From the Ground Up

This report began as an investigation into whether better Environmental, Social and Governance (ESG) reporting and data systems could unlock value for Australian farmers. It reached a broader conclusion. The central problem is not the sophistication of the frameworks, but the direction of the design process. Across Australia and the twelve countries visited in this research, agricultural programs, policy and investment repeatedly failed when institutions defined the problem, designed the solution and then asked farmers to comply. The result was often activity, reporting and administration, but not value at farm level.

ESG provided the clearest example: sustainability and carbon commitments developed at corporate or policy level, pushed down through supply chains without a clear pathway to a valuable outcome on a specific farm. The same pattern held in Australia’s own institutions. Research, policy and finance are all active, but not consistently designed around farmer-defined problems or measured against farm-level outcomes.

Australia does not primarily suffer from a lack of money, capability or goodwill. It already has substantial research infrastructure, public institutions, industry organisations and capital available to agriculture. The deeper problem is that too many of these systems are designed from above. Farmers are often the last consulted, yet the first expected to absorb the cost, risk and implementation burden.

Across the examples in this study where farmer-led design produced genuine outcomes, the same four conditions were consistently visible. Clarity: the problem was specific enough that a farmer could recognise it as real and worth solving. Capital: finance and commercial arrangements were structured to create value for farmers, not simply to extract data or compliance from them. Collaboration: ownership of the problem and the outcome was shared across the parties involved, rather than managed through assembled committees and distant reporting lines. Context: signals moved in both directions. Macro systems shaping decisions on farms, and farm-level reality feeding back into those systems.

The report concludes that no agricultural program, policy, funding mechanism or research investment designed to benefit Australian agriculture will consistently create value for farmers unless farmers are involved in defining the problem and creating the solutions. That is the design test that should sit beneath future reform.

On that basis, the report sets out eight recommendations for redesigning Australia’s existing agricultural architecture so that more value is created on farm:

1. New programs and policies begin with place-based, farmer-recognised problems, mapped before funding decisions are made.
2. Research institutions are accountable for farm-level outcomes, not just activity.
3. Australia’s biggest cross-sectoral problems, the ones no single commodity-based institution owns, are backed by a dedicated outcomes fund, with freedom to take genuine risks.
4. A stronger national "translator" role: a small network of practising farmers connecting Australian agriculture to global trade and policy debates. Bringing what they learn back to farmers, carries signals in both directions instead of just one.
5. Public and private capital reach producers on terms they can actually use.
6. Long-term policy settings give farmers, investors and institutions the stability to act with confidence.
7. Set Agricultural Innovation Australia the independence, resourcing and authority to publish standardised farm-level data, including farm share of the food dollar and natural capital accounting on a fixed schedule.
8. Farmer-to-farmer knowledge networks are enabled, not replaced by top-down extension models that no longer fit how farmers actually learn.

Australia does not need to build an entirely new agricultural system. It needs to change the direction of design inside the institutions it already has, so that programs, policy and investment begin with the people expected to make them work. Ask the farmer first. Then build that.