Building Incentives for Shared Food Security Data

Food security depends on decisions made across farms, supply chains, research institutions, governments and communities. Those decisions are increasingly informed by satellite imagery, climate records, crop trials, market prices, soil measurements, household surveys and logistics data. Yet valuable information often remains isolated in university servers, corporate systems, government departments or local organisations because the people who collect it see few direct benefits from sharing.

The archived IASSIST 2017 conference at the University of Kansas approached this problem through the idea of “Data in the Middle: The Common Language of Research”. Its programme and presentations on big data, deep learning, digital agriculture and global food security remain useful for understanding how technical systems must be matched with human incentives. Data sharing works when it produces practical value, recognises contribution and protects the people and communities represented in the records.

Why shared data matters for food security

Global food security is shaped by interacting pressures rather than a single shortage. Drought, changing rainfall, soil degradation, pests, conflict, transport costs and sudden price movements can affect production and access at the same time. When datasets are connected, analysts can identify relationships that are invisible within one project. A crop scientist may connect field observations with weather data, while a humanitarian agency may combine market information with household purchasing patterns to locate emerging risks.

Data sharing can also reduce duplication. Several organisations may survey the same region using different methods, while other regions receive little attention. Shared standards and discoverable catalogues help researchers locate existing evidence before commissioning another collection exercise. This saves money and allows scarce funding to reach under-measured communities, especially in low-income countries where research capacity is uneven.

The value is practical as well as academic. Timely information can guide seed distribution, irrigation planning, food import decisions and early-warning systems. A government may use harvest forecasts to prepare storage and transport, while a farmer organisation can use local disease reports to coordinate advice. The quality of these outcomes depends on whether data is available in a form that other people can understand, combine and responsibly reuse.

What encourages people to share information

Researchers often receive recognition for a journal article, grant or conference presentation, but data preparation is rarely rewarded with equal status. Cleaning records, documenting variables, removing identifiers and answering reuse requests can take months. An effective incentive system gives visible credit for those activities through data citations, contributor roles, repository metrics and promotion criteria.

Farmers, Indigenous organisations and community groups require different forms of value. They may share information when it brings better agronomic advice, improved access to finance, stronger bargaining power or a clearer voice in policy. Sharing should not be framed as an obligation to provide free raw material to distant institutions. Participants need to see how their information returns to them through services, findings, training or investment.

Commercial organisations respond to a mixture of financial and strategic incentives. A food processor may share supply data if it improves traceability or reduces waste, while a technology company may contribute tools in exchange for access to high-quality, well-governed datasets. Public funding can encourage participation by requiring data management plans, supporting interoperability and covering the cost of responsible release.

Matching rewards to different contributors

No single reward will suit every participant in a food data ecosystem. A university researcher may value a citable dataset and future collaborations; a farmer may prioritise agronomic feedback; and a ministry may seek reliable indicators for planning. Incentives become more credible when they are negotiated at the start of a project rather than added after data collection.

Contributor	Useful incentive	Safeguard or condition
Farmers and producer groups	Tailored insights, benchmarking, extension support and payment for collection	Clear consent, control over sensitive records and understandable reporting
Universities and research teams	Data citations, authorship rules, grants and promotion credit	Documentation, quality standards and reproducible methods
Governments	Shared indicators, forecasting tools and reduced duplication	Legal authority, public-interest tests and secure access controls
Businesses	Trusted data partnerships, efficiency gains and recognised standards	Commercial confidentiality and limits on onward use
Indigenous communities	Data sovereignty, funded governance and community-controlled benefits	Cultural protocols, collective consent and rights to withdraw
Funders and platforms	Demonstrable impact, reusable infrastructure and stronger evidence	Long-term maintenance funding and transparent evaluation

Payment can be appropriate, but it should not be the only mechanism. A short-term fee may encourage data delivery while leaving contributors without influence over later interpretation or commercial use. Benefit-sharing agreements can create a longer relationship by specifying access to analysis, training, infrastructure, royalties or public services.

Recognition needs careful design as well. A dataset citation is useful only when repositories preserve it, journals accept it and institutions count it. Contributor statements should distinguish collection, curation, software development, governance and analysis. These details help make invisible labour legible and prevent data sharing from becoming an unpaid expectation placed on junior researchers or local partners.

Trust, consent and data stewardship

People are more willing to share information when they understand who will use it, for what purpose and under which conditions. Consent forms written in technical language rarely provide that clarity. Projects should explain potential risks, the period of retention, the possibility of linking datasets and the process for handling future uses. Communication must be accessible in relevant languages and suited to local decision-making structures.

Food and agricultural data can appear harmless while revealing sensitive information. Farm boundaries may expose land ownership, income estimates may create financial risks, and geolocation can identify vulnerable households or endangered resources. Aggregation, controlled access, de-identification and secure computing environments can reduce these risks. They cannot replace governance: a dataset may be technically anonymous yet still identifiable when combined with public maps or local knowledge.

Indigenous data governance deserves specific attention. In Australia, the Maiam nayri Wingara Indigenous Data Sovereignty principles emphasise the rights of Aboriginal and Torres Strait Islander peoples to govern data about their communities, lands and resources. Similar principles apply internationally, even where legal systems differ. Community authority, culturally appropriate consent and agreed uses should shape the full data lifecycle, from collection through publication and deletion.

Making data useful across borders

Sharing a file is not the same as making data reusable. A spreadsheet without definitions, collection dates, geographic references or information about missing values may be impossible to interpret. Common vocabularies, machine-readable formats, persistent identifiers and clear metadata allow a researcher in Nairobi, Canberra or São Paulo to understand what a variable means and whether it can be compared with another source.

Interoperability is especially important in global food research. One project may classify farms by size, another by production system, and a third by household income. Soil measurements can use different units, while crop names vary across languages and regions. Standards should support alignment without erasing local knowledge. A shared framework can include mappings between classifications and preserve the original terms used by communities.

The technical capacity to reuse data is uneven. Researchers in wealthier institutions may have better internet access, cloud services and statistical support than partners in rural or low-income settings. Incentives should therefore fund training, local data centres, translation, software maintenance and long-term curation. Open access is meaningful only when potential users have the skills and infrastructure to work with the material.

The IASSIST 2017 programme offers a useful historical lens here. Its attention to big data and deep learning reflected growing enthusiasm for advanced analysis, while sessions on digital agriculture and global food security pointed towards real-world applications. The lasting lesson is that sophisticated models cannot compensate for weak documentation, limited representation or data systems that exclude the people most affected by food insecurity.

Lessons from Australian agriculture and research

Australia has a highly developed agricultural sector, yet its geography creates distinctive data-sharing pressures. Producers in the Murray–Darling Basin manage water scarcity across vast distances, while farms in northern Queensland face different rainfall patterns, pests and transport constraints. Information about water allocation, soil moisture and seasonal outlooks becomes more valuable when it can be combined across catchments without exposing commercially sensitive farm details.

The local market also rewards better coordination. Supermarkets, independent grocers, exporters and hospitality businesses need reliable information about supply volumes, quality and timing. A lettuce grower near Melbourne or a mango producer in the Northern Territory may benefit from demand forecasts, but may hesitate to share production estimates if the information could weaken their negotiating position. Trusted intermediaries, clear access rules and aggregated reporting can make cooperation safer.

Australian research institutions can strengthen incentives by treating data stewardship as a recognised scholarly contribution. Universities in cities such as Brisbane, Perth, Adelaide and Sydney work with producer groups, government agencies and regional communities that hold different priorities. Funding agreements should support local data custodians, return useful results to participants and provide money for maintaining repositories after a project ends.

Customs and expectations around consultation matter as well. Australian projects involving Aboriginal and Torres Strait Islander communities should allow time for relationship-building, collective decision-making and culturally appropriate review. A rapid online consent process may satisfy an administrative deadline but fail to establish legitimate authority. Respectful governance is an incentive in its own right because it demonstrates that participation will not mean losing control over community knowledge.

Building fairer data partnerships

A durable partnership begins with a data charter. This document can state who owns or governs each dataset, who may access it, which uses are permitted, how contributors are credited and what happens if a project changes direction. It should also describe the benefits participants can expect, including reports, tools, training, payments or policy support.

Funders can strengthen these arrangements through staged requirements. Initial grants might support relationship-building and a data inventory; later payments could depend on documentation, community review and a functioning access process. Evaluation should measure more than the number of files released. Useful indicators include the diversity of contributors, reuse by local organisations, time taken to approve access, data quality improvements and benefits delivered to participating communities.

Platforms need to make responsible sharing convenient. A repository that accepts only complex technical submissions will exclude many small organisations. Practical features include templates, multilingual guidance, simple licensing options, version control, audit logs and support from data stewards. Access categories can range from open publication to registered users, secure analysis rooms or community-controlled access.

Fairness also requires attention to power. A large institution should not set all the terms simply because it owns the servers or controls the grant. Negotiations should recognise that data may be a community asset, a commercial resource or a public good at the same time. Shared decision-making, independent oversight and transparent dispute processes make incentives more credible.

Turning shared evidence into public value

Data sharing delivers its strongest results when it is connected to decisions. A climate dataset matters more when it informs planting advice, insurance design or emergency planning. A household food survey matters more when its findings shape nutrition programmes and price monitoring. Projects should identify likely users early and involve them in deciding which variables, formats and update schedules are genuinely useful.

This approach changes how success is measured. A highly downloaded dataset may have little influence, while a modest regional dataset can improve water allocation, reduce post-harvest losses or help a community secure services. Impact assessments should track practical outcomes alongside publications and citations. They should also record who benefited and who may have faced new risks through disclosure.

The most effective incentives combine recognition, resources, trust and influence. Researchers need credit and support for careful curation. Producers need useful feedback and protection from exploitation. Governments need dependable evidence. Communities need authority over information that concerns them. When these interests are designed into the data lifecycle, sharing becomes a relationship with reciprocal value rather than a one-way request for access.

That principle remains relevant to the themes preserved by the IASSIST 2017 archive. Data sits in the middle of research, policy and practice, connecting disciplines that rarely use identical methods or language. For global food security, the task is to build the governance, infrastructure and incentives that allow those connections to produce fairer decisions, stronger local capacity and more resilient food systems.