Serious games can unlock farmer uptake of land‑use decision tools by making model assumptions transparent, demonstrating profitability–emissions trade‑offs, and fitting into farm workflows; however, the evidence is preliminary, coming mostly from small pilots and requiring larger field trials and validation to confirm real‑world effects on land use and emissions.

Main Finding

Serious games can help overcome key barriers to uptake of land‑use decision support tools (DSTs) among farmers and can also be used as a participatory method to design better, more trusted DSTs for achieving land‑use net zero. By supporting co‑design, visualisation of outcomes, experiential learning, and integration with advisory networks and other tools, serious games can increase farmer understanding, change norms, and better link environmental outcomes to farm profitability — all critical for realistic, durable land‑use transitions.

Key Points

• Net‑zero targets will require large-scale land‑use change and behavior change by farmers; actions vary regionally (e.g., forestry, agroforestry, reduced livestock intensity).
• Farmer decision making is complex: economic rationales, behavioural psychology (e.g., Theory of Planned Behaviour, prospect theory), and cultural norms (the “Good Farmer”) all shape choices. Short‑term financial incentives alone often fail to produce sustained change.
• Decision support tools (DSTs) for agriculture exist (carbon calculators, landscape planners, process models such as JULES, E‑Planner, AgLand, etc.) but uptake is limited due to:
  • usability and time costs,
  • perceived lack of relevance to individual farms,
  • trust and credibility concerns (including suspicion of “free” tools),
  • need for specialist knowledge to interpret outputs,
  • poor integration with farm advisory systems and routines.
• Serious games are defined as games with purposes beyond entertainment. In environmental domains they have been used for flood, water, and nexus management and can:
  • act as DSTs by enabling interactive exploration of land‑use choices,
  • visualise future landscape and economic/climate outcomes,
  • provide low‑cost experiential learning and rehearsal of decisions,
  • surface social dynamics and peer norms through multiplayer/roleplay formats,
  • serve as a co‑design method to elicit farmer preferences, local knowledge, and data to calibrate DSTs and models.
• Specific ways serious games can improve DST uptake: co‑design with farmers to ensure relevance; build trust through transparent mechanics and data provenance; demonstrate links between environmental actions and profitability; use intuitive visualisations to reduce interpretation burden; and connect to advisors and existing tools (e.g., carbon calculators, farm management systems).
• Challenges remain: ensuring games are credible and accurate, integrating them with high‑quality data and models, avoiding oversimplification, and addressing digital access/skills gaps (age, training).
• Emerging directions include hybrid systems (serious games + AI/ML models), agent‑based and spatial simulation integrated into game interfaces, and using games to generate behavioral data to improve DST predictive models (ADD‑TREES is an example of AI‑supported DST development).

Data & Methods

• This chapter is a peer‑reviewed literature and conceptual review (no new primary empirical dataset presented).
• Methods: synthesis of academic literatures on farmer decision making, DST design and uptake, and serious games in environmental and agricultural contexts. It draws on theoretical frameworks (behavioural psychology, Theory of Planned Behaviour, cultural concepts like the “Good Farmer”) and applied examples of DSTs and serious games.
• Examples and case tools discussed include E‑Planner, JULES, AgLand, Farm Carbon Toolkit, Agricalc, ADD‑TREES, and literature on serious games applied to flood, water, and nexus management.
• Uptake factors are discussed using a framework that categorises influences into core, modifying, enabling, and driving factors (from Rose et al., 2016; applied to environmental DSTs by Urquhart & Goodenough, 2023).

Implications for AI Economics

• Value of integrating AI into DSTs and serious games
  • Personalisation and targeting: ML can tailor recommendations to farm characteristics, increasing perceived relevance and adoption likelihood.
  • Improved predictive accuracy: AI can combine heterogeneous spatial, temporal, and management data to better estimate sequestration, yields, emissions, and economic outcomes used in games/DSTs.
  • Real‑time feedback and scenario exploration: fast surrogate models/ML emulators can make complex simulations tractable inside interactive games.
• Serious games as a data source for economic modeling
  • Games can generate revealed‑preference or behavioural response data under controlled, policy‑like scenarios (useful for estimating elasticities, adoption thresholds, risk preferences).
  • Multiplayer/role experiments can illuminate peer‑effects and social norm dynamics relevant for diffusion models and policy design.
• Evaluating economic impacts
  • Key econometric and welfare metrics: adoption rates, changes in land‑use allocation, farm profitability, distributional impacts across farm sizes/regions, carbon sequestration delivered, and cost‑effectiveness per ton CO2e.
  • Consider dynamic effects: learning, habit formation, and persistence of behaviour change beyond the game/initial adoption.
• Market and policy design implications
  • Credible measurement, reporting, and verification (MRV): AI‑assisted DSTs and games that improve farmer understanding of MRV can increase participation in carbon markets or payment schemes.
  • Incentive alignment: combine informatics (games/DSTs) with appropriate economic incentives and advisory mechanisms to convert awareness into durable action.
  • Avoid perverse incentives: careful design needed so AI recommendations do not produce leakage, gaming of payments, or risk‑exacerbating choices.
• Design and deployment cautions for AI economists
  • Explainability and trust: opaque AI outputs reduce trust; serious games’ transparent mechanics can help communicate how models produce outcomes.
  • Access and equity: digital divides (age, connectivity, skills) may bias who benefits; consider hybrid delivery (offline, advisor‑mediated).
  • Data privacy and governance: farm data used by AI/DSTs raises ownership, sharing, and liability issues — implications for participation and market functioning.
  • Validation: economic models and AI predictions must be validated against field outcomes; games should not over‑simplify key uncertainties (e.g., carbon permanence).
• Research opportunities for AI economists
  • Use serious games to estimate behavioural parameters (risk aversion, loss aversion, social conformity) for calibrated agent‑based or equilibrium models of land‑use change.
  • Evaluate the marginal value of improved information (via AI/DSTs) on land‑use decisions and carbon outcomes.
  • Study distributional effects and optimal subsidy/payment designs when combined with personalised AI recommendations.
  • Develop metrics for trade‑offs between ecological goals and farm incomes under realistic adoption constraints.
  • Assess impacts of co‑designed AI tools on adoption speed and cost‑effectiveness of achieving land‑use net zero.

Limitations - The chapter is a review and conceptual piece rather than an empirical evaluation; claims about effectiveness of serious games are grounded in literature and examples but require empirical testing in farming contexts. - Much of the discussion is UK‑centred, though international examples are referenced; local institutional, market, and cultural variation matters for generalisability.

If useful, I can: - Draft a list of concrete econometric or experimental designs to estimate behavioural parameters using serious games. - Propose evaluation metrics and an impact evaluation plan for an AI‑enabled serious game pilot aimed at farm net‑zero decisions.