mitigation requires more precise instruments. Three implications stand out.
First, targeting matters. If the goal is emissions reduction, schemes should focus more explicitly on high-emission sources, such as intensive livestock systems, rather than relying on broad eligibility criteria.
Second, performance needs to be measurable. Practice-based payments are easy to administer, but they provide weak incentives for actual mitigation. Linking payments—at least partially—to verifiable emission reductions or adoption of proven mitigation technologies could improve effectiveness.
Third, monitoring must improve. Despite significant public expenditure, farm-level emissions are rarely tracked systematically. Integrating richer sustainability data into the Farm Sustainability Data Network (FSDN), alongside remote sensing and modelling tools, would allow policymakers to distinguish symbolic compliance from real impact.
Lessons beyond Slovenia
Although the study focuses on Slovenia, its lessons resonate more broadly. Many EU member states share similar policy architectures: high AES participation, diverse farm structures, and growing climate expectations layered onto instruments originally designed for other purposes.
For researchers, the study demonstrates the value of combining micro-level data with modern causal inference tools to evaluate complex environmental policies. For policymakers, it offers a cautionary message: ambitious climate targets require equally ambitious policy design and evaluation frameworks.
From good intentions to measurable impact
Agri-environmental schemes play a vital role in supporting environmentally responsible farming. But when it comes to climate mitigation, good intentions are not enough. Without sharper targeting, clearer performance metrics, and better monitoring, AES risk falling short of their growing climate ambitions.
The challenge for the next CAP cycle is therefore not whether AES can contribute to climate goals, but how to redesign them so that public spending translates into verifiable, cost-effective emission reductions—while continuing to support farmers through the transition.
That is a challenge that demands not only better policy, but also better evidence.
Bojnec, Š., & Fertő, I. (2026).
Do agri-environmental schemes reduce farm greenhouse gas emissions? Evidence from Slovenia.
Science of The Total Environment, Volume 1014, 2026
https://doi.org/10.1016/j.scitotenv.2026.181387
