By Sushmitha Viswanathan

Introduction

Aapti Institute has been interested in the potential of data cooperatives as a model for data stewardship. To explore the viability of converting an existing cooperative into a data cooperative by adding a data layer, we have been working with Megha Mandli, a women’s farmer cooperative in Gujarat. Our first visit to Megha Mandli involved a scoping of the key issues faced by the cooperative and its members as well as their main priorities. Through this we learned that a lack of access to credit is hampering the cooperative’s ability to scale. Based on these insights, we formed a hypothesis that adding a data layer on Megha will facilitate access to credit from digital financial service providers. This in turn is based on the premise that the aggregation of personal, farm and financial data of all member-farmers could demonstrate credit-worthiness. To test the viability of our hypothesis, we made a visit to Megha. The agenda of this visit was to gain an understanding of Megha’s current data collection practices and to identify the kinds of data the members are comfortable in sharing. Insights from this visit are crucial to structuring co-design workshops to build the data layer for Megha.

Insights from the field

The members explained that the inability to access credit affects membership, drives up operational costs by negatively impacting logistics, lowers opportunities for members, affects their ability to purchase quality inputs and hinders engagements with experts. Therefore, they saw value in our hypothesis and were eager to understand how pooled data can facilitate access to credit.

It is relevant to note that the concept of data collection is not alien to Megha. The cooperative is presently gathering data such as name, village, mobile number, output, farm size, Aadhaar, farming input needs, input taken and bank details of the Aagewan. Data is collected in a register and is then entered into an excel sheet. While the frequency of such data collection remains unclear, we are aware that the data collection is mostly facilitated by the Aagewans, who are the community leaders for one or two villages.

Megha has fostered a culture of actively exchanging information and tips on farming through the cooperative’s WhatsApp group. We also found out that the WhatsApp group is used to connect members with ladies who could be potential members.

To identify the kinds of data the members are open to sharing, we circulated a questionnaire consisting of a list of illustrative data points. The 20 members who were present at the meeting were required to indicate whether or not they would be comfortable sharing data on the points listed.

We observed the following from the responses:

1. Financial data: Farmers are reluctant to share financial data as they believe that if their financial status came out in the open, government benefits such as ration cards would be revoked. Several members also did not want to disclose Aadhar details due to its linkage with their bank accounts.
2. Personal data: Members were generally open to sharing their personal data. Several members, however, were hesitant to share their Aadhaar details and health records.
3. Farm data: Members were willing to share data related to farm size and information, input quantity and the kinds of crops they harvested. However only around 50% of the members were open to sharing details of their crop yield and price paid at harvest. We believe that their reluctance stems from the fact that the cooperative presently has limited engagement in the output side. It is also relevant to note that the majority of the members present indicated that they are unwilling to share insurance details and details relating to their farmer certificate.

Identifying opportunities for delivering value to Megha

Gaining an understanding of the issues faced by the members of Megha was instrumental to identifying how exactly adding a data layer would help the members. Although our hypothesis revolves around access to credit, the member-farmers were forthcoming in identifying other routes through which pooled data could deliver value to them. These routes fall into the following categories:

1. Knowledge creation and information sharing
2. Insights for planning future action
3. Demonstration of state of affairs

1. Facilitating access

• Credit: Based on the challenges highlighted by the members of Megha, we gathered that pooling data to obtain finances is crucial to facilitate the following:
  - Purchasing more quality inputs
  - Purchasing more quality inputs
  - Purchasing sufficient inputs to meet the demands of all member-farmers
  - Making up-front payments to farmers for their outputs
  - Cover high transportation costs for procurement of output
  - Investing in infrastructure to store output locally
  - Attracting new members
• Insurance: Most farmers have no crop insurance, primarily due to lack of awareness and documentation. Lack of insurance has prevented them from claiming compensation for crop loss from the government. The members believe that collecting and pooling data on loss could be beneficial to getting compensation in the future. It is also worth exploring how data can help with proper documentation to register for crop insurance schemes.
• Government schemes: The members expressed that pooling data is likely to help Megha and its individual farmers claim benefits under government schemes.

2. Building trust:

Concerns on sharing data mainly spring from the lack of trust in the data aggregation process and the digital environment, in general. We believe that once the members begin to realise the benefits of capitalising on their aggregated data, they are likely to place more trust in the process of data pooling, which in turn would encourage them to collect more data. We believe that the trust which is inherent in the existing cooperative structure can be leveraged to create an environment of trust surrounding data.

3. Creating linkages

• Suppliers: Members expressed that data could simplify the process of identifying and selecting potential suppliers. Presently, the members are sometimes unable to procure quality inputs. They believed that pooling data could make it easier to connect with suppliers and enhance bargaining power with respect to prices and quality. They envision that using data, Megha Mandli will eventually become a one-stop platform to procure high quality inputs.
• Market: Members hope that data could link Megha to customers, similar to what is being done in Ucchal district. In the future, it could also facilitate their entry into new markets.
• Members: Megha has been unable to attract members due to low initial investment. They are positive that enhanced visibility of Megha’s activities due to digitization and data collection is likely to attract new members to join the cooperative.

4. Empowering members to make decisions on data and its use:

We expect that adding a data layer over Megha through a co-design process will empower the member-farmers in two ways:

• Members are empowered to make decisions on the purposes of data collection and the use of their data
• Aggregated data will empower members to make business and farming decisions that are in line with their economic and social goals

5. Enhancing efficiency

• Resource management and business planning: The members believe that pooling and sharing data can help them save time by reducing transaction costs and information asymmetry, track prices (input and output), and ease logistical hurdles. They also hope to use data to make business decisions and identify focus areas for expansion.
• Farming insights: Season-based insights on requirement and use of inputs such as seeds and fertilisers is another avenue for using data. Personalised insights based on the crop grown, to maximise yield is also an opportunity worth exploring.

6. Building knowledge

• Expert engagements: The members of Megha are keen about being empowered with the knowledge required to improve farming practices. They expressed their interest in using data to identify and locate experts who can help with evaluating seed quality.
• Knowledge repository: The data cooperative model could facilitate the creation of a repository for knowledge on farming, which is currently being done through WhatsApp. This can also be done for traditional knowledge on the cultivation and preservation of crops as it would not only benefit other members, but also help in preserving this knowledge for the future generations.

Conclusion

Although there is plenty of potential in adding a data layer around Megha, a few challenges have to be addressed for our endeavour to materialise. As evidenced by the responses to our questionnaire, trust continues to be a major hurdle. Since the farmer-members are uncomfortable with sharing financial data, it is pertinent to identify digital financial service providers who are likely to grant credit using alternative data. Since data cooperatives are founded on the values of trust and mutual help, they are effective channels to negotiate with DFSPs on behalf of the member-farmers. Furthermore, since our proposition rests on the premise that the loan will be disbursed to the cooperative, and not individual members, the scope of individual members obtaining personal loans using their data is likely to be limited.

An added dilemma is the need to incentivise data collection. Since data collection and sharing is a time-consuming process which could negatively impact the daily income of a farmer, incentives to collect data need to be identified.

Despite these challenges, our visit has impressed upon us that the member-farmers are not only enthusiastic about adding a data layer around Megha, but are aware about the exact avenues in which data can be used. To move ahead with the co-designing process to build the data layer, it will be essential for us to build greater trust with the members around data sharing and also identify pathways to incentivise data collection.

With these insights in mind, our immediate goals are as follows:

• Identify data flows and appropriate consent models through the first co-design workshop
• Identify DFSPs which will issue credit using alternative data
• Ideate on what incentives to collect and share data should look like
• Explore partnerships with grassroot organisations working on digitization in the project site.

By Shefali Girish

Executive Summary

Gender digital divides have aggravated the already existing inequalities in our society. The lack of control by women over their data has hindered their ability to obtain social and economic value of data. This article seeks to establish why adding a data layer to women cooperatives in particular has the power to redistribute the economic benefits of data.

As part of the 17 rooms project by Brookings Institution to develop a roadmap on building gender data cooperatives, we looked at one use case in particular — Megha, which is a women farmer cooperative under the SEWA (Self Employed Women Association) Federation of Gujarat, India. This paper is an analysis of our field visit to Megha cooperative and our early thoughts and hypotheses on how adding a data layer to the cooperative could provide value to its members and what that could look like. The intent of this article is to demonstrate the value of building a gender data cooperative — pooled data can be used to demonstrate creditworthiness to financial service providers and enable increase in access to credit and receive accurate data driven advice and support for women farmers.

The article is structured into two parts. Part I contains a summary of our field visit. Part II has 2 sections — Section 1 seeks to explain why data cooperatives are the most appropriate stewards of data, the impact of cooperatives on empowerment of women and why organizing gender data cooperatives in the context of agriculture is extremely important. Section 2 explains how Megha cooperative currently functions and attempts to explain how data cooperatives can deploy data better to provide value to its members through our hypotheses. Finally, the conclusion provides arguments for why data cooperative is valuable for the women farmer community and lays down the way forward.

Part I : Notes from the Megha cooperative field visit

• Aapti Institute made a visit to Megha Mandli, which is a women farmer cooperative of the SEWA Cooperative Federation located in Gujarat’s Tapi district. The purpose of the visit was to understand how the cooperative works, its structure, governance, data collection and the challenges they face.
• Megha Mandli has a total of 1001 members or shareholders. The cooperative works closely with the SEWA Cooperative Federation, who helped initiate this cooperative. The SEWA Federation provides specialised services such as capacity building, training, access to markets, access to credit, business process streamlining, compliance and HR/admin, research and communications.
• The cooperative has 14 board members who are democratically elected. For every 10 villages, there exists a coordinator between the board and Megha Mandli called Sankalit Sathi. At the community level, the Aagewan (community leader for 1–2 villages) interacts with farmer members at the ground level.
• Certain decisions will be first decided by the board and then they will conduct general meetings with members of the cooperative.
• The general meetings happen once in a year and board meetings happen bi-monthly.
• Responsibilities of board members include discussing the activities of the cooperative and future plans for the cooperative. This decision/discussion is carried forward to the members; the cooperative also supplies inputs such as seeds, fertilizers and small tools/equipment to members to solve issues that arise due to distance and quantity.

The cooperative engages in 2 main types of work -

1. Organising women farmers and their collective engagement for business related planning. SEWA helps them think through their by-laws until they are registered.
2. Planning business models — The cooperative mainly works on the input side. They focus on activities such as supplying paddy seed or poultry feed/chicks to the members. For instance, when members engage in poultry farming, the cooperative buys input such as chicks, feed, eggs and sells it to the farmers. This helps farmers save time and money as the cooperative negotiates for lower bulk prices while buying input. The farmers then sell it in the village (output).

Part II : Analysis of Megha visit

Section 1

Data cooperatives are the most appropriate stewards of data

Data stewardship has emerged as a framework for community-oriented data governance — a structure that can simultaneously unlock data for public good use while not only preserving the rights and interests of the people whose data is in question, but also seeking value for them from the sharing of such data. Cooperatives serve as the ideal form of existing community-oriented institutions to facilitate data stewardship. The principles that ground the cooperative movement (non-monopolistic, self-managed, horizontal, collective benefits) serve as a promising alternative to the extractive logic of the current data ecosystem. This makes data cooperative an ideal steward for data as it is also based on similar principles. Cooperatives also have established governance structures with democratic decision making and are capable of articulating the needs and concerns of their members accurately. Cooperatives can pool data and allow the sharing of data between various stakeholders while being integrated with the data cooperative. There is also potential for cooperatives to exchange data through federated data flows.

Data cooperatives can thus monetise cooperative structures as an optimal form of economic management. The core principles of cooperatives also include supporting the communities they work with, education and training initiatives — which are extremely important for women as it provides access to social protection. This poses many advantages for data cooperatives to be built upon existing cooperatives. Data is collected and organized in structured systems that can be then used for the benefit of the members (including through sharing with third parties). Thus, data acts as a representational tool that brings value by facilitating the objectives of individual members.

Women collectives and the value of gender data cooperatives

Women’s financial journeys are different from men. The Global Financial Inclusion Database has found that women are particularly disadvantaged when it comes to access to financial services. Women's enterprises usually face higher interest rates than their male counterparts and are discriminated against by banks and must change the way in which they seek funds. A pattern of higher credit access by men when compared to women was observed in all surveyed districts across the states of Bihar, West Bengal and Orissa. Women-owned MSMEs in India faced a whopping gap of 70.37 percent in financing. 90% of women entrepreneurs in the country rely on informal sources of financing.

All of this can be attributed to social biases on the part of banks and financial institutions with respect to the credit-worthiness of women-led enterprises. This is despite most research confirming that women are more disciplined than male borrowers and have a better credit profile despite clear social barriers, including a lack of assets (land ownership), lack of collateral, opportunity and technical skills. Moreover, trust is required for women to invest and use new technology — particularly digital financial services, as there are high stakes of personal finances involved. This, in turn, leads to a self-perpetuating cycle of financial exclusion as women are inhibited from applying for loans which leads to limited access to improved inputs and better productivity. When Micro and Small Women Entrepreneurs (MSWEs) are empowered, there is a corresponding larger-scale impact which benefits the development of the society as a whole . This can be attributed to the tendency of female entrepreneurs to reinvest their money in their children’s health, education and so on. Kudumbashree, for instance, is a programme in Kerala which is primarily organised around neighbourhood collectives that comprise women who belong to high risk families and are economically disadvantaged. This is a remarkable example of organising women cooperatives to encourage them to use pooled resources to meet the needs of members. Small regular savings of neighbourhood groups (NHG) are pooled and given out as internal loans to members. These loans serve as buffers that address any immediate financial shocks. Bank linkage programs of Kudumbashree have made access to banks to neighbourhood groups (based on objectively identifiable parameters to rate NHG) possible without any collateral. Cooperative structures can therefore help in increasing the collective bargaining power to gain access to institutional credit and further the goal of financial inclusion

With the prevalence of women engaging in collective enterprises, gender data cooperatives can help in facilitating access to credit and digital financial service providers. The cooperative can help its members understand the advantages of using new products and help ease the process of transitioning to these products and help with the operational requirements associated with using the services. The cooperatives can help in developing and disseminating gender-specific value propositions for Digital financial services (DFS). By adopting a gender lens while providing training and assistance, the cooperatives can also help minimise risks by monitoring unintended adverse effects to women and address concerns of women by providing them recourse. Gender data cooperatives can thus build trust and confidence amongst its members and increase transparency and accountability while empowering women organisations. A cooperative’s membership therefore can lead to an increase in women’s decision making power at household, group and community levels. A remarkable case study is that of P’KWI Farmer cooperative society in North Eastern Uganda which has proven that membership in the cooperative has increased women’s abilities to influence decision making and has significantly increased their economic well being as they gained knowledge and adopted agronomic activities.

Participation of women in agriculture

Over the last 2 decades, we have witnessed a massive jump in the number of female workers employed in agriculture in India. In 2018 to 2019, 71% female workers were employed in agriculture. This increased participation of women agri-labourers and phenomenon of feminisation is only likely to continue with the increased pressure for men to migrate to urban areas.The dairy cooperatives in Gujarat (AMUL brand) and Shri Mahila Griha Udyog Lijjat Papad are examples of highly successful women leader cooperatives. The problems faced by the farmer community are extremely grave and likely to affect women farmers even more disproportionately — in terms of their right to land, input, market and credit among others. While the contribution of women in the farmer economy is very significant and is likely to increase, they lack voice in decision making and lack access to opportunities. To validate their female agency and to empower them, there is a need to provide services such as capacity building, training and access to credit — all of which women data cooperatives can facilitate while providing women with representation and decision making abilities.

Women are generally underrepresented in formal agricultural value chains, which makes it difficult for female farmers to access agriculture credit. Agriculture credit is a prerequisite for farmers to increase their agricultural output and plays a key role determining the agricultural development of a country. Lack of data on smallholder incomes and credit histories limits the ability of FSPs (Financial Service Providers) to accurately predict cash flows. FSPs are increasingly using data and technology to address the data gaps and conduct risk assessments while creating credit products for the farmers, especially women. This would enable them to design financial products tailored to farmer’s crop cycles.

There are innovative agriculture digital financial service business models that can capture the wealth of farm and farmer data for credit risk assessment and help facilitate data sharing partnerships through APIs. Startups offer loans to unbankable customers using big data (mobile money data, call data records and so on) which is processed by algorithms to predict a person’s likelihood to repay a loan. For instance, Apollo Agriculture offers a short-term input credit product bundled with agronomic information and advice and weather index insurance. The company uses alternative data sources to perform credit risk assessments. They combine farm and farmer data collected by their agents with satellite data to predict value chains, crop yields, crop cycles, housing ownership, animal/livestock ownership and road access. Apollo then applies machine learning to assess creditworthiness and design credit products tailored to farmers’ production cycles. The loan is disbursed and repaid using mobile money. These companies use alternative data to provide credit for farmers which can greatly benefit them and data cooperatives can assist in the organizing, collection and sharing of such data while protecting the privacy rights and interests of the farmers.

Why organizing women as data cooperatives in context of agriculture is important

Agriculture, as a sector, has a long and rich history of cooperatives. Among the 300 largest cooperatives in the world, about 30% are found in the agriculture and food industry sector, as per International Cooperative Alliance. Agricultural cooperatives throughout the world have played a significant role in organising the small farmers who are responsible for 80% of the world’s food production. These cooperatives tend to be well organised, and are present at different levels, with some acting as a hub for smaller, more localised cooperatives.

Cooperatives allow those who are prone to being exploited and cannot enter the market system to gain strength to do so collectively. They have been the best suited institutions for economic and rural development, especially for women. Women cooperatives have empowered women by enabling access of credit to women at affordable and low interest rates and the consequent promotion of self employment and small scale industries. A study conducted by ICNW states that there is 1 woman member against 3 male members in a cooperative and participation of women, especially in the decision making of the cooperative has significantly increased over the past couple of years.

Cooperatives provide small farmers access to raw materials, advisories on better farming practices and sources of credit. They are also an important factor in obtaining better prices for farmers for their yield, by allowing for a stronger bargaining position. Large scale cooperatives in the agriculture and animal husbandry sector in India also have a very successful track record — AMUL, IFFCO and KRIBHCO being notable examples. These cooperatives are adept with technology adoption and they have large membership bases with a great deal of trust between the members and the cooperative structure. The presence of large cooperatives in the agriculture sector is also witnessed across numerous jurisdictions, making it an ideal vehicle to pilot a roadmap to build data cooperatives.

We believe that data cooperatives, either built on existing cooperatives or newly created, pose many advantages as they are cost effective and provide easier access to data providers and data users. These data cooperatives can act as data custodians to ensure farmers stay in control of their data while taking into account the regional perspective and the associated needs in an already established level of trust.

Data cooperatives can help in pooling of data from various sources, provide support on how to evaluate benefits of farm data use to derive value such as getting access to capital. This can be done by focussing on the direct concerns to farmers, incentivising farmers to use their data and presenting data driven solutions in an understandable and easily identifiable way. Data cooperatives can pave the way to utilise the potential for farmers to use different data sources to create data products that can increase the value of products (for purpose of sale/marketing). For instance, combining data regarding feed consumption, vaccine used, organic fertiliser used, digital equipment performance and so on can help increase the value of the poultry or harvest that is up for sale.The data cooperative will integrate multiple data points to get real time information and recommendations for farmers. The data cooperative can thus act as an intermediary which shares data with data driven apps, agritech companies and ensures that farmers benefit from the value of data. Therefore the data cooperative can undertake the task of marketing their data on conditions set and agreed upon by farmers.

Why organizing data cooperatives for farmers in the context of climate change is important

Changes in temperature, frequency and severity of droughts, floods and precipitation cycles have made it very difficult for farmers to harvest their traditional crops, combat proliferation of pests and diseases and retain their nutritional value. Populations in developing countries (that are primarily dependent on agriculture) are more likely to be affected by climate change impacts on agriculture and agriculture in turn contributes largely to climate change due to greenhouse emissions. Small and marginal farmers account for 86% of the total farmers in India and they lack the ability to afford advanced machinery or improve modern farming practices and hence are hampered by low productivity. This has left them very vulnerable and food insecure.

Current data collection efforts are top-down, disparate and the datasets are siloed — farmers have very little to benefit from them. Bottom-up data cooperatives can help reverse current extractivist data practices where farmers can exert more control of their data and make decisions to benefit from such use. Further the diverse environmental conditions and weather patterns in countries like India, is another area where improved data use can greatly benefit farmers. Data can also be used for digital agriculture or smart farming and can provide opportunities for farmers to learn about regenerative practices and understand the impact of their practices to the environment. Data cooperatives with appropriate governance mechanisms can help local farmers collect relevant datasets which could be shared with researcher and agritech companies in return for services and hyperlocal farming advice best suited to their circumstances and needs.

The Birchip cropping group, for instance, aims to improve the lives of farmer communities in Australia by providing support and tools for better farm management practices. The UAWC project promotes collective sustainable models of agriculture and enables women rural communities in Palestine to better respond to their community needs while facing climate change and under territorial occupation.The project assists cooperatives in selecting drought resistant crops and adapting to climate change. It was found that as women participation in the economy increased, it improved women’s influence in local policymaking and they found their voice in their own household and in the public sphere.

Section 2

How can Megha deploy their data better through a data cooperative

How does Megha currently function and the issues

Currently the relationship between the cooperative and farmers is such that it is a community-based organisation led by women farmers and provides advisory support to farmers on both agricultural and non-agricultural activities.

1. The cooperative supplies input such as paddy seed, poultry chicks/feeds. They buy from input suppliers at scale and sell it to the farmers according to their demand.
2. The cooperative however does not involve itself in the output activities of the farmers, i.e., the processing, marketing and sale of their produce.
3. In our conversation with Megha we realised that they have the desire to scale at the output side but are hindered by the lack of capital to do so.
4. They have till date not taken any loan including from the SEWA cooperative Bank — this suggests that they are extremely risk averse. This is because the SEWA Cooperative Bank provides loans only to individuals and not cooperatives. The member farmers have taken credit from SEWA during the pandemic for seed purchase.
5. Unlike Megha, most farmer cooperatives have an increase in share capital and maintain savings. For example — Mulkanoor farmers cooperative, which deals with collecting milk from farmers, processing and selling it — has both group savings and higher share. In the case of Megha, the membership fee is INR 10 with 1001 members/shareholders and they don’t maintain any joint savings (this could either be attributed to a possibility that access to credit and microfinance was something that the cooperative did not want to handle) — which means that there is already a paucity of capital.
6. Currently the cooperative provides an advisory role with the help of the SEWA Cooperative Federation. They also organise training programmes for farmer-members and enable access to government and SEWA offered social protection services.
7. The co-operative makes decisions (such as determining the demand for input by farmers) on an estimate and is likely to suffer losses if the estimate is not accurate.

Our hypothesis is that pooling of data through data cooperative structure can help ease the access to capital and hence scale output activity and generate more profits.

How can Megha deploy its data better

Data cooperatives can help deploy data better to provide value to farmers of Megha cooperatives. Our hypotheses are that data cooperatives can manage pooling of data and enable farmers to raise money as a collective and increase access to capital; use the pooled data to provide better data driven insights with the help of AI; help farmers get insights from each other in the community on best practices and negotiate for better services based on farmers needs while protecting their data rights. This section attempts to demonstrate how some of our hypotheses of data cooperatives provide value to farmers.

Hypothesis 1: Data cooperative can manage the pooling of data and help increase access to capital

The data cooperative can collect data from various data sources to build an economic identity for farmers and get aggregate data to demonstrate credit-worthiness of the cooperative to get access to loans. For instance, in the case of small and marginal farmers in Dharwad, it was found that education status, landholdings, irrigation facilities and income levels were the main criteria to assess credit. Therefore many such data points can be used to build an economic identity for each farmer.

Data variables that are loan worthy include — access to reasonable irrigation facilities (irrigated areas have greater demand for institutional credit), health records of farmers and access to hospital facilities, data regarding soil health (Kisan Vigyan Kendra provides for low cost soil tests for micronutrients) and the kinds of crops (whether it is commercial or high value) that ensure consistent return. All these could serve as fundamental building blocks of data for long term loans.

A combination of agricultural (input,produce yield, quality, cash aid for produce) and non-agricultural data (personal data, bills) and gender, family size and occupation will help create an economic identity for farmers and the aggregate data will help get access to credit for cooperatives. Musoni MicroFinance Ltd is an example of a data driven organisation that seeks to serve the unbanked population. They design loan products for farmers with flexible repayment periods. After an analysis of several data points such as KYC, farming data — the product is designed in a way that caters to the farmer’s diverse needs and the crop cycle.

Here are the following ways in which the data layer in Megha can manage pooled data to get access to credit.

a. Input side

Issue

Megha supplies input to the farmers by buying it from input suppliers, however they make the payment upfront while buying such input.

How can data cooperative help

• The proposed data cooperative can collect data on the farmers such as — who receives input, the price they are paid at harvest, whether they repay credit and so on. This collection can happen via community leaders in the village or with the help of digital solutions. For instance, Farmforce — a digital procurement solution allows cooperatives to track interactions with farmers such as input supply. Examples of data points typically captured in the app include farmer ID, crop information, farm geolocation data, farm size, crop cycles and a photo of the farmer.
• After its collection, the data cooperative can show the aggregated data to FSPs to establish the it’s collective ability to repay credit.
• The FSP on profiling the creditworthiness of the cooperative, can pay directly to the input supplier for which the cooperative will repay later.
• The cooperative then recovers the loan amount by deducting it from the farmers output (produce) payments.
• The data collected could also be used by the cooperative to accurately estimate the actual demand of input of the farmers while buying from the suppliers.

b. Output side

Issue

In the case of Megha, the lack of data on farmer activities (output) such as total sale, profits, farmers’ incomes and their creditworthiness limits the ability to predict future cash flows and extend financial products to farmers like loans and insurance.

How can data cooperative help

• On the output side, The various alternative data collected on farming and farm data can be shared to FSPs for estimating credit rates. The aggregated data can be used to establish creditworthiness and get cash advances/credits from FSPs. The Kisan credit card crop loan, for instance, marks short term credit requirements to include expenses incurred from the cultivation of crops and post-harvest activities.
• On receiving credit, the data cooperative can disburse and monitor the use of credit for agriculture production. Financing agriculture production using credit improves quality and predictability of output. Providing cash to farmers as per their needs can be an effective incentive for farmers to sell their share to the cooperative.
• The data cooperative can use the credit to buy the produce of the farmers at the and sell it to a group of agribusiness buyers/retailers or open market

c. Tailoring non-agricultural loans to farmers’ cash flow cycle using data

Issue

Farmers lack access to credit for their non-agricultural needs

How data cooperatives help

• The data cooperative can use the combination of farm data, personal and non-agricultural data collected to build an economic identity for individual farmers.
• This can be accessed by farmers with the help of data cooperatives and used to get individual personal loans. The data cooperative can help digitise and support the entire loan process from application of loan to collection of data, monitoring and repayment.
• The data cooperative can thus leverage this data to negotiate with FSPs for loans that cover a farmer’s non-agricultural needs that are tailored according to the farmer’s crop cycle.
• Alternatively, the cooperative as a whole can take out loans for non-agricultural needs and then lend to farmers behaving as a non-profit micro-credit entity.
• Using the data shared, loans can be designed to fit the crop calendar of farmers. For instance, Advan used a credit scoring model for school loans that combine agriculture and financial data. In this particular case — the loans were disbursed at the start of the school year and during periods of negative cash flow and repay the investments during the harvest months

Hypothesis 2 : Data cooperative can help maintain and digitise savings to increase access to capital

Issue

Megha currently does not maintain savings. Irregular cash inflows and regular cash outflows add to the difficulty in managing finances for farmers. This makes savings extremely crucial to meet agricultural consumption and household needs when cash inflows are limited and to cope with financial shocks. To cover expenses, the farmer also would require flexible access to credit, enabling them to make long term investments that increase their productivity, in addition to their savings for the year. Savings increase deposits and both of these factors help in increasing access to capital. For farmers to contribute to greater share and savings, there is a need to establish an extremely high level of trust.

How can data cooperatives help

• The proposed data cooperative can have members contribute equal amounts to a group fund or a linked savings account. This can help increase access to loans.
• Market-appropriate savings vehicles for all women-owned small and medium enterprises increases productivity and earnings in developing countries. Savings build members’ capital for consumption, emergencies and investment and protect them against borrowing from money lenders — they give access to loans as well as to get insurance from financial service providers. Savings products for women farmers in rural areas become extremely important considering the seasonality of their incomes. Digital savings culture has been proven to enable farmers in both formal and informal value chains to address short term financing needs. Creation of data systems by digitising and linking savings groups for farmers will help capture funds and their transactions with it — thereby building farmers’ financial history.
• This could also reflect how the savings group operates, generating a record of individual and group balances and transaction history and hence unlocking new opportunities, predominantly access to credit . Data cooperatives can help maintain this and provide assurance to members on security of their savings with a transparent system in place. These voluntary savings can provide flexibility of withdrawing cash at times of need. APIs can be used to integrate the digital wallets of individual farmers with savings accounts. Mobile money enables Digital finance services (DFS) in the agriculture sector — it acts as an alternative to cash payments and acts as an entry point to provide savings and credit products for smallholder farmers
• For instance, Vodacom Tanzania launched M-Koba in partnership with TPB Bank in 2019. Here, the M-Koba account is managed by the group leader that adds members to the group using mobile numbers. Members send money to the group savings account from their digi wallets — each member can request a group savings loan automatically.

Hypothesis 3: Data cooperative can help manage the pooling to provide data driven expert advice

Issue

After a preliminary analysis of samples of data collected that were shared by the cooperative and information gleaned from the interview, we realised that the data collected digitally by the cooperative is not extensive and maybe limited to data regarding seed purchase and sale list.

This is a barrier to getting accurate data-driven expert advice on daily activities of farmers. Currently, such decisions are broadly based on an estimate and on occasions, they have suffered monetary losses for the same. This is likely to reduce if decisions are data-driven and will match the demands of the farmers. The data cooperative can use data management and analytics capabilities.

How data cooperatives can help

The data co-operative can also partner with agri-tech companies and ensure responsible data sharing to get information regarding climate, soil quality and geo-location data. The data cooperative can manage such data and analyse it to make informed decisions.

Data sharing is essential for the growth and scale required to close the financing gap for smallholder farmers and to make better informed data-driven decisions with consent of farmers. The data cooperative can enable this by aligning itself at a strategic, operational and commercial level with a common understanding of data protection and processes for its collection

Conclusion

Why deployment of data through a data cooperative is valuable to the women farmer community

Through a data cooperative, pooled data can be used to demonstrate creditworthiness to financial service providers and enable increase in access to credit and receive accurate data driven advice and support. A data cooperative can enable the transition to digitising data collection and aggregating a variety of data points that could enable an accurate estimation of credit risk. Farmers have both agricultural and non-agricultural needs and building a digital economic identity can help them get access to loans for non-agricultural expenses such as education and rent. Financial service providers also benefit as it reduces their credit risk. This economic identity can be created using a multitude of data points (such as household events, assets or transaction history) to present a more dynamic and holistic picture. Women thus have the ability to control and also create data products that can increase the value of their product.

Data cooperatives also have the potential to negotiate with Financial service providers to ensure that members are not adversely affected when using new forms of services. Existing AI systems perpetuate bias and bring women and marginalised communities into a credit system that is detrimental to them. Data cooperatives can ensure this risk is mitigated. Stewards (such as data cooperatives) by virtue of their links with a diverse community of data producers, can curate, standardise and supply representative datasets to train AI models and reduce the scope for bias. They can facilitate control over one’s data and decide the terms for data use in AI development and demand accountability from data users through grievance redressal mechanisms. ‘High-risk’ applications (such as AI integration for credit scoring) that could result in denial of essential services come with significant compliance requirements. There is potential for data cooperatives to conduct assessment audits and be continuously involved in the evaluation of the social consequences produced by deployment of AI systems amongst its members. Community leaders within the cooperative could also perform the role of human oversight into AI systems and mitigate the discriminatory effects of automated decision making systems.

While making the value proposition to women and demonstrating the gains of data cooperatives and adoption of new technology, it is important to gain their trust. Women need to be sure that technology will work in their best interests and perform its purpose reliably. Especially with digital financial services, women are inclined to take help from agents more often before they feel comfortable using these services. Data cooperatives will help enhance this digital trust by playing the role of intermediary that uses data to achieve the needs of its members and facilitating credit and business management training as required. Data cooperatives can also enhance trust in the DFS provider that the cooperative partners with and enhance trust in the product itself — that the product will work as designed and for the benefit of members. To further this trust, data cooperatives can facilitate training and awareness on how to use these products. The data cooperatives also ensure that the user’s experience with the technology must result in positive expectations with respect to its functionality, usefulness, and reliability. The data cooperative can also ensure that AI systems used are not exploitative and establish grievance redressal processes to resolve service issues in a timely manner.

With the prevalence of women engaged in agriculture in developing countries, data cooperatives can share data and partner with external parties that run agricultural extension programs supporting specific needs of women and are a trusted source of information on a range of topics. As previously noted, trust again plays a critical role in understanding the true advantages and challenges with using the new product or service and data cooperatives can play a major role in enhancing that digital trust.

Way Forward

We believe that data cooperatives can use pooled data in an easily accessible or interoperable format to access credit and invest on behalf of farmers in a more cost efficient manner. The cooperative can be used to test data solutions and gain insights from farmers on various decisions made around agricultural activities. A phased approach to credit, starting with small, relatively simple and short-term products, can introduce farmers to a credit system and help them build credit history that unlocks larger, longer-term credit. The data cooperative model can be used to generate societal value and help groups evolve mechanisms of participative governance and negotiate better.

With Megha — we seek to test our hypothesis and to understand whether adopting the data cooperative model will be helpful to the farmers. We seek to establish a finalised data cooperative design for Megha women farmers cooperative created through a community co-design process and to address essential architecture and governance issues. To achieve this — we will make field visits to understand the willingness of farmers to share their data and most importantly, involve them in the co-designing process of building this data cooperative layer.

If you are interested in learning more about this work or collaborating with us on any of these questions, please get in touch with us at contact@aapti.in.

By Suha Mohamed & Vinay Narayan

Despite new modes of monitoring and capturing data, the ocean and its complex interdependent aquatic ecosystems are still relatively unmapped. Roughly 85% of it remains ‘unexplored’ — due to a vast set of operational and cross jurisdictional challenges. Yet this situation is quickly evolving — we are now in a period defined by the UN as the ‘Decade of Ocean Science for Sustainable Development’, accompanied by global calls and efforts by academic and scientific communities to strengthen our understanding of marine habitats and functioning, while nuancing our awareness of how human interventions (pollution, fishing activities, development, etc.) affect these conditions.

In July, several countries also adopted the new Lisbon declaration at the United Nations Oceans Conference. Actions that stem from this declaration include: “strengthening data collection, recognising the role of indigenous people in sharing innovation and practices to reducing greenhouse gas emissions from international maritime transportation, especially shipping”.

However, in a world where data is also intricately intertwined with driving profit for private sector entities, there is growing commercial interest in collating data on oceanic ecosystems for activities like deep-sea mining for minerals — projected to be highly damaging for marine ecosystems. This calls for greater exploration into how we can incentivize a move away from extractive physical and data-related practices towards a paradigm of interventions which would meaningfully orient actors around common goals of oceanic conservation and data stewardship.

Additionally, while there is momentum building around sustainable development-oriented data efforts, aside from progressive rhetoric, many of these initiatives are still largely exclusive processes. Consequently, communities most vulnerable to effects of climate change are left out of conversations around sustainability. The United Nations Oceans Conference for instance was criticised by small-scale fishers and indigenous people who argued these sessions were inaccessible and excluded them during important decision-making processes. This lack of representation and ability to participate in shaping systems of conservation and resilience ignores the long-standing role of indigenous communities in ‘stewarding’ these resources. These systemic forms of exclusion also trickle down to the design and deployment of data-collection and governance efforts.

Therefore, while there are a growing set of efforts to monitor, reconcile and coalesce data that pertain to marine conservation and fishing — these must be designed to also be inclusive of indigenous modes of data collection and governance. Further, there is also a need to find new approaches to strengthen participation and agency within existing systems.

This case study review focuses primarily on the evolving landscape of fisheries or oceanic data management to understand how it is currently being captured, managed, preserved and shared. This area of inquiry was selected as our research highlights this domain has a growing set of existing stewards and organisations that are spearheading efforts which are trust-preserving and fisher-centric. Deriving lessons from these organisations can be crucial in understanding how they can be scaled across regions and better inform global conservation efforts.

Understanding the fisheries management ecosystem and role data plays within it

Fishing represents one of the largest maritime industries in the world, both in terms of the number of people involved as well as the economic value associated with it. Fishing from capture fisheries, i.e., fishing from natural sources such as seas, rivers and lakes for commercial purposes, constitutes a predominant segment of the larger fishing industry. As of 2017, around 120 million people were involved in capture fisheries, with more than 90% of these people involved in small-scale fisheries. Aside from being a significant source of employment, capture fisheries are also an essential source of nutrition for millions around the world — making capture fisheries critical for food security and nutrition. However, as of 2018, over 35% of fisheries were overfished with this number only rising in the subsequent period.

Fisheries management has been a critical part of the functioning of international organisations for decades, but has taken on renewed importance in light of increased demand for marine life for commercial purposes leading to an uptick in illegal, unregulated or unreported fishing. The goal of fisheries management is to promote and facilitate sustainable fisheries which ensure that marine life is harvested at a sustainable rate so as to prevent the decline of aquatic life population over time. The Marine Stewardship Council, an international non-profit, foregrounds three main principles in their assessment of sustainability of fisheries: (1) Sustainable Fish Stock; (2) Minimising Environmental Impact; and (3) Effective fisheries management. Given the socio-economic impact that fisheries have, ensuring the growth of sustainable fisheries is paramount. This is evident in the call by Members of the European Parliament to increase funding and investment opportunities for sustainable fisheries.

Existing challenges around data and their impact in the fisheries data ecosystem

A critical factor in ensuring the continued sustainability of fisheries is the availability of “fisheries data”, data that usually includes biological information about the exploited fish and associated species, economic information about the fishermen and the markets for the catch, and information about the environmental conditions that affect the productivity of the species. The FAO has been developing an international framework to address fisheries management with a number of instruments governing various aspects of the same. In accordance with this, as well as other international obligations (such as the sustainable development goals), there have been a number of regional and national frameworks around collection of fisheries data for better fisheries management. However, despite this, there remains a paucity of adequate and reliable fisheries data for many countries.

This issue is particularly severe in developing and low-income regions containing numerous small-scale fisheries. This paucity of data has consequent effects on the ability of governments to implement policy to adequately tackle illegal, unregulated or unreported fishing, thereby allowing for unmonitored fishing that has severe socio-economic consequences, most importantly for members of the local communities in such areas. An additional issue with the collection and analysis of fisheries data is that fishers and local community members have typically been treated only as a source of data, as opposed to being looked at as contributing scientists. In addition to having deleterious consequences on local fishing communities, such practices also lead to the ignorance or loss of indigenous knowledge that can vastly impact sustainability in fishing. For example, a recent study of indigenous practices around oyster fisheries in North America and Australia revealed key takeaways that can be used to inform sustainable fisheries management globally. Similarly, another study on the use of traditional practices in the harvesting of Kelp along the northwest coast of North America concluded that customary management practices and social norms rooted in indigenous knowledge promote system-wide social–ecological resilience, and importantly, can empower local decision makers and democratize the science and practice of natural resource management.

Similarly, citizen science fisheries initiatives that place an emphasis on building capacity of fishers and local community members to collect accurate and high-quality fisheries data have also shown great potential in addressing the paucity of fisheries data faced by scientists and policy makers. Citizen science projects developed in collaboration with small-scale fishing communities in Mexico’s Pacific Ocean, Gulf of California and Caribbean Sea resulted with the projects providing a wealth of data and opportunities, for fishers, researchers and managers — with the data generated being used to evaluate stocks and habitat, detect the impacts of climate change and increase knowledge on fisheries. Furthermore, it has also empowered a network of local leaders, who are now transmitting a message of environmental stewardship in their communities.

Promising efforts to responsibly collect and steward data related to fisheries management

In this context, through our research at the Data Economy Lab, we have seen a range of new initiatives that seem to intertwine high-levels of community involvement, participation and stewardship of data for fisheries management. These illustrate the possibilities that community-oriented data-driven fisheries management may hold and must be further unpacked if we are to understand and forge its pathways for replication, adaptation and scale at a global level.

This blog focuses specifically on four cases, where the first, PescaData has been featured in an episode on our video series, Tracking Stewardship. In addition, it also showcases: Abalobi, DTU Fangstjournalen, California Groundfish Collective and Dakshin Foundation’s work.

PescaData

Tracking Stewardship: Empowering fishing communities through data-driven insights & innovation

The video below features a conversation with Stuart Fulton from Comunidad y Biodiversidad — COBI, a civil society organisation that focuses on marine conservation and sustainable development. Their product PescaData is a software as a service that enables the digitalisation of catch records, supports individual fishers in sharing information with their fishing organizations/collectives and enables access to insights on self-generated data. Our discussion highlights key challenges PescaData aims to solve for and outlines their approach to delivering greater value and data justice to fishing communities and collectives in Mexico.

Learnings and areas for further development and application

• COBI may be best defined as an ecosystem enabler organisation that is building the supportive technical infrastructure, tools and capabilities required for communities to practice self-governance over their data. In their experience, this means supporting new end users (collectives) with the onboarding process. Going forward, their aim is for their end-user organisations to be able to navigate their platform — without requiring extensive engagement.
• COBI is invested not only in delivering what it believes may be valuable insights back to communities but is keen to understand how this information is used by users for future decision-making and adaptive management. This evaluation and feedback loop can further product development and ensure more responsive design.
• Acknowledging and integrating pre-existing, traditional knowledge forms and practices from communities is key to PescaData’s vision and design. Their knowledge sharing mechanism enables a peer-to-peer exchange of information — allowing for digitisation of these practices and collaborative enrichment of knowledge sets.

Abalobi

Abalobi is a non-profit social enterprise based in South Africa that seeks to “contribute towards thriving, equitable, climate change resilient and sustainable small-scale fishing communities globally, through the joint development of Technology For Good”. They facilitate this through the use of mobile apps that relate to seafood traceability, fully documented fisheries, fair and transparent supply chains. The provision of mobile apps is supplemented with initiatives designed to deepen community cohesion and entrepreneurship. The ABALOBI FISHER app provides fishers with an e-logbook along with services of data dashboards, basic bookkeeping, permits safekeeping and weather forecasting. The Abalobi MARKETPLACE app provides the local fishers with a marketplace where registered chefs or restaurants can purchase fresh, traceable seafood and post-harvest products offered by approved fishers. This digital marketplace is made available with an integrated area-specific approach to maximise logistical efficiency. The ABALOBI MONITOR app is a digitised community catch monitoring platform for use at the landing site and along the shoreline. Using tablet-based technology, the platform is designed to improve community catch monitoring by providing monitors with dashboards of data collected per fisher and landing site. It also enables communication with fishers, other monitors and fisheries managers via an integrated chat platform.

What does it aim to solve for in the data economy?

• Providing control over data to the data producers — Being acutely aware of the asymmetries in the data economy, Abalobi is very insistent on the fact that the data collected using their apps belongs first and foremost to the fishers. Abalobi assists the fishers with visualisation of the data they collect using the app, but any external entities that want access to the data are directed to communicate directly with the fishers, with Abalobi acting mainly as an intermediary that holds the data for the fishers.
• Empowerment through collective action and market information — The balance of power in the traditional fishing system meant that fishers were at the short end of the stick with both the government — in terms of seeking permits or petitioning for other policy action — as well as the marketplace — with fishers being price takers who were invisible in the market. Abalobi assists fishers by allowing them to use the app as a tool to showcase their activities and their Bona Fida status while applying for permits. The MARKETPLACE app places the fishers front and centre to potential buyers who are interested in connecting with local fishers, with this also addressing issues of overfishing wastage, and loss of value of the species in the broader community.
• Preserving and sharing local knowledge — Fishers have a wealth of traditional knowledge that can inform sustainable fishing practices. However, in most instances across the globe where traditional knowledge is concerned, methods of accessing and sharing this knowledge are typically extractive. Abalobi provides a creative and respectful way for fishers to record and share their traditional knowledge within the community.

Learnings and areas for further development and application

• Abalobi is a steward-like initiative that provides fishers with the tools to capture, visualise and interpret their data. They also provide a marketplace that connects fishers with potential buyers, but aside from this, governance structures ensure that control over the data rests with the fishers who have final say over who their data is shared with and how it is used.
• While the Abalobi mobile apps require the use of the internet, they have been designed through iterative loops that incorporate the concerns of the fishers based on their lived experiences. The apps are designed to be very minimal in their use of data — thus accounting for regular concerns faced by fishers.
• Existing market dynamics reinforce an individualistic culture among fisheries. To address this, Abalobi works with local champions, who are typically early adopters interested in experimenting with digital technologies. By providing organisational and business development support, they have a network of fishers who are integral to the onboarding process and who in time can create formalised collective structures.
• In addition to the provision of the mobile apps and data visualisations, a big part of Abalobi’s work includes community engagement and empowerment. Through their work, Abalobi realised that women played a critical role in the entire fishing process. Women in the fishing communities were logging the catch for their husbands or sons, and they were primarily driving the programme — but conventional gender norms meant their contributions were typically invisibilised. Abalobi ran a woman empowerment programme, where they worked with small groups of women and provided them a safe and healthy working environment. This helped them unlock their own intrapreneurship setup and to coordinate it. Going beyond that, Abalobi has worked with these women as core partners in this process, and over time have seen this lead to the development of a strong collective structure.
• A crucial element in the growth of this model is to build trust within the community and make sure that there is real engagement on the ground. This is facilitated through providing support for anybody who is interested in using the technology. As part of the onboarding package, Abalobi provides a lot of training and capacity building around some of the essential concepts. This includes providing basics of financial literacy, as they realised that many fishers tend to use the app as an accounting tool. Lastly, they explore creative and innovative ways to convey and explain concepts — such as contract clauses that are graphically displayed like in a comic book. All of this helps build greater trust within the community, which in turn helps with the increased adoption of the application.

DTU Fangstjournalen

Supported by Technical University of Denmark (DTU), Fangstjournalen is a citizen science project that aims to sustain specific fish populations in Denmark. Fishers or anglers generate this data through an app by logging a range of parameters that include location of catch, duration of trip, species caught and details on its size. Through the aggregated data, DTU can identify the health of certain fish populations. For anglers, the value of this data and benefits related to the app include — greater knowledge around fish populations and their biological characteristics, regular updates on regulations and zoning rules. If anglers are interested, the app also allows them to share their catches with their peers or keep this information private — sharing this only with DTU. The app also enables a degree of social networking for anglers — that can possibly generate the sharing of practices or function as a feedback loop for DTU to better understand what value can be better generated for citizens as a result of data collection. Some interesting use-cases that have been identified through this citizen science project include the identification of parasites/zoonotic diseases, or threats to certain populations like the red skin disease in salmon/trout species through the collection of photos. These forms of knowledge can guide early public health interventions where necessary and be used to generate guidance on handling and consuming fish.

In addition to this non-personal data, Fangstjournalen is also keen to chronicle the wellbeing and interests of anglers — these forms of data also feed into possible collaboration avenues — where anglers can work closely with academics on specific research questions. The methodology of data collection and governance is structured around the FAIR principles — where data must be Findable, Accessible, Integrated/Interoperable and Reused. To start with this involves storing the data in a secure repository (hosted by the university) and the creation of a data management plan. Certain codes, like the DOI, are ascribed to datasets in order to make these ‘Findable’ and ‘Accessible’ to DTU researchers. Data is structured considering open data standards for better ‘Interoperability’ which can facilitate collaboration with international researchers or other citizen science projects — for instance one that seeks to monitor the population of tortoises in a specific region of Denmark. Other data can be ‘Reused’ to inform and surface the health and functioning of certain marine and freshwater ecosystems. Reuse is predicated on standards that make it reliable and high quality for further research.

The quality of the data collected is maintained by referencing it against more traditional methods of data collection and a critical analysis by the team to scan for any emerging anomalies — an absurdly high catch rate or incorrect registration, etc.

Fangstjournalen also implements a number of safeguards to preserve the privacy of data that may possess personal attributes. Sharing is made conditional upon certain license-oriented agreements that it helps facilitate between researchers and students. This prescribes possibilities and conditions around its usage. In order to foster greater scientific awareness and knowledge, Fangstjournalen makes some of this aggregated, de-identified data publicly available through the platform and their Facebook page.

What does it aim to solve for in the data economy?

• Efficient and cost-effective processes — The volumes, spatial and temporal diversity of data required to understand fishing practices and conditions (oceanic and species) can often be a cost-prohibitive and time-consuming process.
• Tapping into the potential of citizens — Traditional means of capturing data through surveys and questionnaires are often complicated to establish and can instead be supplemented with citizen generated, hyperlocal data that can be scaled throughout different parts of the marine ecosystem — rivers, lakes, streams and sea.
• Bridging information asymmetry — There is often an asymmetry of information and limited opportunities available for citizens (anglers) to participate actively in research or knowledge-gathering/development efforts that could facilitate their own livelihoods and operations.

Learnings and areas for further development and application

• Fangstjournalen operates as a steward-like initiative that takes on the role of securing data, providing analytics to fishers/anglers and is committed to identifying new streams of value/benefit that can be redelivered to citizens. With the support of DTU, an academic institute, Fangstjournalen is able to draw domain expertise and facilitate real-time collaboration for the re-use of data for furthering knowledge and innovation. Academic Institutions like DTU may also be well-positioned to offer a great deal of technical support — hosting data, implementing quality metrics/standards and carrying out necessary maintenance.
• Data collected by anglers can highlight important aspects of the human experience, dynamics and welfare related to fisheries management that can inform more responsive policies and practices.
• Based on reflections from the DTU team, including the lead researcher and founder of the platform, they shared that while an electronic platform can facilitate ease in the collection of data, the scalability and relevance of findings is predicated on having a critical mass of users (anglers) actively engaged in data collection. Understanding how to continue to onboard anglers will require a thorough dissemination strategy and engagement with the angler communities — both those in informal loose collectives and those with more organised presences.
• Additionally, retention of users and generating a more active usage of the platform an area presents room for further development — a possible solution they have identified is to expand the pool of anglers to international tourist anglers. This may solve for additional concerns around the lack of diversity or representation of the angler community within their existing datasets.
• Further analysis and thought may be lent to exploring how anglers could be a part of the data governance structure — more specifically where the streamlining usage and sharing agreements could present use-cases for collective benefit. For instance, the sharing of data could present benefits around reforming fisher registration processes, support greater access to markets or be used in advocacy for financial products or in shaping better zoning areas for sustainable fishing practices.

California Groundfish Collective (CGC)

In collaboration with the Nature Conservancy (TNC), the California Groundfish Project brings together fishers from collectives across the West Coast of the state and community leaders with the intent to advance fisheries management and conservation practices. The Nature Conservancy supports the collective in leveraging innovative technologies and facilitates the sharing of data through an app called eCatch.

Their overarching vision is to revive groundfish fisheries (that primarily focus on catching species like the petrale sole, sablefish/black cod etc.). Therefore, much of the data collected can be used to monitor and create necessary guardrails to: prevent the overfishing of species and bycatch and encourage adaptive management — where fishing of these more abundant, healthy species can be made possible.

Through this program, fishers are also encouraged to practice alternative methods to trawling.

“With the goal of enhancing the productivity of the fishery while ensuring healthy ocean ecosystems, TNC worked with fishermen to demonstrate the value of voluntary cooperative management in solving these economic and resource challenges under the IFQ system”

What does it aim to solve for in the data economy?

• Establishing trust and deepening collaboration between stakeholders in the fishing ecosystem through data-driven sustainability and conservation-oriented practices — As a result of unsustainable fishing practices, Central California’s Morro Bay fishery was declared a national disaster in 2000, due to its decline in species and subsequent ban on fishing of certain populations. This led to economic decline of the industry and loss of jobs. While ‘conservationists’ and ‘fishing partners’ were previously been defined as at odds, through consultation and collaboration with both sets of stakeholders they were able to identify common areas for societal value to be generated — and this was identified through data collection. In this instance, fishers can be provided value to maximise their catch, without harming endangered or overfished species. As a result of this collaboration and related data captured, this fishery ecosystem has been revived and is now home to a thriving, sustainable fishery management system.
• Enriching and filling in data gaps to build resilient and responsive climate management around the fishing industry through multi-stakeholder and other non-traditional data sources — The paucity of high-quality data on fishing management and marine ecosystems is well-established — CGC aims to address this by pooling data from a range of fisher associations, academia and other partners. Sources of data also include video and sensor data from fishing vessels — found to be a more cost-effective method of collection.

Learnings and areas for further development and application

• California Groundfish Collective is structured in some respect as a bottom-up steward in that it provides for a governance mechanism that brings together disparate stakeholders and enables existing collectives to tangible benefit from the collective value of data.
• Facilitating the creation of fishing collectives of previously independent, unaffiliated fishers to enable collaborative risk management and compliance with fishing regulations. The existing federal regulations in the U.S impose an ‘Individual Fishing Quota’ to manage groundfish trawl fisheries — to limit fishing of certain species. However, this presents certain economic constraints for fishers as it hinders the ability for them to harvest the necessary volume or meet their own catch requirements. Additionally, if individual fishers fail to comply, they risk being shut down by regulators. To address this, CGC instituted ‘a voluntary mutual insurance pool of quota — known as a risk pool.. If a fisherman incidentally catches overfished species, the collective quota holdings are used to cover their quota requirements as long as they operated in accordance with the agreed upon terms of the risk pool.’. This highlights the relational value around how data may be pooled to attain collective benefits, and overcome individual level challenges — therefore forming what may be defined as data solidarities.

Dakshin Foundation | Fisheries co-management in the Lakshadweep Islands

Closer to home, similar initiatives are also being piloted by organisations like Dashkin Foundation, a non-profit that focuses on conservation and natural resource management.

In their ongoing program ‘Fish for the future’, Dakshin works with the local fishing community in the Lakshadweep Islands and have launched a community-based fisheries monitoring programme. At this point, they support fishers and boat owners in capturing day-to-day data on their practices in logbooks. In the future, this work will be supplemented by a mobile application and a platform that seeks to bring together in addition to fishers, government agencies and civil society with the goal of leveraging this data for sustainable resource management.

What does it aim to solve for in the data economy?

• Institute systems for monitoring of fishery resources that are long-term, participatory and cost-effective. Involving local fishing communities in monitoring provides a platform for cost-effective data generation on a much larger spatial and temporal scale as compared to researcher-led monitoring.
• Create a platform for co-management of fisheries that centres fishers. Regular monitoring enables fishers, who are the primary stakeholders in the fishery, to see patterns in resource availability and can also create a sense of stewardship towards the environment in their minds. This is all the more relevant for data-scarce systems such as the Lakshadweep Islands.

Learnings and areas for further development and application

• In the stewardship ecosystem, Dakshin may be currently best categorised as an ecosystem enabler in that its activities at a grassroots level create the relevant capacities, skillsets while in parallel, also surfacing existing indigenous practices around conservation and environmental decision-making.
• By involving communities and placing them at the core of the knowledge creation process, the project has also enabled the democratization of knowledge.
• While the program is designed on the basis of extensive stakeholder mapping, the continued engagement with the community over the course of the program has led to further learnings that affect system design. It is therefore key to have a system of continuous engagement with the community that will help iteratively build out a co-management system that places them at the centre.

Learnings from community-oriented fisheries management initiatives

In addition to the case study analysis of five organisations, we also carried out a review of emerging literature on ocean conservation and sustainable fisheries management. Below, we have highlighted a few additional findings that can aid global conservation and related stewardship efforts.

• Entrust existing custodians of ecosystems with the agency to guide conservation efforts and interventions instead of penalisation

One approach to biodiversity conservation has been to demarcate fixed zones as Marine Protected Areas(MPAs). Within these zones data can be collected to monitor the population of specific species to better understand their wellbeing and ensure their productivity. To enforce the protection of these spaces, several economic/legal provisions and penalties are often set in place — to prevent overfishing, exploitation or destruction of habitats.

A study in Indonesia found that enabling indigenous peoples to have agency over protecting and sustaining these MPAs is a far more effective approach than enforcing penalties on these communities. Similar recommendations have been made with respect to participative coastal management and protection efforts in other parts of the world. This reinforces the need to ensure that indigenous peoples are at the helm of protection efforts and that subsequent data collection can take place under their leadership and guidance.

• Encourage the development of knowledge systems that combine traditional and indigenous knowledge with innovative data practices and diverse sources (iOT, geospatial, citizen-generated)

While mainstream data collection efforts are often seen as incompatible with indigenous knowledge and practices, few efforts are under way by the public sector to bridge this divide like the Queensland Government’s Indigenous Land and Sea Ranger Program and the Australian Government’s Reef Trust these both involves deepening partnerships with ‘traditional custodian groups’ to better understand and combine ancestral knowledge with ‘western biological knowledge’, explore how its re-use with their participation and consent and identify its possible future value for greater conservation efforts.

• Empower women as key governing actors in the marine and fisheries management ecosystems

Although their labour is largely invisibilized, women play a fundamental role in fisheries management and aquaculture systems across the globe — and comprise nearly 70% of its workforce. Adopting a gender-based lens should be enmeshed into the design of sustainable fisheries-based interventions and programs. From the perspective of data stewardship, requires a few considerations: first, responsibly collecting gender-disaggregated data and carrying out an analysis of the role women play in these respective fisheries/aquaculture ecosystems. Second, positioning women in key leadership positions when it comes to decision-making around management and data-related efforts. Lastly, relevant technical and infrastructural investments must be put in place to ensure that women are fairly compensated for their role and labour in the sector. Organisations like Abalobi keenly recognise the inequity and barriers women face and have responded by equipping fisherwomen with the capacity, tools and skills to be data-owners and leaders in onboarding community members to the foundation.

Fishing for Sustainability — The promise of equitable data stewardship was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

In February, the Data Economy Lab set out to understand the vast landscape of data gathering and governing efforts in the environmental conservation and sustainable development space. Our aim has been to identify the key stakeholders (policymakers, stewards/steward-like initiatives, funders and research institutes) in the space, diagnose their key challenges and explore their needs.

The following blog highlights the role and possible responsibilities we see civil society organisations playing in the broader stewardship ecosystem. It does so by exploring a set of assumptions that position civil society organisations as ‘viable’ stewards, particularly in the domain of environmental conservation and sustainable development. It goes on to share our emerging findings of how these entities roughly fit into set of categories — where each raise a different set of questions and have varying demands. These findings urge a more nuanced approach for those in the ecosystem looking to accelerate the uptake of stewardship.

A recap of our early mapping

We found a range of efforts to collate environmental/ecological data archives or repositories (e.g Environmental Data Initiative’s Data Portal), participatory or citizen-science oriented data collection initiatives (e.g WaterCAN) emerging networks (e.g Environmental Data & Governance Initiative and echo Network) that aim to harmonise and make environmental data openly accessible. Efforts to render traditional and indigenous knowledge, practices and interwoven histories (e.g Kalahari San Master Tracker Programme and Keeping Culture) digital platforms were also found to be spearheaded by regional coalitions, non-profits and international institutes.

Through our research and conversations, we recognized that many of these initiatives are also led or supported by Civil Society Organisations (CSOs). CSOs engaged in environmental justice efforts are now keenly recognizing the adjacencies to data justice — an evolving field of discourse that makes explicit the systematic injustices that data-led and digital technologies can maintain or exacerbate. Similarly, those invested in building locally led adaptation-based solutions are beginning to explore how principles of offline participation, inclusion and accountability may translate in digital contexts.

This evolving trajectory finds resonance in the field of data stewardship. Data Stewardship is a paradigm which explores how the societal value of data can be unlocked while considering what it takes to empower individuals/communities to better negotiate on their data rights.

Are CSOs best placed to steward environmental and ecological data?

Employing the lens of stewardship in the environmental and ecological data ecosystem, resulted in more closely analysing the role that civil society and community-based organisations can play in stewarding or supporting steward-like initiatives in thriving. In a previous piece, we explored how CSOs are well-suited to act as bulwarks against data extractivism and may play a central role in building a more equitable data landscape.

This thinking was guided by several core assumptions; first, an active and diverse civil society is considered a key pillar of democracy. CSOs are not only capable of ‘calling out’ or holding other stakeholders accountable but are increasingly playing a fundamental role in bridging relations between community, public and private sectors. Most recently, CDP, a global non-profit that describes itself as the “gold standard of environmental reporting with the richest and most comprehensive dataset on corporate and city action”, pioneered a ‘Non Disclosure Campaign’ to engage and incentivize the planets highest impact companies (read: polluters) to disclose environmental data that relates to climate change, water and forest security. While their campaign and engagement resulted in 60% greater participation and transparency from some companies, it also shone a spotlight on those that have yet to comply — a list that expectedly names Exxon Mobil and Saudi Aramco, but more surprisingly also includes Tesla. These inter-stakeholder efforts are necessary as siloed environmental data impedes our visibility around the impact of these industries on the climate and hinders the ability to mitigate these risks through interventions and regulation.

Second, existing community-based infrastructures (for e.g self-help groups, unions, cooperatives) can play a crucial role in mobilising communities around data and identifying ‘data priorities’ — existing relationships of trust may position these entities as viable ‘stewards’. Further, grassroots and community organisations may already possess structures for individuals/communities to be meaningfully represented or alternatively, actively participate and make decisions. These mechanisms are foundational to stewardship and if harnessed, will not require creating these processes from scratch.

Digital Democracy, a civil society organisation that co-built a free digital toolset for earth defenders called Mapeo for documenting and monitoring, operates along these principles. Their technology is employed directly by communities as well as by partner organisations and enables them to document, own and share data on human rights conditions and land practices in ways that reflect their interests and outcomes.

Third, CSOs may possess an existing duty of care, loyalty and imperative to minimize harm toward their beneficiaries — an extension of these responsibilities toward safeguarding rights in the digital realm is possible. Most recently an updated playbook was released in partnership with the International Federation of Red Cross & Red Crescent Network to support organisations to embed data protection and responsible data use across the data journey and value chain. Similarly, Data.org’s ‘data maturity assessment tool’ provides guidance to social impact organisations on how to best leverage the value of data for impact. These efforts are foundational to fortifying civil society’s capacities to better protect and uphold digital rights.

Lastly, many CSOs already engage in capacity building and skilling around technical and financial literacy — it is possible that these frameworks can similarly be employed to generate awareness on data rights and literacy. Organisations like Body & Data operate in this nexus and carry out critical work in advocating for feminist oriented digital rights and awareness.

If these assertions are valid, it stands to reason that civil society organisations will continue to require support from the ecosystem in order to effectively play the role of a steward.

Unravelling the stewardship spectrum — categorising CSOs

To get a pulse for how stewardship is perceived and what organisations’ core needs are, we classified our conversations and the key questions that CSOs raised into the following categories:

1. Early stage stewards/steward-like initiatives
2. Mid-stage stewards/steward-like initiatives
3. Supportive/Ecosystem Enabler organisations.

The first two categories can be best described by referring to the data stewardship ladder illustrated in the figure below:

Figure 1: The Data Stewardship Ladder

Type 1: Early stage stewards are characterised as being relatively nascent in their digital journeys and depending on the regulatory frameworks (or lack thereof) that pertain to their jurisdiction, may be newly contending with questions of data governance (consent, privacy, safeguards, risks). This positions most of these organisations in Step 01 or in between Step 01 and Step 02. These early-stage stewards may be at a risk of mirroring the same patterns of data extractivism and irresponsible usage, even though their intent may to be to protect their communities and beneficiaries against digital harm.

Type 2: Mid-stage stewards/steward-like entities appear to be better equipped with the resources, vision and ability to prioritise matters of privacy and data governance and as a result are more likely familiar with consent (and relevant breakdowns), harms and the role of data subjects in these processes. Many of these mid-stage stewards are located at Step 02 of the data stewardship ladder (which may in part be owed to compliance with data protection regulations in their jurisdictions) and are looking to transcend to Step 03. Areas of focus for mid-stage stewards include: prioritising scalability and financial viability, fostering multi-stakeholder partnerships and building out robust technology systems.

Type 3: Supportive/ Supportive/Ecosystem Enabler organisations often aren’t regarded as stewards themselves as they don’t possess data assets. Many of these organisations are either technology providers/aggregators (creating the tools or infrastructure for bottom-up stewards to better govern data), academic institutions/research organisations/non-profits (building the discourse and resources around data rights, governance and justice) or funders (directing resources to non-profits or community-based organisations to implement technology or data-based solutions or innovations). As a result of these interactions, these players are keenly involved with and play a vital role in both supporting and building the relevant capacities of stewards to better function in the ecosystem.

Key Questions from CSOs

The questions outlined below have been drawn from interactions and conversations with organisations that include Dasra (through the Data Empowerment Challenge), Subak (through their cohort of climate non-profits), echo network (a consortia of academic, research, civil society and public agencies that mobilising around climate change, sustainable development and environmental conservation), International Civil Society Centre Berlin (Digital Debates) and a range of other community-based organisations that are actively collecting and stewarding data. They are not exhaustive and have been shared to illustrate the range of demands and requirements organisations may possess across the spectrum.

Type 1 — Early Stage Stewards/Steward-like initiatives

Questions

• What are the regulations on data and its governance? What do we need to consider for us to remain compliant? How does this differ between domestic and international partners/clients
• What are the best ways to collect consent considering our beneficiaries are not necessarily engaged in questions of data?
• Does consent need to be collected retroactively?
• What are ways in which communities we work with and our own intermediary network (who are often facilitators in the data gathering process) can be introduced to the nature, value, and harms of data?
• Who are the personnel that need to be hired to reconcile activities on data and its governance?
• How do we prioritise building or embedding privacy principles into our organisation from its inception?

Type 2 — Mid stage stewards/steward-like initiatives

Questions

• Many partner organisations (grassroots/community groups) don’t have assessments in place to evaluate how data is collected, whether consent is collected and what harms may exist — how does this affect us as potential stewards of this data going forward?
• How do we ensure the accuracy and or validity of information uploaded onto the platform or unlocked from other stakeholders?
• How do we go about identifying funders with similar values or align with our broader vision? Who are these stakeholders in the Global South?
• What are the pathways to convert pilots/paid POCs with public sector agencies into more robust business models?
• What are the trade-offs when considering building customised technologies/platforms vs PaaS services? In either case, what are the ways in which feedback loops in usage and needs can be built and maintained?
• What is the process of harmonising data from different sources or pre-existing stewards of data (open-source portals/dashboards or databases)?
• What is the process by which the curation of data must be decided? Is this a process that is carried out once?
• What needs to happen to ensure this is responsive to the community of users?
• How can accessibility of this data be factored in?
• We only use existing datasets that are already in the public domain — what does this mean for us as stewards?
• Do we still have a responsibility to consider safeguards?

Type 3 — Supportive/Ecosystem Enabler Organisation

Questions

• How do we define data ethics and privacy?
• How can we think more critically about the harms that emerge both upstream and downstream in terms of our data collection efforts?
• Is de-identification of data sufficient as a mechanism to protect data?
• What specific harms can emerge from the collection of non-personal data? How do we identify both the communities and other areas that this might produce intentional or unintentional consequences?
• What are the processes by which indigenous knowledge and resources could be effectively shared ? What are aspects of quality, privacy and transparency that need to be considered?
• What are the incentives or driving factors that would push our organisation to effectively partner with existing data gathering networks instead creating/curating our own data resources?

Possibilities for application and the way forward

The questions shared above can act as litmus test and help orient actors in the field who are building resources, tools and guides with greater direction on what the needs and demands of civil society may be going forward.

These insights are also being leveraged by the Data Economy Lab to shape an upcoming playbook. The playbook aims to guide for stewards and their supportive entities (policymakers, funders and civil society organisations) on what steps to take to build and inhabit a thriving and diverse ecosystem of responsible data stewardship. It also specifically considers how these plays will map to specific domains — starting with environmental conservation & sustainable development and followed by health and urban governance.

To learn more about our emerging work, reach out to us at contact@aapti.in

Situating Civil Society Organisations on the Stewardship Spectrum was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

When the Data Economy Lab was established in 2019, our vision was to map the landscape of stewardship, delineate its various operative models and document practices of existing stewards or steward-like initiatives across the globe. Our report furnished a common taxonomy to unify the vocabulary around data stewardship and presented a classification framework that built on analysis of existing use cases of stewards. A key area of inquiry for our research has been the sustainability and scalability of these entities. To this end, our paper on the ‘Principles for Revenue Models of Data Stewardship’ helped frame our initial thinking around incentive structures and related business models and guided further research around these entities.

While our research has highlighted the existence of several organisations or initiatives pushing for a more fair and just data economy, it has also become clear that not all of them have been able to survive the test of time. This has prompted a retrospective analysis of the stewards we had previously studied to understand their status — are they still operational? What are their business models? How have they managed to incentivise participation of individuals, communities or other end-users? Have some stewards pivoted in focus or approach in response to an external trigger from the legal or policy environment? What has contributed towards the growth of some stewards and the dissolution of others? Is this a function of funding, demand or capacity?

This blog attempts to tackle a few of these questions and provide a snapshot of the ecosystem at present — three years since we kickstarted the Data Economy Lab’s research on data stewardship. To this end, we’ve looked back at the 70+ stewarding entities to understand how they have sustained or scaled by building partnerships, securing funding, incentivised use of their services and/or expanding their teams. Drawing from our insights, we’ve highlighted common patterns and trends we’re seeing in the stewardship space.

These findings will feed into the research we’ve charted for the next phase of our research at the Data Economy Lab. Going forward, we will continue to identify barriers as well as enablers for stewardship. This will require unpacking the technical, social and financial levers required to create a thriving ecosystem for stewards and other stakeholders to co-exist.

Methodology

We started by updating our existing database of 70+ stewards to include details on their operational status in order to establish if the steward continues to remain active. This database contains a longlist of stewarding entities which operate across 4 sectors — environment and sustainability, healthcare, consumer welfare, and mobility and smart cities. Their operational status was examined through metrics such as their publications, collaborations with partners, funding models and their legal status. Lastly, the metrics also helped us glean how certain entities have persisted through the years and functioned as scalable models.

Limitations

The status of these models was determined through pre-existing data and secondary research which may not always be reflective of the actual picture on ground or sufficient in deriving the full picture of a steward’s activities. Where stewards were found to not be functional, it was not possible to identify the cause or experience that led to this outcome — which may be further explored in the future through carrying out in depth interviews. Further, the sample size is in itself limited and primarily draws from the Data Economy Lab’s network of actors in the stewardship landscape.

Result and analysis

Common trends in the stewardship landscape

• Stewarding entities in the health sector are particularly resilient

It has emerged that stewards in the health sector are particularly resilient, with only 1 of 24 (4%) entities examined which are no longer operational. On the other hand, 33% of entities in the environment and sustainability sector (3 out of 9) are currently inactive, with corresponding figures for mobility and consumer welfare sectors standing at 23% ( 3 out of 13) and 18% (5 out of 27) respectively. This could be attributed to the pandemic which saw an increase in demand for digital health solutions that consequently led stewarding entities to establish several continuous partnerships with healthcare management companies, data providers and successfully incentivised participation of citizens to contribute their data. On the whole, 58 of 72 stewards considered continue to remain operational since they were first examined as part of DEL’s research in 2019.

• Blockchain technology as a means to enable individual ownership of data

Stewarding entities like Geens, and BurstIQ reflect the trend of applying decentralised technology solutions such as blockchain to enable individuals to exercise control over their data.

Decentralised web3.0 claims to solve existing issues with data ownership by offering both sovereign data ownership and social/ trusted ownership. ‘Data Unions DAO’ is an interesting scalable marketplace for data which is community-owned and controlled. Data Unions allow individual users to join the organisation through smart contracts and aggregate data and monetise the same in a controlled manner with the ability to vote on how and where data is used through an explicit DAO governance protocol. DAO organisations like Streamr and Swash offer Data Unions as their products. However, it remains to be seen how such blockchain-based solutions can build and sustain ‘data unions’ in the long run, as they would require a critical mass of users to generate data for sale. Additionally, these “decentralised” solutions seem to have a limited focus on matters relating to governance, participation and representation within the structure of a ‘data union’ — foundational qualities of unions in the offline world.

• Data trusts remain largely esoteric with limited practical resonance and application

There has been growing interest in data trusts as a model for stewarding data. This could be observed, for instance, in the spike in inquiries about data trusts on search engines worldwide over the past year. This can likely be attributed to the fact that data trusts have been positioned as promising governance models whose core functions include: institutional safeguards to manage data misuse, empowering individuals and communities and providing ethical, architectural and governance support for trustworthy processing of data.

Of the 72 models documented in Aapti’s database, there exist only 2 data trusts — both of which have been disbanded since 2019. This includes Sidewalk Toronto’s Civic Data Trust and the ODI Wildlife Techhub data trust pilots. Consequently, the lack of experimentation and practical application around data trusts is a function of its unique legal status, with genesis in common law trust frameworks and attendant fiduciary obligations attributed to trustees. As a result, data trusts have limited feasibility beyond common law jurisdictions due to varied political and legal histories of different countries that determine data rights afforded to citizens as well as their enforceability of trust laws within the country. Considering these challenges, it is imperative to look to other models of stewardship like cooperatives and collaboratives that have demonstrated greater potential for empowering communities through alternative mechanisms of equitable data governance that do not rely on common law trust frameworks.

• Private sector initiatives tend to be more sustainable when compared to entities funded by the government

Most private sector initiatives tend to be more sustainable when compared to organisations that are funded by the government. Out of 12 data exchanges in the database that were funded largely by the government, 25% ceased to function. Our analysis revealed that of the 11 projects which have been disbanded, 60% of them were either funded by the government or multilateral organisations such the UN and World Bank.

• Most EU-based private sector initiatives generate revenue through subscription or membership models

Most private sector initiatives and cooperatives examined in this study (such as Geens/GebiedOnline/Schluss/Midata/PolyPoly) generate revenue by incentivising members to pay a small subscription fees for secure storage of and control over their data. All of these organisations are based in the EU — perhaps suggestive of an appetite for ordinary individuals to engage and avail the services of stewards.

• Evolution of the policy landscape has meant that data stewards are now in focus

Innovative models cannot achieve their full potential if the landscape is lacking in mechanisms of accountability, robust data governance, advisory protocols and tangible pathways to enable community participation in data decisions. Emerging policy developments that serve as enablers to stewardship include the Data Governance Act by the European Union which refers to the role of intermediaries in the digital economy and how stewardship can channelise socially productive data use to benefit the public. Similarly, through the introduction of the consent manager (CM) framework in India, regulators are beginning to recognise the extractive digital economy and offer the CM framework as a means to uphold user agency and data rights.

Scalability and sustainability of models in the data stewardship landscape

• Some models have scaled due to partnerships with data providers, technology firms and local state agencies

The pandemic saw an over-reliance on digital health solutions that promise to overcome the bottlenecks of conventional health systems which were otherwise overwhelmed and unable to manage stresses created by COVID-19. In such a milieu, data intermediaries in the health sector have played a crucial role by stepping up investments in technology and capacity building through innovative partnership and data sharing among public health, technology firms and healthcare providers. Redox is an interesting use case that provides data standardisation services through its unified protocol for electronic health records maintenance — bringing providers, insurers and patients under one framework. Saluscoop is a health data cooperative which facilitated citizen participation in health research as a part of its CO3 project, expanding the scope for collaborative health data governance.

Collaboration with state agencies comes with immense potential for scalability for stewards given the availability of data that rests with public bodies. For instance, in 2022, Findata undertook a joint project with Finnish municipalities to use data compiled from data controllers to examine how cities can bridge gaps in delivery of welfare services. The main advantage was that Findata could access and use data across diverse registers which helped provide a multifaceted perspective on the state of welfare services provided to families through analysis of their educational, income and health data. The Yale University Open Data Access (YODA) Project acts as a data intermediary to facilitate the sharing of clinical research data between members of academia, government, and private industry for the purposes of meta-analyses, replicating trial results, building upon prior bindings, and conducting secondary analyses.

Elsewhere, Strava Metro partners with public agencies such as Departments of Transportation, Metropolitan Planning Organisations, counties and cities like Sydney, Oslo and Sao Paulo to improve infrastructure for bicyclists and pedestrians.

In India, the Foundation for Ecological Security which houses The Indian Observatory program brings together 1800+ disparate government and open datasets to create contextually-sensitive decision support tools on natural resource management. These examples are indicative of the need to foster greater engagement/collaboration between public sector agencies.

• Some models have rebuilt their business structures to improve product offering and financial sustainability

Certain stewards have restructured their business models to make their products and services more accessible and/or automated. For instance, Idaho Health Data Exchange underwent a significant revitalization of its business model and technology stack and service offering by partnering with Orion Health (provider of health management solutions). Subsequently, its migration to the Amadeus platform proved to be beneficial in enhancing value for patients. This is because Amadeus is a highly scalable and open platform whose API layer allows for 3rd party developers to build new capabilities, enabling best health services applications to be developed while fostering innovation.

Such use cases illustrate how data stewards need to forge creative business models that provide sustained sources of funding and revenue for themselves. Other examples that stand testament to this insight include Luna DNA whose governance model centers around recognizing the value of individuals’ data and compensating them for sharing genomic and health data for research. The Luna model ensures that users are provided agency over their data and are adequately compensated for its use, ultimately creating value for its members as well as revenue for itself.

• Some models engage and empower communities as a way to unlock societal value of data

Cooperatives like Drivers Seat imports gig worker data and seek to help drivers harvest industry data and reclaim control over their data, furnishing a valuable precedent for community mobilisation and group enfranchisement around data. Gebied Online, another cooperative, seeks to impact greater control over member’s personal data and is driven to reinvest profit back for the benefit of the member community. Abalobi is a data cooperative that allows fishermen to collect data and negotiate the conditions and parties with which such data is shared, thereby increasing their bargaining power.

Other examples of similar stewards include Variant Bio which works with historically marginalised populations to facilitate people-driven therapeutics. At Variant Bio, prior to the start of each research project, communities are consulted to determine the objectives of the study and ensure that their data is used within a framework that focalises community prerogatives.

Networks of solidarity in the stewardship ecosystem

The stewardship ecosystem has expanded since 2019, building communities of concern around equitable data governance. The stakeholders within this ecosystem seek to support each other along a few important pillars outlined below:

• Funding & mentoring support

Initiatives such as Mozilla Foundation’s Data Futures Lab provide a space to explore new approaches to data stewardship by providing funding and facilitating collaboration has onboarded four organisations — Place Trust, Driver’s Seat Cooperative, Drivers Coop and Digital Democracy as grantees. As part of this program, organisations receive $100,000 in funding and access to a network of experts to implement user-centric data governance policies. These projects will help pilot new models of stewardship and provide much-needed financial impetus within the ecosystem. Elsewhere, Open Data Manchester, another non-profit, has convened a working group on data cooperatives for stakeholders to connect and learn from each other. Zebras Unite is another data cooperative which connects its members and investors with each other, which demonstrates the potential of democratic networks and infrastructure to raise capital.

• Technology, infrastructure and capacity building support

Community-oriented tools like Mapeo built by Digital Democracy was designed to capture environmental data and evidence of human rights violations, and have received support and funding from organisations like the Open Data Institute (ODI), McGovern Foundation and Mozilla Foundation. The ODI is also lending technical assistance to Mapeo to build inclusive tech for indigenous communities and develop protocols for indigenous data sharing.

PolyPoly, an open source EU cooperative, is another ecosystem enabler which merits exploration. They are leveraging decentralised EDGE computing solutions in order to enable resource saving, scalable and GDPR-compliance. The interaction between enterprise (which supports business interaction and creates products like EDGE solutions), cooperative (which connects individual users) and foundation (which deals with governance of members’ data) are mutually profitable and ensures secure data flow across its supply chain.

Dasra in partnership with Bloomberg, Societal Platform, Stanford PACS and the Bill and Melinda Gates Foundation are hosting a series of initiatives under the Data for Good Exchange 2022 India. The aim is to empower organisations to create data-driven impact for alleviating India’s most vulnerable communities. These initiatives include learning lab sessions to surface solutions capacity building on ground and supporting NGOs with design grants.

• Community mobilisation & participation

Ecosystem support can go a long way in empowering community-led stewards in the environment and sustainability domain. Indigenous communities have been guardians of traditional knowledge, but are seldom involved in governance of data derived from and relating to them. In this context, the role of ecosystem enablers such as the ODI and ICES are pivotal to the journey of indigenous communities, as they begin to take reins over their data.

Data for Black Lives (D4BL) invests in research and spreads awareness on how algorithms and other technologies have furthered the legacy of marginalisation and invisibilization of Black communities within mainstream science and technology discourse. D4BL’s policy working groups (that include lawyers, scientists and technical experts) are building frameworks that relate to data governance — and uniquely acknowledge how data and technologies amplify historic discrimination and systemic oppression. Open Humans is a non-profit funded by grants from Shuttleworth and Knight Foundation that has adopted a participatory approach to healthcare research where patients are involved in framing research questions and are consulted periodically.

• Principle-building and inter-stakeholder collaboration

Global non-profit organisations like OMF, NUMO collaborate and act as ecosystem enablers. These multi-stakeholder and multi-domain networks bring together important players to help harmonise understandings about data governance, navigate diverse interests and incentives around data and in some cases also create common standards or guiding frameworks or principles. For instance, they help in launching privacy principles for mobility data to assess policy or technical decisions and their implications on privacy. OMF, an independent nonprofit, offers a collaborative environment for municipalities, civic nonprofits, corporate stakeholders, subject experts, and the public to work together in order to solve emerging transportation challenges that affect communities.

Cities for Digital Rights, a network of cities committed to promoting digital rights in the urban context through city action, has a sandbox that aims to become a testing platform for innovative products and services, as well as to support the development of local ICT and entrepreneurial ecosystems and to generate data to scale up and export these products.

The way forward

Our retrospective scan has highlighted a few areas that must be further investigated. To start with, it is necessary to better understand the appetite for stewardship at an individual and community level. Early research conducted by Streamr Network and Swash illustrates how consumers are curious about the opportunity to take control of their data and money income helps consumers understand the value of data. Demonstrating the possible use cases and more clearly showcasing the tangible value of data can be a pathway to engage consumers and may be useful in creating stickiness around products and services offered by data stewards. This is an important area to address as without a critical mass of users, it is exceedingly difficult to scale and sustain many stewarding platforms.

One of the takeaways from our research is that users demonstrated an incentive to share and participate in data decision-making where they felt a personal resonance or alliance with particular issues (e.g. contributing to climate action efforts or powering more inclusive research studies). In some cases, mobilisation around data and its governance has been propelled by a failure on behalf of governments in addressing data related injustices — stewards are therefore perceived as intermediaries that can facilitate collective action and solutions.

One of the roles that stewards often play is to make data accessible to diverse stakeholders, considering varying levels of data literacy and perceptions of its value — this necessitates building more inclusive data governance architectures on the basis of stakeholder consultation and involvement. There is a need to develop both the social and technological infrastructures that can drive data accessibility. Participatory methodologies that foreground stakeholder consultations with the purpose of identifying concerns about data sharing and delineating data priorities can empower communities to be involved in data governance. While there are several steward-like initiatives thriving across sectors and use-cases, our research shows that there are many who face challenges in sustaining and scaling. Varying support from the ecosystem and regulatory systems, policy pathways and governance principles will be required for stewardship to gain more ground. Therefore, a sector-wise ecosystem diagnosis into the regulatory, infrastructure and the funding models of these organisations is needed to unpack and understand each of them.

If you are interested in learning more about this work or collaborating with us on any of these questions, please get in touch with us at contact@aapti.in.

This work is supported by Omidyar Network

State of Data Stewardship was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

Burgeoning digitization continues to move beyond personalised services, and now plays the role of transforming governments, policymaking processes and our integral relation to the state as citizens. As a part of this trajectory, there has also been ongoing development and deployment of artificial intelligence (AI), iOT and blockchain based solutions by the public and private sectors. There is little doubt that these technology and data-driven solutions can be pivotal in creating impact — helping governments identify and design for gaps in service delivery and program implementation, better collaborate with one another and define new strategies and innovation to tackle pressing global challenges like our climate crisis.

However, as underscored by the experiences of individuals and communities across the world, without intuitive offline and online architectures for control, decision-making, transparency or accountability, there is a risk of further marginalisation (evidenced most clearly in the lack of equitable access to vaccines) — pushing those with limited access or literacy of technology and data to be further sidelined. We can no longer assume that data-driven, predictive or automated solutions alone are sufficient in meeting the evolving needs of our society. Those committed to solving last-mile access challenges and building inclusive digital public infrastructures must consider the role of existing human architectures, intermediary organisations, and support networks in addressing these issues.

In the Data Economy Lab’s journey to build a responsible, agential and safe digital world, we have put forward data stewardship as one part of a broader set of necessary solutions. Data stewards are intermediary organisations or institutions that seek to empower individuals and communities, providing them with the ability to control or make decisions that relate to their data and allowing them more directly benefit from its value. Pushing the boundaries of participation beyond largely ineffective notice and consent mechanisms, stewards can be embedded with a greater duty of care and loyalty towards their subjects. This can extend to instilling meaningful accountability and safety mechanisms for principals and creating pathways for more substantive involvement in data decision-making and mechanisms for accountability.

It’s also worth noting that the increased proliferation of artificial intelligence and technologies will depend on the availability and aggregation of high-quality, trained datasets. Acknowledging these realities, further exploration is required into the labelling of the data (and who carries out this labour) and what frameworks are necessary to ensure the collection, governance and use of this data are ethical and rights preserving. Particularly for global challenges like climate change that necessitate these forms of data exchange from multiple sources, stakeholders and streams — an ecosystem of stewards will be a necessary piece of the puzzle.

Taking this into consideration, we envision that a diverse ecosystem of data intermediaries or guardians will be required to thrive in addition to, or often as part of, democratic public digital infrastructures. These stewards will play a number of roles — both technical and organisational — but will be driven by a larger agenda: to unlock the societal value of data while empowering and safeguarding the rights of individuals and communities.

Where we’ve gotten so far

Our research on data stewardship began with an exploration of the landscape, to understand how the concept has been translated on-ground and in what form. The taxonomy report we released in 2020 was one of the first of its kind that sought to delineate various models of stewardship that include: data collaboratives, personal data stores, trusts, exchanges, and marketplaces. This later was expanded to include data repositories, cooperatives and other organisations that act as ‘ecosystem enablers’.

Recognizing the diversity in application across use-case and sector, we realised that our investigation required a sectoral, problem-led lens that could account for the nuances in how stewards could be designed and deployed. To this end, we carried out deep dives to understand both the opportunities and domain-specific challenges that may arise when considering stewardship’s application across sectors focusing on healthcare, agriculture, climate change, civil society and mobility/smart cities.

Feedback and conversations with startups or emerging stewards led to the realisation that our insights must be actionably surfaced back to the community (researchers, emerging stewards and policymakers). Responding to this, we released the Stewardship mapper which outlines the variation in structural choices and considerations while building a steward. Taking this a step further, we transformed this into an interactive guided questionnaire, the Stewardship Navigator. This tool (released in beta mode) has been designed to help organisations better think through aspects related to organisational structure, business model, services provided and how tethered stewards can be to their end users, beneficiaries and society at large. We’ve documented our journey and vision for the tool in a detailed blog post.

More broadly, our guiding vision with this research has been to illustrate the opportunities and possibilities stewardship presents that could lend to a more equitable data lifecycle and value distribution system. To make this real, we’ve focused on uncovering new and existing models of participation, responsible data exchange and inter-stakeholder collaboration around data and its governance. The Tracking Stewardship video interview series we produce, showcases these efforts and provides founders and pioneers with a platform to share their experiences.

The way forward

Our axis of inquiry till this point has focused on the steward itself — what form it should take — depending on sector, use case and purpose, mechanisms it could employ to enable participation and representation and the business models it should be anchored by to maximise privacy while limiting possibilities for extraction of data. However, our research indicates that while there is a growing set of stewards or steward-like initiatives across sectors and use-cases, many face challenges in sustaining and scaling. Differing regulatory environments and varying support from the ecosystem (funding, infrastructure, appetite for stewardship/data intermediaries) must be further unpacked to understand what support is best required and where.

Therefore, going forward, we believe instantiating stewardship will require top-down and bottom-up support. From a national, regional and international level, policy pathways and governance principles will be required for stewardship to gain ground. Ensuring these systems remain community-centric and accountable also necessitate building the capacities, infrastructure and models for engagement to empower beneficiaries (or data subjects) from the bottom-up.

Taking this into account, our research aims to delve deeper into these foundational layers required for stewards to thrive at an ecosystem level. We aim to do this by analysing a few different dimensions including the regulatory landscape and funding ecosystem. We also aim to understand both the enablers and barriers that stewards have come up against at a technical, legal, social and economic level.

A key component of this research and analysis will be to uncover how stewardship will fit into ongoing policy formulation globally. Current debates around the Data Governance Act in Europe have already made reference to the role of intermediaries in the digital economy. It highlights how stewardship will be a critical bridging mechanism for data subjects and data controllers and may channelise data use in ways that benefit the public. In a similar vein, the Indian regulatory ecosystem is re-imagining their consent architecture, in an effort to uphold user agency through the introduction of a consent manager framework. While these developments are promising, further dialogue, consultation and research backing will be required to meaningfully flesh out functions of the steward including its advisory capacities and structures for accountability and grievance mediation. Focusing on these new approaches will also be important if we are to understand opportunities to adapt these frameworks and structures in other parts of the world.

Sectors & Core Focus Areas

Building on our existing sectoral expertise, our research aims to expand on our understanding of the state of data stewardship in the following areas and will explore a number of related questions or sub-themes:

Healthcare — Availability, Access, Innovation

• Strengthening public health management and interventions through technology & data (enabling regulation and defining ecosystem and infrastructural capabilities)
• Powering inclusive health research, innovation and treatment possibilities through responsible data sharing (Defining frameworks and building blocks for safe data collection, usage and exchange at municipal, national and regional levels)
• Building and scaling patient-centric stewards that empower individuals & communities while safeguarding rights

Urban Governance (Mobility, Smart Cities) — Infrastructure, Citizen Services, Civic Participation

• Strengthening city planning, public service management and delivery — building foundations for robust multi-departmental/-stakeholder data exchange — creating privacy-centric public digital infrastructures
• Data driven policy-making, governance and planning for a sustainable future — defining data ecosystems for new energy, waste and resource management
• Creating inclusive, accountable and effective citizen-centric systems online / offline — design principles, safeguards and participation mechanisms/structures

Environment & Sustainability — Climate change, conservation and sustainability

• Data solidarities & systems of resilience — supporting indigenous and local knowledge systems in enhancing adaptive capacities
• Enhancing the unlocking, pooling and collective value of data –research, community empowerment, green innovation, policymaking
• Mainstreaming bottom-up & citizen science efforts — participatory governance of data for conservation, climate-related policies and investment

Methodology

To understand unique needs and provide actionable recommendations in the form of guides, toolkits and audiovisual content, the Data Economy Lab aims to use a global case-study led approach that explores stewardship across sectors and jurisdictions to drive forward this research. This will involve the following steps

• Sector-wise Ecosystem Diagnosis — Dive into an in-depth review of the policy environment and emerging discourse around data privacy, protection, governance and analysis. Specific focus will be placed on one country as a case-study.
• Expert Interview & Consultation — Carry out a set of interviews with key stakeholders in each sector: data stewards, enabling organisations, peer research institutes/organisations, nonprofits/civil society organisations and policymakers to understand their experience — What is the state of stewardship in the respective geography, who is doing what, where are the stewards located, how do we make this solution real — what is needed?
• Dynamic Policy Tracking & Response — Track upcoming developments in the domestic (India) and the international policy landscape to understand the positioning and role of stewardship and offer necessary comments to encourage a more reflexive, citizen-centric framing or approach.
• Co-producing knowledge with partners — Engage with peer research institutes and existing organisations looking to play the role of a responsible steward and leverage this experience to understand unique pain points and surface possible design choices and policy experiments. Document these learnings with the support of the partner to underscore gaps and share findings for other organisations with similar use-cases.
• Translate findings into actionable tools and knowledge repositories — These findings will be brought together in the form of an overarching report that defines the way forward for stewardship. The report will be composed of an analytical overview (drawn from the diagnosis phase), a set of playbooks or tools/guides (extrapolated from expert insights, consultation and co-designed with our partners).

If you are interested in learning more about this work or collaborating with us on any of these questions, please get in touch with us at contact@aapti.in.

This work is supported by Omidyar Network

By The Data Economy Lab team

The Data Economy Lab 3.0 was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

The current ecosystem of health data sharing

Health data governance assumes significance due to the highly sensitive nature of information under question, potentially revealing confidential data on medical history, racial or ethnic background and financial status of individuals and the broader communities they belong to. While data sharing promises to significantly improve healthcare quality and health systems performance, unchecked sharing has adverse ramifications for access and affordability of health services for individuals, particularly when such information exists in siloes, dispersed across multiple sources. Individuals and communities — primary generators of this information — have little visibility on how their data is used by entities concerned with collecting and processing data.

In addition to challenges pertaining to availability of reliable and quality data, current ad-hoc, top-down contractual mechanisms for health data sharing employed by corporations, providers and public agencies do not address public concerns about trust, privacy and misuse of data. This highlights the distressing consequences of predatory mechanisms which expressly exclude communities from participating in data governance, deepening the asymmetries of our expanding data economy.

Need for stewardship

The extractive model of data relationships characterised above follows a growing realisation that access to data is highly lopsided and unequal, necessitating a framework which is grounded in democracy and inclusivity. Recognising collective rights over data is a significant step towards rebalancing the data economy, foregrounding user agency and accountability as imperative considerations for building responsible data sharing mechanisms.

To this end, data stewardship is emerging as a framework for equitable data governance, simultaneously unlocking societal value of data while respecting the autonomy of individuals and communities. Aapti’s research, alongside that conducted by ODI and Mozilla, have demonstrated the virtues of an alternative paradigm for data governance whose benefits can accrue to all stakeholders in the ecosystem — individuals, communities, private entities and the public at large.

Health data stewardship

Stewardship assumes importance in the context of the health sector where data sharing for public good and innovation is at odds with data rights and privacy of patients involved. During the current pandemic where digital health solutions promise to compensate for the failings of overwhelmed health systems, their lack of transparency has been a cause for alarm. A flagrant disregard for user agency is a dominant feature of such solutions which depend on instituting new data pipelines to support their services. For instance, a patient consulting a general physician through an interactive voice response system (IVRS) platform has limited knowledge about how the recorded conversation, likely containing sensitive health information, would be used/shared by the service provider.

Creating user-centric models of stewardship is imperative to empower individuals and communities to participate meaningfully in data decisions. Our research has relied on case studies of organisations operating as health data stewards, demonstrating the real impact of democratic governance frameworks. Examples include SalusCoop, a health data cooperative whose members can review medical research proposals and consent to share their data for specific projects. It also enables democratic decision-making through a general assembly of members accorded with one vote each.

Similarly, Ciitizen is a private collaborative platform that enables individual patients and larger patient advocacy groups for cancer and rare neurologic disorders to collect, digitise, and share their health information for research with pharmaceutical companies as well as hospitals for treatment management of individuals. The platform has introduced innovative modes of implementing user-centric health research wherein patients and advocacy groups are consulted while framing partnerships with third party entities who would gain access to their data, deciding the terms and purpose of such collaborations. Individual consent for participation in a specific research is essential to the operation of Ciitizen.

These models perform the twin tasks of unlocking the societal value of data by empowering communities whilst concomitantly creating public good solutions — evolving a potentially diverse set of data sharing mechanisms between patients, providers, academics and pharmaceutical companies to use sensitive health information for research and treatment.

Health data governance

While the merits afforded by stewardship have been outlined, the specific implications it holds for various stakeholders ought to be unpacked in light of the various governance considerations it involves. Here are three recommendations to implement a more equitable framework for health data governance for regulators and organisations looking to steward sensitive health information:

1. Increase transparency in health data sharing agreements: Existing data sharing agreements between healthcare providers, technology companies and hospitals remain opaque to patients and caregivers who are the primary data generators. The systematic exclusion of individuals from data sharing agreements has an invisiblising effect on their position within the data value chain. There is a need to increase transparency of agreements for patients and the larger public. This can be achieved by implementing mandatory disclosure requirements of stakeholders with respect to health data use/sharing.
2. Conduct regular audits: Regulators and citizen groups should conduct regular audits of data sharing practices among providers, private corporations and public agencies to ensure transparency, accountability and procedural fairness in the process of sharing. This step goes beyond conventional internal review board assessments which are merely concerned with protecting the privacy of individuals involved in research, while comprehensive data audits can highlight broader opportunities for health data sharing for social good.
3. Accentuate user agency in data governance: Users should be organised and empowered to negotiate with healthcare providers to have greater say in who uses data and to what end — mechanisms like data cooperatives (ex: Salus.coop, MiData) can be institutionalised to actualise this. Enhanced agency can impact existing patterns of consent — marking a departure from the current blanket and uninformed processes to more dynamic, process-based and granular approaches. These mechanisms can also make sure that data generators (patients and communities) are compensated either financially or through individual-centric insights obtained from their data.

As data value chains continue to evolve against the backdrop of the pandemic, there is a need to forge creative solutions which place patients and their caregivers at the heart of health data governance frameworks. Balancing data openness and innovation to inform public health policy, research and disease management without compromising rights, agency and security of data generators is the primary aim of health data stewardship — a mechanism whose essential roots lie in fairness, transparency and accountability.

This article was originally published in Business World on May 13th 2021

The lack of a cohesive government response in the latest wave of Covid-19 has forced citizens and civil society to fend for themselves. Citizens are acting as intermediaries, supporting others by pooling data and amplifying demands for oxygen, medication and hospital beds. This collective resilience is powered by the exchange of critical information, but has made the ability to collate, find and share verified data a key determinant in life or death situations.

To better sort through the high volume of requests and information, apps, websites or simple tools have been built to aggregate resources. Open source trackers are helping users log requirements, view compiled information, track beds and even identify opportunities to volunteer. Others are spreading awareness on how to generate more effective, customised search queries on Twitter.

Regardless of these efforts, most information is still scattered across platforms and mediums — reconciling details from images, texts and tweets is complex. Another growing challenge is the validity and relevance of this data. Several volunteers are manually verifying sources (and tagging as #verified) to ensure numbers are valid and that medical support is still available.

What also remains to be solved for, is how to protect our digital rights and privacy — a fundamental right. This is critical to address now, given that many people are openly broadcasting details like phone numbers, blood types, addresses, medical status and histories online to amplify needs and information. While this may be necessary to currently share, what happens when the crisis is over?

The exposure of personal, sensitive data on platforms is concerning as it could potentially be misused to further harass, surveil and censor citizens. Doctors, front-line workers and suppliers whose numbers are now in the public domain are inundated and burdened by calls. Twitter accounts sharing information on suppliers of oxygen and medicine have also recently been shut down or ‘shadow banned’. Past data breaches also point to the risk of this data being commodified and sold without consent to third-parties. Volunteers who recognize these risks advise blurring or hiding sensitive information before posting (or re-posting), however, this may not be a sustainable solution.

These are reminders that we need to retain agency and be active participants over our data journeys, especially in context of calamities. Citizens have largely taken control over relief and access efforts (albeit out of dire circumstance and necessity), and are witnessing the value of a bottom-up approach to crowdsourcing and sharing hyperlocal data. However, existing platforms/systems are ill-equipped for effective bottom-up data governance.

Bottom-up data governance systems can help put power and agency over data, its use and protection back in the hands of individuals and communities. This starts with recognizing that civil society organizations can play the role of emergency guardians or stewards of data. We already trust that many grassroots and community-based organizations possess a ‘duty of care’ to work in the best interests of those they serve, it’s time for this to translate to data and digital rights.

Lessons can be drawn from existing organizations on how this could work in practice. Digi.me streamlines user data from existing sources and facilitates consent in real-time. PolyPoly’s polyPod vault is similar but data decisions and the development of their technology take place through a cooperative — potentially well suited for a community-oriented solution. Swash has built plugins that capture and pool data from existing social media platforms to enable users to better derive value. In theory, easily embeddable tools like these could trawl for and aggregate data to strengthen matches between requests and resources, help visualise the proximity of nearby services and map and alert suppliers on aggregate demand.

Considering the growing technical divide, which renders those with digital literacy and networks the power and privilege to access support, data stewards can also serve as important bridges to connect the last-mile. Stewards can represent and surface requirements of those who have been invisibilized, and potentially negotiate on their behalf. The ongoing pandemic has made clear that crises have calamitous impacts on our data rights, and we need to create systems to become active participants in our data journeys — to protect and harness the value of our data. Bottom-up governance models are a step toward ensuring that these approaches remain community-oriented and can scale and help existing efforts be more effective.

India’s Community-led Crisis Response: How Can Bottom-up Data Governance Help? was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

1. ‘Communities’ as conceived by the Committee under Section 7.2 is ambiguous and risks causing more harm than good to their interest

The report makes communities the owners of NPD, with a community being defined as “any group of people that are bound by common interests and purposes, and involved in social and/or economic interactions.” A community could be virtual or delimited by common geographic, livelihood, economic or social interests. However, It is possible that a group of individual data principals sharing a common interest, say subscribers of a milk delivery service by BigBasket Daily in a particular locality, might not be aware that they are indeed a “community”. How such communities would create a Section 8 company, Society or Trust and choose a Data Trustee to represent and protect their interests is challenging to comprehend, particularly when the individual data principals involved in common social or economic interactions do not view themselves as a community.

As also, the Report is unclear about the terms of association between data principals and a larger community. Consent, particularly community consent, is another dimension that the Committee has not considered in report. In its current form, individuals must merely consent to the anonymisation of their data and are given an option to withdraw the same, prior to anonymisation. An effort must be made to involve communities and/or their representatives at every step of the data value chain — from collection to processing through to data exchange and use by third parties — to ensure that their data is used for social benefit.

2. Data Principal v. Community: Lack of vision in organising principles for Communities and Data Principals; lack of procedure for protection of rights

The NPD Policy has stated that the creation of regulatory structure including Data Trustees, Data Custodians and Data Processors, the Non-personal Data Authority (NPDA) and High Value Datasets (HVDs) will help Data Communities exercise rights over datasets that are relevant to them.

It is important to note that this objective has not been achieved. Paragraph 7.7 states that Data Trustees are “obligated to establish grievance redressal mechanisms so that the community can raise grievances”. It is unclear as to why the burden of coming up with the procedure of instituting grievance redressal mechanisms has been laid completely at the feet of the Data Trustee, without principles on how it ought to be designed. At the level of an initial analysis, the process should comply with tenets of principles of natural justice and sound administrative law. Excessive delegation is a violation of these principles, and the Committee must elaborate on the form of recourse available to Data Communities in order to engage with the activities of Data Trustees.

3. The Report requires a clearer articulation of the harms from which it seeks to protect Data Principals and Communities

The Report states that Data Custodians and Trustees have a right to act to minimize harms, and they are in fact obligated to carry out this task by virtue of their duty of care towards the relevant Data Communities. What forms these harms take, however, has not been specified.This lack of detail leaves stakeholders in the dark as to what possible claims or causes of action they may have. Definition of harms under the Report cannot be left to the imagination of stakeholders.

The Report must lay out the processes involved and form of submissions to be made as part of the grievance redressal requirements prescribed in the Report under Paragraphs 7.2(ii), 7.7(ii) and 7.4(iv). Clarity of procedure and uniformity of process is necessary for transparency in protection of harms against Data Communities and Principals. The Report in its current form ignores Constitutional principles of procedural justice.

4. The Report fails to make a convincing case for the NPDA outlined in Section 7.10 and avoids any discussion on the possibility of sectoral overlaps that could manifest as enhanced regulatory burden for data-driven businesses.

Despite its attempt at articulating the nature and roles of the NPDA, the Committee has not made a convincing case for the institution of an independent regulatory authority. For one, a variety of existing mechanisms could achieve the same ends as that of the NPDA. An MoU between two consenting parties, say a hyperlocal delivery platform and a govt. agency, for release of data relating to self-employed gig workers in order to ascertain the number of informal workers in a city, could result in a similar disclosure of data, without the regulation of an overarching entity such as the NPDA.

Further, the Committee has failed to consider the ramifications of regulatory overlaps as a result of the creation of an NPDA. Firstly, the jurisdiction of DPA vis. NPDA is ambivalent inasmuch the presence of confounding variables such as mixed datasets and re-identification of NPD creates challenges for outlining clear adjudicatory mandates for the two authorities. In this context, the regulatory parameter for NPDA is undefined and requires more deliberation on the principles that compel the creation of the NPDA. As also, the Committee proposes that HVDs be excluded from copyright protection. The rationale offered is that formulation of HVDs does not entail any creativity or skill, but mere compilation of existing “fields of data”. However, the existence of certain “fields of data” might be a product of innovation and skill, producing competing claims for “trade secrets” protection simultaneously. In such a case, the mandate of the NPDA to compel disclosure by Data Custodians will necessarily be in conflict with the existing IPR regime.

The Report also suggests that sectoral regulators can delineate protocols for data sharing, in addition to those imposed by the sector-agnostic horizontal NPDA. This could result in a situation of regulatory excess where data-driven businesses will have to comply with multiple legislations: the PDP Bill in cases of personal data, the NPD framework for anonymised data and a third set of compliance obligations outlined by sectoral regulators. Efforts must be made to harmonise compliance requirements across regulatory bodies such as the DPA, NDPA and other sector-specific authorities to ease the consequent regulatory burden on businesses.

5. Dichotomy between the functions of a non-profit under Paragraph 7.2(ii) and a Trustee under Paragraph 7.7 of the Report, leading to inadequate protection against harms from processing of NPD

The role of the Data Trustee is to protect the interest of its user groups and communities. The Data Trustee firstly plays a pivotal role in the creation of the relevant HVD. However, the community relevant to the HVD cannot approach the Data Trustee to express their concerns on its usage through a grievance redressal mechanism. In order to do this, they are required to approach a different non-profit, as described under Paragraph 7.2(ii). Any engagement on harm experienced by Data Communities is required to go through this alternate regulatory route, as opposed to the better-established framework of the Data Trustee under the Report. Given that articulation of economic interests of the Community is in the hands of the Data Trustee, it is unclear as to why the prevention of harm has been delegated to a different entity. Both these processes aim to protect interests of the same group — the Community impacted by the HVD. The Report adds to regulatory burden on Data Communities when different platforms are required to be approached for different concerns.

It is recommended that the responsibility of protection against harm be given at least the same strength of grievance redressal process and accountability standards as that of protecting the interests of Data Communities in granting access to HVDs. Potential harm should be an explicit consideration in the Data Trustee’s decision in granting access for HVDs to any data requester.

6. Anonymisation and data sanitization with respect to “mixed datasets” are loopholes in the framework that could be used to circumvent mandatory data sharing obligations under Section 7.4 (iii).

The mandatory data sharing obligations sought to be imposed on Data Custodians and Data Businesses by the Revised Report introduces complications of regulatory arbitrage and enhanced compliance burden on private entities. This is highlighted by the clauses on anonymisation and governance of mixed datasets — wherein Data Custodians/Businesses could simply alter their terms of service and choose not to anonymise data or claim that they operate with mixed datasets (containing both PII and NPD) — which could be used to circumvent the NPD framework itself as explained below.

This above is problematic because the creation of structured HVDs for mandatory sharing, including raw and factual datasets, is a resource-intensive activity which could potentially disincentivise Data Custodians and Data Businesses from anonymising data at all. In fact, it is possible that mandatory data sharing could engender practices of regulatory arbitrage — a phenomenon by which Data Custodians could circumvent mandatory data sharing obligations by explicitly choosing not to anonymise data. Regulatory arbitrage is all but common in the tech industry where companies like Facebook have modified its terms of service to escape stringent privacy guidelines under the GDPR.

In the Indian context, Data Custodians could avoid the mandatory disclosure obligations imposed on them by either choosing not to anonymise user data or claim that such data is a “mixed dataset” which is “inextricably linked” to personal data of users, thereby bypassing the mandatory data sharing obligations. Consequently, the Data Custodians would not fall under the purview of NPD framework, but that of the Data Protection Authority mentioned in the PDP Bill. This glaring loophole in the NPD framework is only strengthened by a lack of coherence in the definition of a “mixed dataset”, specifically the scope of what constitutes as being “inextricably linked” to personal data. Whether such a link is determined by economic or privacy considerations is unclear although their association to personal data might be tenuous, at best.

Alternatives to Mandatory Data-sharing: Creating a voluntary sharing ecosystem

A plausible alternative to mandatory data sharing obligations imposed by the NPD framework is the creation of “data marketplaces”. Data marketplaces would provide an avenue for data providers/suppliers to offer curated datasets at a price where data consumers/requestors can purchase or subscribe to such datasets. Similarly, the proposed EU Data Governance Act, 2020 makes provisions for “data altruism”. Data altruism refers to a mechanism of data sharing in which data subjects (data principals in India) consent to processing of their personal data, or permissions of other data holders (custodians, processors) are obtained to authorise the use of their non-personal data without reward, for purposes of public welfare, such as scientific research or enhanced delivery of public services.

For more information about Aapti’s submission to the Committee, please visit the link here.

Comment on the Revised Report by the Committee of Experts on Non-Personal Data Governance Framework was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.

The need for Responsible Data Governance for Impact

Accelerated by the pandemic, civil society organisations (CSOs) are increasingly capturing data or partnering with other entities to unlock relevant information for societal benefit. Data-led efforts are critical, particularly in times of crisis when the lack of access to reliable administrative data can result in threats to individual and community livelihood, rights and dignity. For instance, Jan Sahas, a community-based organization that focuses on workers’ protection and safe migration in India, stores and manages massive volumes of data through their new initiative, the Migrants Resilience Collaborative, which includes a database of over 1.5 million migrant workers to facilitate relief, help them find employment and enable access to government services.

In contrast to past resistance and distrust of technology and data, the value of data is increasingly recognized by CSOs — whether it is to guide decision-making, act as an auditor for upward and downward accountability, help assess gaps or identify needs in service delivery, or be used to measure impact. This change in approach has in part been bolstered not just by the ubiquity of data, but also the vast data requirements for accessing any goods and services, made available by the government and private sector.

The increasing burden and demand on CSOs to be more data-led assumes that CSOs already possess technical skill sets and capabilities, when in fact, many are just starting their ‘digitization’ journeys — moving offline or analog systems online is a challenge in it of itself. Furthermore, without a focus on building foundational technical architecture or creating principles for collecting and governing data or investing in capacity, data efforts by CSOs often land up being “bandaid solutions”, at best.

As data becomes more ubiquitous to the work of CSOs, it’s important to ensure that these processes do not de-link data from the experiences and realities in which they are produced. In the face of rapid datafication, this also means being cognizant that vulnerable groups are most threatened by the lack of agency, transparency, and accountability around data which has often been used for further exclusion, surveillance, and exploitation. These concerns are amplified by data security risks. Even for larger international institutions, being able to securely manage data and enforce governance measures has been a challenge. For instance, in 2019, as a result of a data breach on a UNICEF learning portal, data on 8000 learners was released.

Taking note of these realities, a responsible data steward could preempt these risks and create necessary safeguards to protect users/beneficiaries against malicious attacks. More broadly, a steward’s responsibility would also be to ensure principles of data minimization are applied to prevent the overcollection of sensitive data, impose access controls or restrictions and demand transparency around data use and sharing — all elements that currently remain obscured. However, without a complete understanding of their role in the data economy and capacity to operationalize the role of a responsible data steward, CSOs may end up reinforcing data extractivism, instead of leading the resistance against it.

Imagining non-profits & community-based organizations as Good Data Stewards

Existing regulations like GDPR initially have already required civil society organizations to rehaul their systems for the purpose of compliance. Establishing data storage and handling policies were previously seen as a way to maintain ‘organizational resilience’ against potential threats or attacks by adversaries. Now, emerging policies are taking this a step further. NGOs and non-profits may be expected to play a more active role as ‘data custodians’. The proposed Non-Personal Data Protection Governance Framework in India, for example, suggests that communities must organize, and exercise their data rights through a non-profit organization.

A few organizations have started to demonstrate an interest in being responsible stewards of their community data, but the instantiation of protocol and governance practices still prove to be complex. The World Food Programme published a set of practices on data management/handling but were called out when an audit revealed glaring gaps in protection and safeguards when assessed for their own implementation.

Examples like these indicate a few areas of concern, particularly around capacity. To begin with, many smaller nonprofits often work in remote areas with limited bandwidth and maybe early in their stages of engaging with online systems for data collection. They also often do not have dedicated technical resources or may not be able to divert significant funding into building/adopting technical infrastructures that are human-centric. As a result, they tend to rely on outsourcing technical solutions or products — where possible.

There is little doubt that leveraging existing design tools for data collection, storage and management can prevent organizations from needing to recreate the wheel, save costs and most often increase efficiency. However, commercial tools and products that tend to be readily adopted by civil society, are often not designed to maximize privacy by design. In other cases, even where data collection tools are customized for nonprofits requirements and needs, our emerging research indicates that the ability to collect consent, govern data decision-making, facilitate agency of data generators are not typical services these platforms offer. These data governance aspects are usually left to the will and control of the end-user (in this case, CSOs) who may not prioritize or commit to these responsibilities — especially if they have limited resources or are in their early stages of operations.

While it has been observed that some non-profits are more readily understanding the potential value data can hold and are increasingly engaging in its collection — it is still often treated as an abstract or unused asset. A study carried out by Everyaction showed that 60% of don’t use data for decision making. Another major finding from this research showed that in organizations that do collect and use data, 49% of employees were unaware of its collection or governance. This is telling of organizations ability to fully take advantage of data but also sheds importance on how these processes may not be transparent and accessible to understand for stakeholders within and outside of the organization.

Although these concerns are important to surface, CSOs are still well placed to commit to a ‘duty of care’ on data rights — as an extension of their existing commitment to the individuals, communities and groups they serve. Recent writing is beginning to realise this role and argue that non-profits must play a bigger role in data rights as most organizations in this space are well embedded in communities and have established long-standing relationships of trust — a critical foundation for responsible and representative data governance.

If we are to envision the contours of this new role, CSOs could potentially also act as effective bulwarks against the range of private, profit-driven companies whose motivations lie in data extraction and unconsented brokerage. Similarly, CSOs could lead the way for individuals and communities to access and exercise their digital rights and strengthen their collective bargaining position when interacting with the State.

CSOs can also play a critical role in creating opportunities for greater representation or participation in data decision-making. This proposition also raises the concern of imposing an additional burden, similar to ‘consent fatigue’ around data governance on individuals and communities, particularly those from disenfranchised or vulnerable groups. It also remains to be seen whether individuals and communities would be motivated to engage in these affairs of data governance — particularly where a clear value proposition may not exist.

In thinking about these questions and how to collaboratively build solutions to enable a greater degree of transparency, accountability, participation, and security around data captured or unlocked by civil society we believe a layered approach to data stewardship is necessary.

Our emerging research and conversations with civil society organizations across the globe have helped us identify potential best practices that are useful to draw from. We’ve spotlighted these case studies by categorizing them as two broad, initial steps in the process of transforming into a data guardian:

STEP 1: Data literacy & community-building

Build communities of concern to equip them with knowledge, skills and power to reclaim control over their digital present and futures.

• Data for Black Lives was founded with the purpose of “using data science to create concrete and measurable change in the lives of Black people”. Their broader manifesto of “Data as protest. Data as Accountability. Data as collective action.” is both a call and active effort by civil society to wield data as a tool to empower communities of colour in the U.S. The organization’s #NoMoreDataWeapons campaign does this by catalysing policy action and building public awareness around Data Weapons that they aptly describe below:

“Data Weapons aggressively expand state surveillance efforts, exponentially increase the ability of law enforcement and state actors to relentlessly punish and criminalize Black communities, limit the movements of Black people, and effectively transform Black neighborhoods into zones of heightened patrol and policing, mirroring the prisons and detention facilities that hold millions of our people captive everyday” — Data for Black Lives

• Conversations with social enterprises like Abalobi, which empowers fishing communities across countries in Africa, revealed the importance of closely engaging with their beneficiary communities to both enrich data collected and assess data needs through Monthly Data Discussions. This inclusive fora allows individuals from the fishing community to take control of their own data narratives and better understand how it can reflect their needs and interests.
• Organizations like Data Refuge work to surface narratives and use storytelling as a method to ensure their methods of data collection are community-driven, collaborative and reflect on-ground realities often obscured by mainstream data collection mechanisms.
• Citizen-science projects like DECODE or TULIP Smart City bring together individuals to play an active role in collecting data on their neighbourhood and broader environment and helps them understand how they can use its insights to lobby and enact change in their own communities. Encouraging active participation can build trust and awareness around data and its collection which can be critical grounding necessary for engaging in further dialogue on issues like privacy and governance.

STEP 2: Leverage or transform existing community infrastructure

Facilitate participation in data governance and strengthen digital collectivization opportunities.

• Worker Info Exchange is a non-profit that supports trade unions and grassroots initiatives to help workers unlock and gain access to data and powerful insights that are key to enhancing collective bargaining power.
• The Lighthouse guide is an outcome of a collaborative effort by Prospect, Digital Public, Duke Center on Law and Technology, and small scale which helps trade unions assess and explore how they can become better stewards of their data.

What needs to be unblocked?

In addition to our evolving research and analysis, there is already work being done to further support this journey for non-profits and social enterprises. Based on workshops and conversations with various civil society organizations, Stanford PACs has outlined how an organization’s operational values can define technical choices and legal considerations for non-profits.

Due consideration must also be taken of the earlier mentioned challenges around capacity, technical infrastructure/tools and funding.

Recently, the McGovern Foundation has called for philanthropies to prioritize support to non-profits in order that they can develop stewardship capabilities necessary for data empowerment. Their recent merge with the Cloudera Foundation and the new ‘Data and Society program’ also holds promise for CSOs and helps them gain access to grants, ‘technical expertise, short accelerator programs, and multiyear partnerships’.

Similarly, the Wellcom Trust has partnered with Sage BioNetworks to launch the ‘Wellcome Data Re-use Prizes’ to encourage innovation around data re-use. Sage Bionetworks is also engaged in piloting a global mental health databank.

Institutions, philanthropic foundations, and impact ventures like Omidyar Network and Rockefeller Foundation are also key stakeholders in this ecosystem, who have recognized barriers to responsible data governance for CSOs and are exploring funding pathways to promote research, pilot use-cases, and scale efforts.

However, questions remain that we should continue to discuss:

• Do CSOs have the power and mechanisms to hold private and public stakeholders accountable for data decisions?
• What are the responsibilities and mechanisms that might need to be put in place to both empower and safeguard individuals/communities?
• What are the technical capacity requirements and skillsets required to transform this into an operational reality for CSOs?

Empowering Civil Society Organizations to be better Data Guardians was originally published in Aapti Institute on Medium, where people are continuing the conversation by highlighting and responding to this story.