Governing Food System Data: is the UN Committee on World Food Security Up to the Task?

In 2021, The World Bank published a major report arguing that “the digital revolution has enormous potential for transforming the agriculture and food system in ways not previously seen.” Indeed data-driven technologies—from AI-enabled plant breeding to digital farm platforms to online food retail—are being rapidly brought online in countries around the world.

But who is going to benefit from this so-called transformation?  And how will digital food systems address the immense inequalities that lead to persistent food insecurity? Answers to these questions often pivot on how and by whom data is governed. Next week, the UN Committee on World Food Security (CFS), an inclusive multilateral body established after the 1972-3 food crisis, will begin negotiating a set of policy guidelines on “Strengthening FSN Data Collection and Analysis Tools for Food Security and Nutrition.” The negotiations could not be any more timely. The CFS has a momentous opportunity to address key data governance issues through an inclusive, multilateral process. But whether the CFS rises to the task remains an open question.

The Power of Food Systems Datafication

While modern states have long sought to collect agricultural data in order to regulate and respond to fluctuations in food supply and food prices, by the twentieth century agricultural datafication became a central process through which states sought to exercise power beyond their national territories.

In the 1930s, for example, the United States promoted a project of collecting agricultural data on a global scale in an effort to advance its own export interests and models of agricultural modernization. But in recent years, the analog collection of statistical data has been outpaced by new digital technologies. Machinery, sensors and mobile phones now collect immense quantities of granular data. These technologies have fundamentally reshaped how power is exercised through datafication. As the OECD explains in a recent report, “the real revolution in the sector is the increasing capacity to produce and use data that was not previously technically or financially feasible.”

Datafication today encompasses a wide array of organisms and communities, purposes and activities, and applications across the full supply chain. For example, digital platforms exist to map soil microbial data; to track digital agrifood trading, and to monitor labor on warehouse floors. With the expansion of computational power and ever-shrinking microprocessors, data is now driving digital technologies at multiple food systems scales: from seeds developed on gene chips, to farm operations guided by ‘internet of everything’ sensors, to global information systems whose satellites aggregate scans across millions of hectares. This breadth signifies opportunities for a variety of food system actors: fintech services for farm insurance sellers, smart packaging for food companies, logistics for the transport sector, and e-commerce platforms for retailers.

As actors in this competitive space seek to generate new economic value from data and data infrastructures, they are recognizing the value of datafication in making new connections between previously siloed links in the agrifood supply chain, as well as between this chain and an almost limitless web of biophysical, technical, economic, and social data in other sectors. As a result, datafication of agro-food systems is now being pursued by a diverse range of actors including not only incumbent agri-food corporations, but also tech giants, the financial sector, a growing ‘agri-tech’ sector, governments, and international institutions.

However, for many of these system-wide applications to be useful, developers need more than just granular data; they also require “foundational data”—that is, statistical data on population size, land use, and other information historically collected primarily by the state. Unsurprisingly, we are now seeing a widespread push by corporations, philanthropies, and some governments to encourage states, especially those with few resources for such extensive data collection, to do so.

The Policy Recommendations

The CFS was founded in 1974, in the wake of a major world food crisis. After yet another food crisis in 2008, it was reformed to include the participation of the private sector, other international institutions, and civil society and Indigenous communities who forged an essential and autonomous part of the CFS: the Civil Society and Indigenous People’s Mechanism (CSIPM). The CSIPM includes organizations from 11 constituencies—smallholder farmers, pastoralists, fisherfolk, indigenous peoples, agricultural and food workers, landless, women, youth, consumers, urban food insecure and NGOs—and is today one of the most important spaces for civil society organizations working to eradicate food insecurity and malnutrition. While the CFS remains primarily a multilateral body in which states are the only members, its inclusive approach has been widely touted as a model for democratic global governance.

Developing policy recommendations on “Data Collection Tools and Analysis for Food Security and Nutrition” was first proposed in 2021 by the Bill and Melinda Gates Foundation with support from the United States. From the outset, the CSIPM challenged the narrow framing of the proposal, suggesting that beyond just data collection and analysis, it also needed to address data governance and the impact of digital technologies. Despite these concerns, the narrow frame persisted and CFS’ High Level Panel Experts (HLPE) developed a report on the topic. In March 2023, a draft set of policy recommendations was published by the CFS that included input of all current participants.

CSIPM’s view, however, is that the policy recommendations remain both insufficient and problematically construed. Despite a strong analysis by the HLPE that emphasizes both the significance of data for food security and nutrition, as well as the risks of datafication, the draft policy rehearses a persistent narrative that more data is needed. That is, “data gaps” are to blame for a variety of social and environmental ills.

Developing Data Governance?

The CSIPM has consistently argued that data governance must be at the core of any policy recommendations on data. But what does data governance mean? Globally, divergent perspectives exist among economically powerful states. The US, EU, and China, for instance, exemplify different approaches, with the US pursuing a ‘light touch’ regulatory approach, the EU a market-driven and human-rights oriented approach, and China an approach driven by state security concerns. While in many ways these actors are attempting to export their own regulatory approaches in the context of food systems, they are facing push-back from social movements and civil society organizations for three reasons.

First, food producers have developed a heightened awareness of the potential for being exploited through the data that they themselves generate. They understand their power as data producers—that they make data on their territories, farms, and about their relationships represents and encodes their knowledge practices. And they know this data can be appropriated and used in a variety of extractive and accumulative ways. They are also well-organized into a variety of collective structures, from co-operatives to unions to peasant associations, which are voicing deep concerns about the collection and use of their data. For this reason, large-scale farmers in North America have developed some data governance systems such as the Ag Data Transparency Evaluator through which they seek greater control over their data.

But as food systems in the Majority World (Global South) become increasingly datafied, small-scale food producers are challenging the governance models being exported by the US, EU, and China, none of which offer protection from the agri-food corporations that seek to amass their data and extend their oligopolistic control over global food systems. Therefore, farmers across the Majority and Minority world want new approaches to data governance that protect not only their privacy, but also their sovereignty and autonomy.

Second, civil society and social movements are already deeply concerned about the impact that corporate power is having on environments and communities through climate change, food insecurity, and global biodiversity loss. Industrial food systems now contribute almost one-third of global greenhouse gas emissions, are responsible for rising food insecurity and diet-related illness and are the main cause of the loss of global biodiversity.  Producers and consumers worldwide worry that datafication will only ‘lock-in’ unhealthy and unjust food systems. Yet current models of data governance fail to constrain corporate power and demands for data are increasingly putting pressure on governments to outsource data collection and infrastructure to the private sector.

Third, civil society movements and actors, particularly in Majority World countries, recognize that the US, the EU, and other powerful states continue to wield geopolitical control through expanding food systems digitalization. For example, several Minority World states, via the World Bank, are backing the 50×2030 initiative, a 10-year, ~US$500 million project that aims to increase the capacity of 50 low- and lower middle-income countries to produce “foundational data” by digitizing existing public records, collecting new digital data, developing farmer registries, and building national survey programs for agricultural and rural data. Emphasizing  the importance of the state in building the infrastructures necessary for datafication, as well as in producing foundational data upon which the private sector may draw, the World Bank urges national governments to pursue “‘no regrets’ policy actions” to maximize the benefits of quickly transforming the food system. Forging public-private partnerships to advance so-called innovation ecosystems, however, could easily undermine the sovereignty of Majority world countries and their people in the absence of robust agrifood data governance.

State surveillance and data-enhanced military capacities represent another prominent concern, both in terms of wider geopolitical pressure on countries that resist digital agriculture in favor of agroecological transitions and in terms of technologies entering the farm. Recently, for example, Landus, Iowa’s largest farmer-owned cooperative, announced the launch of Landwerx, an “agricultural innovation hub” that is partnered with Defensewerx, an organization that works closely with the US Department of Defense. The firm’s government partners have included U.S. Special Operations Forces, Department of Homeland Security and the CIA. This is but one example of how surveillance technologies are being rolled out across food systems to monitor farmers and food-chain workers—from the seeds they plant to the hours they work.

Put otherwise, without adequate global governance, deeply corrosive trends in data control and application are set to continue. Smallholder, Indigenous, and other rural communities know this pattern all too well and are now leading the way with calls for a redirect: to treat food system data as a public good, as a knowledge commons, and/or in terms of data sovereignty.

Thickening data governance

Given that the CFS exists to support global policy convergence and governance to ensure food security, developing data governance models that respect and protect the ability of small-scale farmers to feed themselves and their communities should not be controversial. In its report, the HLPE recognized the need for thicker forms of data governance through its model of the “data informed decision-making cycle” at which governance is at the very center.  But some vocal governments, including the United States have continued to insist that data governance should not be included. They argue that data governance should be developed in other more appropriate arenas, ones where they likely have more leverage to assert their propertarian approach to data.

At best this is a call to “forum shift” to other less politicized arenas, such as the UN Statistical Commission. At worst, this is a disingenuous attempt to undermine multilateral decision-making and instead promote forms of data governance advantageous to its own political economic interests. Yet the CFS was reformed exactly for the purpose of establishing inclusive and democratic governance, not to eschew it.

The CFS’ policy recommendations need not detail the exact ways that data should be governed; it is critical that national governments and communities have the ability to determine the regulations and rules that are best suited to their social and ecological contexts. Food producers and other constituencies are also still developing models and approaches to data governance based on changing forms of technology. Nonetheless, the CFS policy recommendations can be a useful resource if they address four issues.

First, the guidelines should clearly distinguish data types by the infrastructures through which data are collected. They should elaborate on data responsibilities and norms accordingly. The draft recommendations fail to distinguish between data collected from public infrastructures, such as national statistical offices, and those collected by the private sector through mobile telephones or soil sensors. These infrastructures entail different forms of consent, accountability, use, and interests. In conflating these different forms of data, the policy recommendations instead provide a pollyanna-ish appeal for “data-informed decision-making,” as if any form of collection that entails timely and granular data will help address food insecurity. But international institutions have recognized that this is far from true.

Public and private sector data collection raise distinct concerns given their differential interests and forms of accountability. Public data collection can support public wellbeing when it is used to ensure adequate food supply, make inequalities visible, and monitor climate and environmental change. But it can also infringe on human rights and can aid discriminatory surveillance. Meanwhile, data collection through digital technologies and private infrastructures raises distinct concerns including informed consent, lack of interoperability, information asymmetries and the expansion of corporate control. When these two infrastructures and forms of data collection are combined, they are most dangerous.

Second, the guidelines must provide clear normative guidance on data governance in the context of food security and nutrition. The CFS’ mandate focuses on food security and protecting the right to food. It has the opportunity to elaborate how, for example, how the six dimensions—availability, stability, access, utilization, agency, and sustainability—of food security should be considered when developing data governance in the context of food systems. What kind of disaggregation is necessary to ensure access? How should individual and collective forms of agency shape data governance processes? Given that ensuring the right to food requires participation in decision-making to uphold the respect-protect-fulfill frame of the right to food, what kind of participation should be required in data collection for food security and nutrition?

Other sources of normative guidance should also be included, such as the UN Declaration on the Rights of Indigenous Peoples (UNDRIP) and the UN Declaration on the Rights of Peasants and Other Peoples Working in Rural Areas (UNDROP), which both elaborate guidance on protecting indigenous local knowledge. These frameworks are important because they specify the rights, duties, and responsibilities of different actors. Standards such as FAIR and CARE can also be useful in designing equitable data governance, but both these and the UN Declarations have been removed from the most recent drafts after the US suggested they were cherry-picked and not multilaterally agreed upon.

Third, the text should provide clear frameworks for the governance of public data. The most current draft has adopted the EU’s approach to data— “as open as possible, but as closed as necessary.” However, this does not address private appropriation of public data. Expansive data is produced and held by the public sector, including agricultural, demographic, meteorological, geographic, biometric, and fiscal information. In recent years, government, industry and civil society bodies, have campaigned for unrestricted and free access to public sector information, “hailing it as indispensable for government accountability and digital innovation, among other causes.”

However, studies show that private sector actors hold the advantage when it comes to utilizing open public sector data, especially in the context of governmental austerity. For example, in the 1990s a coalition of energy industry giants, including Enron, developed a new financial instrument to guard against non-extreme weather events. When this “weather risk” market first developed, traders sought out public meteorology datasets in the US and UK, enabling a multibillion-dollar private market in weather derivatives contracts to emerge on the back of public data. More recently, Uber—valued at USD $82.4 billion in its IPO—relied on London’s real-time traffic data to develop its lucrative (and highly worker-exploitative) ride-sharing app.  It has been argued that market value generated from the exchange of these products and services can generate indirect, if not direct, benefits to the public sector, such as stimulating growth for the economy and thus fostering opportunities for taxation. However, evidence to support such claims is thin, and without clear restrictions around “as open as possible,” public data is likely to constitute a taxpayer subsidy of private agribusiness and ag-tech. Private firms would then accrue open data yet without democratic accountability for how the data is collected or used.

Finally, the draft text does not adequately ensure against monopolization of data by the private sector. Given that the recommendations specifically encourage governments to explore innovative approaches to data collection, including through digital technologies, and that digitalization of food systems is a major goal of many international institutions and foreign donors, it is imperative that the CFS develop clear guidance and how to ensure that digitalization does not exacerbate inequalities. Indeed, numerous studies have already revealed the challenges that digitalization poses with regards to equitable agricultural development. The text does not address issues of intellectual property rights, interoperability, the transparency of source code, or other mechanisms that can prevent lock-ins and ensure that digitalization does not only benefit “first movers” and multi-national agri-tech companies.

The Challenge for the CFS

Transnational data governance is an area of global contestation. Yet few opportunities exist to debate these issues in the multilateral context, where data subjects and data producers are directly involved in decision-making. With these Policy Recommendations on Strengthening FSN Data Collection and Analysis Tools for Food Security and Nutrition, the CFS now has an invaluable opportunity to address what is already emerging as the next frontier of technological struggle—the debate over who has the power to datafy. The CFS should not shy away from data governance, but should directly take up the challenge.

About the project

Places and populations that were previously digitally invisible are now part of a ‘data revolution’ that is being hailed as a transformative tool for human and economic development. Yet this unprecedented expansion of the power to digitally monitor, sort, and intervene is not well connected to the idea of social justice, nor is there a clear concept of how broader access to the benefits of data technologies can be achieved without amplifying misrepresentation, discrimination, and power asymmetries.

We therefore need a new framework for data justice integrating data privacy, non-discrimination, and non-use of data technologies into the same framework as positive freedoms such as representation and access to data. This project will research the lived experience of data technologies in high- and low-income countries worldwide, seeking to understand people’s basic needs with regard to these technologies. We will also seek the perspectives of civil society organisations, technology companies, and policymakers.