The Data Governance Deep-Dives held by Aapti institute and the Digital Commons Networks aims at broadening a discussion around data stewardship. Models of data governance, among other data governance aspects, are discussed in a closed-door meeting of professionals, practitioners, researchers, and organizations. The sixth session of the Data Governance Deep-Dives explored issues around data governance in the agricultural sector. This was centered around a presentation by a non-profit organization -working for the empowerment of farming communities through technological interventions -on their data governance model for a data exchange platform currently under development.

As an organisation working at the intersection of agriculture and technology, they collected various datasets on farmers, farms, and farming activities. To leverage this data for the improvement of the livelihood of farmers, they created an open-source protocol. The protocol sought to give greater power to farmers over their data created at various points in the production cycle. Farmers are given the controls to share their data with different actors in the data ecosystem directly, ensuring safe use of and agency over their data to avail preferred services.

The discussion involved key points on data sharing -trust-building, farmer incentives, usage safety of data, and stakeholder engagement through the platform in the agriculture sector.

Protocol to data governance – an alternative to platforms?

In simple terms, protocols are a set of rules for standardizing communication. HTTP (HyperText Transfer Protocol) is an example of a protocol used for communication between web browsers and web servers. Protocols in the agricultural sector, with adequate controls, can be used to enhance data governance by facilitating communication between different platforms in the ecosystem. This allows sharing information from the data provider to the data consumer through peer-to-peer connectors with appropriate usage policies and rules. This is different from other platforms -such as Amazon and Uber-which are run by a single centralized authority and have a water-tight policy and rules framework. Protocols ‘enable’ communication between existing platforms and applications. A protocol, therefore, solves broad data ecosystem problems of trust and configuration which a centralised platform cannot adequately solve.

Participants within the protocol

Farmers are the primary data providers. They share their data with organizations -participants. These participants carry out the actual exchange of data or provide services to the farmers. These participants are allowed to be a part of the protocol network after authorization by a steward. This steward runs the protocol from a specific geographical location or jurisdiction.

Usage policies and their role in the data governance model

An important element suggested in the data governance of rural farmers was the implementation of customizable usage policies. Through these configurable usage policies, the protocol aims to build trust in data providers/farmers by allowing them to decide how, why, and when their data should be shared and used. Farmers seemingly exercise complete control over their data This model suggested configurable usage policies and granular consent provisions safeguard the farmer’s interest over their data. Data wallets are a part of these usage policies.

Data Wallets

Another key element introduced by the organization was data wallets. Open Data wallets are application-based data containers that allow data owners to keep their data secure and within their control. These wallets primarily function as consent managers at the farmer’s level. They are included within the framework of the protocol’s usage policies. The farmer can consent to share data collected by an organisation with other organisations within the ecosystem. Data wallets are open as they allow data providers to give granular consent on factors such as the purpose or use of their data.

Role of Connectors for data sharing

Connectors are applications or software that data providers and the data consumers have to download for peer-to-peer sharing of data. Connectors are used to avoid third-party interference in the data governance process. The protocol works efficiently through these connectors.

Role of the Steward

Organizations have to register with a ‘steward’ or a central body. This steward acts as a licensing authority for the participants in the protocol. The steward provides authentication through subscription, to allow safe exchange and use of data. Participants will be ‘trusted’ only after receiving a subscription from the steward. The steward is, therefore, a gatekeeper that permits participants to be a part of the trusted data users. The steward could be a ministry or a consortium of researchers or professionals. Through the presentation, the organization underlined the power of the steward as a clearinghouse and a gatekeeper that allows or disallows organizations to be a trusted part of the protocol network.

Need for such protocol

The presenting organization highlighted a threefold benefit of the protocol

1. Currently, at a system level, farmers approach organizations in different siloes to receive services or benefits. A peer-to-peer sharing protocol would connect these siloed organizations, allowing easy access to services for farmers.
2. Presently, farmers lack trust in organizations, given the possibility of data misuse. Through this protocol, trust and collaboration could be enhanced, as organizations must be authenticated, must comply with the usage policies, and share data only after being consented to by the farmers. Through the discussion, a question on measures taken by the organization for building trust and data literacy was raised. Answering the trust-building measures, the organization has adopted a three-fold solution: No third-party data storage or control, the power given to farmers to customize usage policies for data sharing and modify their consent to usage and sharing of their data. According to them, data literacy at a farmer level is still lacking, but organizations in the agriculture sector take compliance with data protection policies on priority.
3. The problem of lack of control over data is solved as farmers control their own data whilst the protocol ensures complete and safe usage of their data through features like the usage policies and data wallets. Farmers gain bargaining power and receive better access to benefits and services through an equitable protocol network.

Community Governance of Data

Farmers share their data directly through connector applications with the data users -third-party organizations/participants, to receive better services, benefits, and incentives. This sharing maximizes the usage of data. Organizations use the data for their business models in return providing benefits, services, or incentives to the farmers, as decided on in the usage policies. Each organization has to comply with the usage policies and share the data only if consent is acquired from the data-providing farmer. The participant organizations earn the trust of the farmers by trading consented data in return for better services for farmers.

The interoperability of data between organisations of the ecosystem -that carry out different functions -allows users (farmers in this case) to avail better information and services in a localised manner. Similarly, this data can also be used for certification from agencies for users to realise better prices.

Questions that remain

• Need for data literacy
  The burden of consent and configuration depends on the farmers and their scope of knowledge. Hence, the feasibility of this model for data governance is still to be tested from an understanding of how farmers will take control, make decisions, and informed choices on how their data will be used.
• Identifying Data Stewards and their role
  In community-based consent models, it is important to ensure that data stewards best represent the interests of the community. It is, therefore, crucial to explore how potential data stewards would reflect this. Information disclosure and transparency are key to ensuring this. For effective, comprehensive consent controls, it is important to define the terms for use of data.
• Adequate Recourse Mechanism
  Using the data whilst complying with the usage policies needs monitoring; violation of such policies requires actionable deterrents. It is still to be seen how a suitable recourse can be provided to data providers ( farmer community) for breaches on the part of data consumers (organisations).
• Compliance with data policies
  A steward requires cross-jurisdictional and geographical knowledge on changing data policies, whilst simultaneously supporting the community in safeguarding their data rights. Stewards must unburden the community from granular compliance considerations, It is yet to be seen how stewards in this model will fulfill this role.