Behavioral Science and the Future of Agriculture

Precision agriculture uses remote sensors to analyze the needs of individuals crops. Because each crop and field is different, the technology allows for precise information to maximize yield per crop. With this technology, farmers can be selective in their resource use and only deploy resources like water and fertilizer in locations where they’re needed.³ Precision technology also improves dairy and livestock farming, with sensor systems that measure animal behavior and health, helping farmers’ decision-making.⁶

Overall, precision agriculture is an incredibly helpful tool for decision-making optimization. This technology has several key benefits: reducing resource costs for farmers, reducing the risk of over-fertilization, and improving sustainability.^6,7

Technological advances allow for autonomous vehicles and equipment to be employed in fields 24/7, increasing productivity, reducing food waste, and protecting farmers’ safety. One example of AI technology in the agriculture industry is facial recognition software for cows, which provides farmers with the ability to track individual animals’ health in detailed ways. Other uses of AI include using data to decide where to apply which type of herbicide, and predicting upcoming weather patterns to help farmers in their decision-making.¹⁸

However, the uptake of these technologies is low.6 Researchers found that in Germany, only 10–30% of farmers use new tools such as these.¹⁹

A regularly discussed setback is the cost of new technologies. For some farmers, the technology’s upfront costs are too high, given the potential risk of crop failure due to poor weather. The variability of crop prices year by year can also make farmers more careful with their capital investment.^1,9

Additionally, many emerging technologies involve the use and integration of data across different products. Evan Fraser, a professor at the University of Guelph who researches farmer behavior, states that data interoperability, data governance, and cybersecurity are the most significant challenges for farmers adopting the technology. Ransomware hacks are threatening in any context, but could be devastating for farmers; Dr. Fraser provides the potential scenario of cyber hackers remotely taking control of a poultry farm ventilator, which would have drastic consequences for the farm.¹⁰

Both cost- and data-related factors are contributors to the slow adoption of agtech, with due cause. These challenges considerably impact adoption and will need to be solved through technological improvements to decrease the cost while improving security.

But behavioral science can also help speed up agtech’s adoption, and policy makers should start to consider its use. Luckily, research is increasing in this sector, showing effective behavioral interventions that improve technology adoption.

Factors like age, gender, experience, attitudes, and beliefs explain a lot about a farmer’s inclination to use new technology. In assessing precision agriculture technology adoption, researchers from HEC Montreal and the University of Buckingham found that factors including ease of use, knowledge of technology, and perceived usefulness all significantly explained the variance in farmers’ adoption of the technology.¹¹ These results show that one should consider these individualized qualities in order to create effective communications about threats, and the recommended responses to them.¹²

Cognitive biases also get in the way of technology adoption in farming, including heuristics and cognitive dissonance.

One research group found that heuristics had a significant part to play in farming behavior. Heuristics, or rules of thumb, are cognitive shortcuts used to process information efficiently. For example, in Kenya, farmers use visual cues to identify high-quality dairy cows. By recognizing these rules, most of the time, farmers can make accurate decisions.

Yet, in other scenarios, these shortcuts prevent optimal farming decisions. For instance, farmers in Mozambique traditionally aim to plant their cotton by mid-December each year. However, the time of year does not actually dictate the optimal time to plant cotton. A better indicator is the saturation of the soil after rainfall. As Mozambique, like other parts of the world, is seeing increasingly unpredictable patterns of rainfall, the farmers’ “rule” was at risk of inaccuracy. The researchers helped farmers switch to a rain-based rule of planting cotton after the first large rainfall, consequently increasing their crop yield.¹³

Cognitive dissonance also contributes to a reluctance to accept new information or techniques. Cognitive dissonance is how our brains respond when presented with competing information to our original beliefs. While cognitive dissonance can result in positive behavior change, our minds often rationalize continuing with irrational behavior that aligns with our existing beliefs.

Take, for example, a Queensland peach grower, who believes to have accurate knowledge of a well-known pest, the fruit fly. In this case, the farmer believes that the pest is not a credible threat, and they know how the pest arrived and its movement around the farm. When the farmer’s neighbor begins installing fruit fly traps to reduce the risk of damage to their crops, the farmer faces a choice: ignore the neighbors and proceed with their existing beliefs, or adjust their beliefs and install traps.

Psychologists explain that individuals will actively avoid situations that conflict with their beliefs due to the cognitive strain caused by changing previously-held thoughts. It is much easier to do what you’ve always done and view the world how you’ve always seen it.¹⁴

Because personal beliefs and perceptions have such a great impact on technology adoption in agriculture, experts have recommended simply educating farmers about various risks, and how technology can be used to minimize them.

In central Tanzania, Seetha et al. (2017) used focus groups and farmer learning sessions to encourage farmers to seriously consider the problem of aflatoxin infection. Two years after educating farmers on the infection, farmer understanding of the negative effects increased from 19% to 82%, and the frequency of contamination with the disease decreased from 44% to 5.9%. Similarly, in Wisconsin, farmers who participated in an educational workshop on nutrient management plans changed their behavior significantly, using more nutrients and conducting soil testing more frequently.¹²

Evidently, education for farmers can significantly change behavior, but more important is how this education is delivered. Experts recommend considering personal factors, like age, in developing educational materials, focusing on demonstrating the technology’s value and ease of use, and ensuring farmers’ concerns are heard in educational sessions.

Many experts have recently called for a more holistic approach in future interventions that accounts for farmers’ social spheres. Social networks are incredibly important in understanding and changing behavior. One study found that one of the most significant influences on a farmer’s utilization of a tool was their community: a farmer was more likely to try an innovation if a friend recommended it. The same situation was true even among agricultural experts, with one study finding that agronomists used a particular technological tool based on their colleagues’ recommendation.¹⁶

A different research project invited farmers to an event hosting other local farmers who presented humorous plays on farming safety topics. After a week, the researchers found that 67% were considering making safety changes, and 42% actually did.¹² This study shows the impact that social norms and networks have on changing behavior. When trying to inspire beneficial behavior change, understanding an individual’s network and how they receive information is incredibly impactful and should be considered in all interventions moving forward.

While the current behavioral literature has produced encouraging results on low-cost behavioral interventions, Rose et al. stated that several gaps exist in the research today. Going forward, we need:

• More robust, long-term studies of farming behavioral change: There is still a lack of high-quality research specific to the agriculture industry, including the use of control populations in studies.
• More work to understand personal traits and impacts on behavior: Rose and colleagues recommend better understanding how cognitive and emotional factors affect farming behavior.
• Lack of knowledge on how targeted policy tools may work: While knowledge about individual behavior change is increasing, little knowledge exists about which policy tools are more likely to get results in different contexts.¹²

Overall, the world’s changing agricultural needs require new practices and the use of new technologies to keep up. While these technologies are increasingly developed, several factors influence farmers’ hesitancy to adopt them. However, behavioral insights and nudges are useful to help understand factors beyond cost and data security that influence adoption of agtech products. Research shows that personal factors, cognitive heuristics, education, and social norms shape how farmers perceive new technologies, and should inform low-cost interventions to encourage further adoption. While there is still a need for more research on proper policy and the use of nudges, what exists in the research today shows a promising future for improving farming practices around the world.