Yes, Transformative Impact Through Digital Agri Information Services is Possible.
Digital agricultural information service has rarely delivered a transformational impact for farmers. But Gramhal advocates that it is plausible. We make this rather forward statement based on our experience in the field and study of existing literature. First, some context.
Missing Transformative Impact
Agriculture is the source of food for 7.5 billion people on Earth and the primary source of livelihood for 2 billion people. Hence, the objective of all agricultural transformation efforts rests somewhere on the continuum of increasing productivity per acre to feed the growing population and ensuring better income gains to keep farmers out of poverty.
The potential of digital technologies to deliver agricultural information services was unlocked to reach an exponentially large number of farmers. The government utilises a basket of ICT tools such as TV channels, community radios, SMS, call centers, mobile apps, etc. Increasingly, the private sector has also made in-roads in providing digital information services, especially in the agri-input space.
Over time, multiple studies have been undertaken to understand the effects of digital technologies on farm outcomes. One study , looked at the impact of providing farmers with market and weather information on their mobile phones. They found no statistically significant effect on the price received by farmers, crop value-added, crop losses resulting from rainstorms, or the likelihood of changing crop varieties and cultivation practices. Another experiment with Indian potato farmers, found that improving farmers' access to price information alone is unlikely to have positive outcomes on farmgate prices. In a meta-analysis study of existing RCT studies about the effects of digital technologies, it was found that agricultural information through mobile technologies in sub-Saharan Africa and India increased yields by only 4%. Further, the odds of adoption of recommended agrochemical inputs was only 22%. Yet another study in Uganda, showed that the effects of introducing a mobile phone-based marketplace for agricultural commodities were meaningful but not statistically significant. And, the benefits were greater for larger-scale versus smaller-scale farmers.
This shows that the existing research around digital agri information services has not been able to conclusively establish transformative impacts on an individual level. However, it is argued, given that there are hundreds of millions of farmers and the cost per farmer of digital agri information service is low, the limited impact has an appealing benefit-cost ratio. At scale, even the small income gains per household will make the intervention enormously cost beneficial from an overall welfare perspective.
Gramhal acknowledges these findings. However, our team believes that digital agri information services can have a transformative impact based on our understanding of the 3 emerging sub-themes.
1. Context and Countermoves determine the translation of information into impact
Over the past three years, our learning has been that – give information to a pool of farmers, and you would get an impact on the continuum of negative and transformative positive. This happens
because every farmer has their own context and faces different countermoves by influential stakeholders in their ecosystem. Following is an experiment by Gramhal that proves this learning.
In November 2020, we provided door-step quality inspection service and personalized price information to 300 maize farmers from 20 villages in Chhindwara district, Madhya Pradesh.
We found that at the individual farmer level, socio-political and interpersonal factors determined how the information provided translated into price gain. Factors at the intersection of gender, landholding, literacy level, caste, relationship with buyers, and ability to interpret information and use it in negotiation, acted as enablers for some and barriers for others.
On the other end, the buyers (village-level aggregators/VLAs) also responded to this new reality with different countermoves. Some VLAs started creating a narrative that the price information given by Gramhal was incorrect. Others, became hostile to farmers and threatened to boycott the household for any future purchase. However, we also had farmers who were able to counter the moves of village-level aggregators and earned 30% higher profits after negotiation. On average, our intervention had only small positive gains.
Our team's learning has been that the impact observed in the existing literature is small because extension interventions rarely acknowledge that socio-political constraints bound farmers. Ownership of land and means of production, gender relations, caste dynamics, and powerplay during farmgate interactions between smallholder farmers and buyers determines how well an agricultural information service is received. Every external intervention starts a game between farmers and other stakeholders, each countering the moves of the other. Solutions that consider farmers as simple agents and predictable maximizers in a static context are most likely to produce no or small impact.
In the last two decades, the majority of the digital agricultural information interventions took this marginalization a step further as smartphone ownership, connectivity, and digital literacy become additional means of production.
2. Availability of low-cost two-way communication technology
Another reason for the poor impact of existing digital tools is their dependence on one-way technology to serve farmers. Mobile phones are used like radios where information and knowledge are being pushed using SMS to farmers rather than hearing from the farmers and providing customized information based on their needs. In the last five years, we have seen massive technology growth that allows transmitting two-way multi-media content at a lower cost. This creates potential for a greater impact.
3. Bundling of agri-information services is necessary for enlarging the size of the pie for the farmer
The current research body has mainly studied interventions providing singular solutions – like crop advisory, weather, price information, or market connection. Farmer requirements keep changing throughout the crop cycle, and intervention around a single service has a higher probability of dropping the impact ball in the middle. A bundled service delivery can collect data from farmers at different times and keep improving its personalization and quality of service. Different services will create their small impacts, and together the impact will be larger than the sum of its parts.
In conclusion, taking a design thinking approach is the need of the (transformative impact) hour. Innovation for digital agri-information tools should rest on three foundational design themes:
Taking an agile product development approach, where the solution keeps evolving as other stakeholders make countermoves.
Creating a rapid feedback mechanism from farmers by creating a digital community with two-way communication
Bundling of agri-information services that increases the size of the pie for every farmer.
Bergquist Falcao, McIntosh, Craig (2021): Search Cost, Intermediation, and Trade: Experimental Evidence from Ugandan Agricultural Markets. CEGA Working Paper Series No. WPS-173. Center for Effective Global Action. University of California, Berkeley. Text. https://doi.org/10.26085/C3759K
Fabregas R, Kremer M, Schilbach F (2019). “Realizing the potential of digital development: The case of agricultural advice,” Science Dec 13; 366(6471): eaay3038. doi: 10.1126/science.aay3038. PMID: 31831641.
Fafchamps Marcel, Minten Bart (2012). “Impact of SMS-Based Agricultural Information on Indian Farmers,” Oxford University Press on behalf of the World Bank. © World Bank. https://openknowledge.worldbank.org/handle/10986/19077 License: CC BY-NC-ND 3.0 IGO.”
Mitra Sandip, Mookherjee Dilip, Torero Maximo, Visaria Sujata (2018). “Asymmetric Information and Middleman Margins: An Experiment with Indian Potato Farmers,” The Review of Economics and Statistics 2018; 100 (1): 1–13. doi: https://doi.org/10.1162/REST_a_00699