Beyond yield maximisation: identifying optimal agricultural productivity thresholds for environmental sustainability across 16 countries
Meserret NalçakanPurpose
Relationship between agricultural productivity and environmental quality has emerged as a critical challenge in sustainable development policy. While agricultural intensification remains essential for global food security, mounting evidence suggests productivity maximization without environmental consideration may be counterproductive beyond certain critical thresholds. This study investigates non-linear relationship between agricultural productivity and environmental quality, representing 74.2% of global corn and 68.7% of global wheat production over 2004–2023.
Design/methodology/approach
Multiple estimation techniques are employed, ensuring methodological robustness. Analytical framework incorporates comprehensive pre-estimation testing, explicit threshold identification using both parametric and non-parametric approaches, and bootstrap-based inference. Cross-sectional dependence is addressed through multiple tests and estimation methods, while second-generation unit root tests and cointegration procedures validate long-run relationships.
Findings
Significant yield thresholds for corn and for wheat are identified, beyond which environmental quality deteriorates substantially. Cross-country analysis reveals substantial heterogeneity in threshold values, strongly correlating with development levels. Developed countries consistently exhibit higher sustainable productivity thresholds compared to developing countries. Findings of this study provide empirical evidence for Environmental Kuznets Curve relationships in agricultural systems, fundamentally challenging the conventional productivity–environment trade-off paradigm. Identification of quantitative thresholds offers sustainable agricultural policy design, suggesting optimal rather than maximum productivity should be the policy objective. Results provide targets for sustainable intensification strategies and emphasise the importance of institutional development and technology transfer in raising sustainable productivity ceilings, particularly for lower-income countries.
Research limitations/implications
While our Agricultural Quality Index represents a significant advance in measuring agricultural sustainability, it may not capture all dimensions of agricultural–environmental relationships. Future research should explore alternative sustainability metrics that incorporate biodiversity, soil health, water quality and ecosystem services more comprehensively. The 20-year time horizon, while substantial, may not capture long-term environmental effects or secular trends in agricultural development. Longer time series data would enable better understanding of threshold evolution and stability over time. Additionally, country-level analysis may mask important sub-national variation in agricultural systems and environmental conditions.
Social implications
From an economic welfare perspective, the ‘optimal' productivity level identified in this study represents the output point at which the social marginal benefit of additional intensification equals the social marginal cost, inclusive of environmental externalities. Formally, this is distinct from the profit-maximising output of an individual farm, which ignores unpriced environmental costs (e.g. nitrate runoff, Greenhouse Gas emissions, biodiversity loss). The welfare-maximising threshold is therefore lower than the private optimum, and the gap between them represents the magnitude of market failure that policy must correct. Countries currently operating below our thresholds face minimal trade-offs: productivity gains are associated with efficiency improvements that reduce per-unit environmental costs. Countries above the threshold face genuine opportunity costs: maintaining food output targets requires accepting higher environmental externalities unless offset by technological improvement or institutional reform.
Originality/value
To the best of our knowledge, this research provides among the most comprehensive multi-country threshold analyses for major food crops using panel autoregressive distributed lag methodology, contributing to a growing body of evidence on non-linear agricultural–environmental relationships.