In June 2026, Astrid Mastenbroek successfully defended her PhD thesis, Smallholder seed choices under risk and uncertainty, at Wageningen University. Her research began with a question familiar to many working in seed sector development: why do smallholder farmers continue using familiar seed varieties, even when improved varieties with higher yield potential are available?
The answer, Astrid argues, goes well beyond affordability and access. Drawing on her experience working in the seed sector and research with maize farmers in Uganda, she explored how trust, habits, household responsibilities and perceptions of risk shape farmers' decisions. We spoke with Astrid about her research journey, the findings that surprised her and what they could mean for SeedNL and its partners.
“Information is necessary, but often not enough”
In a nutshell, what is your PhD about?
My PhD started with a question I kept encountering in seed sector development: why are smallholder farmers not buying improved seed varieties, even when they know about them and their potential benefits? We often hear that improved seed is too expensive, or that farmers lack information or access. These are all important factors, but I wanted to look beyond them.
What interested me was how farmers actually make their seed choices. We tend to assume that if a farmer knows a variety produces higher yields, can access it and can afford it, purchasing it is the logical next step. But farmers do not necessarily make their decisions that way. Their choices are also influenced by previous experiences, habits, trust in suppliers, cultural expectations and the opinions of people around them.
I wanted to understand what happens before a farmer even considers purchasing seed. What catches their attention? What makes a product credible or worthwhile? And what might prevent them from considering a different variety in the first place?
For this, I used insights from behavioural science, including a model called COM-B. It looks at three things that shape behaviour: capability, opportunity and motivation. In simpler terms, whether someone is able to do something, has the opportunity to do it and actually wants to do it. That last part also includes less conscious influences, such as habits and emotions.
Although my research focused on maize farmers in Uganda partly due to the research grant, I believe these questions are relevant to seed sector development much more broadly.
What made you want to research this topic, and why is it important?
My interest came directly from working in the seed sector. I had spent years working with local seed businesses and farmers, and I became increasingly interested in how both made their decisions.
I remember discussions with local seed businesses about how to price their seed. Sometimes their starting point was simply what they had paid for the foundation seed, rather than what farmers might actually be willing to pay. We were also working on things like Quality Declared Seed, which was intended to provide farmers with a more affordable option. These discussions raised my interest in whether making seed cheaper would necessarily make farmers want to buy it.
I also noticed that we, as researchers and practitioners, sometimes approach seed adoption quite linearly. We develop or introduce a variety with better characteristics, demonstrate that it performs well, and then wonder why farmers are not buying it. There is often a silent assumption that the benefits should speak for themselves.
Yet, smallholder farmers' lives are much more intricate than focusing just on a single crop or variety. Farming decisions are closely connected to household needs, social relationships and other responsibilities. What may look like an obvious investment to someone working in seed sector development may not be so obvious from the farmer's perspective.
Over time, my research question shifted. Instead of asking why farmers were not adopting improved varieties, I became more interested in how those varieties enter farmers' decision-making in the first place. What makes them credible, attractive and worth considering?
What was the process of doing and eventually writing the PhD like?
It was a roller coaster, with plenty of ups and downs. I had a broad idea of what I wanted to investigate, but working out how to translate that into rigorous academic research was quite a journey.
I tried to involve researchers from different disciplines and worked closely with Ugandan colleagues and other research institutions. That was important because everyone brought different expertise and ways of looking at the questions. Some of my research was qualitative, spending time in communities and really trying to understand people's experiences and perspectives. Other parts involved economic experiments to investigate how farmers valued seed and made decisions under different circumstances.
One challenge was learning to approach questions as a researcher rather than a practitioner. As a practitioner, you are often looking for solutions and ways to improve things. In research, you have to be much more precise about what you can actually conclude from the evidence. My supervisors kept pushing me to narrow the questions and be more specific.
Some papers came together relatively quickly, while others took years. But towards the end, things started to fall into place. I began to see more clearly how the different findings connected, and how they could contribute to our understanding of farmer decision-making. The final year was particularly exciting because I could see the potential value of the research for practitioners, not just for academia.
What are the two or three findings you think are most important?
One important finding is that providing information does not necessarily change farmers' willingness to pay for seed.
In one experiment, we tested whether explaining seed certification and quality standards would make farmers willing to pay more for certified maize seed. The information improved their knowledge of certification, but it did not increase their willingness to pay.
That is interesting because providing more information is often one of our first responses as practitioners when adoption is low. We organise demonstrations, train farmers or work with model farmers to explain the benefits of improved varieties. Those interventions can certainly be useful, but we should not automatically assume that information is the main constraint.
For example, farmers might understand the benefits of certified seed perfectly well, but still not trust the supplier or believe that the label guarantees what is inside the bag. Giving them more information about certification may not address that underlying concern.
Another finding concerns the importance of what farmers already have and know.
In one experiment, we gave farmers a pack of hybrid maize seed and explored how much money they would want to receive to give it up. We compared this with how much farmers were willing to pay to acquire the same seed. Interestingly, farmers valued the seed much more highly once they already owned it.
This is something economists call the endowment effect. In everyday life, we often place more value on something we already have than on the same thing when we do not own it.
It made me think about farmers' continued use of home-saved seed. Farmers may have a variety they know and trust, and they may not feel a strong reason to change. We all behave like that in other parts of our lives. Think about buying the same breakfast cereal every week. You’ve liked it for years, and you don’t spend hours comparing all different types of cereal in the supermarket every time you go grocery shopping.
I suspect something similar happens with seed, although I could not directly prove that these unconscious habits explain farmers' continued reliance on familiar varieties, but I think it is an important area for further research. The difficulty with studying unconscious behaviour is that as soon as you ask someone why they do something, you are already making them think consciously about it.
A third important finding is that we need to look at agricultural risks in the context of the entire household.
We often treat farming decisions as separate economic decisions. A farmer invests in seed, expects a certain yield and weighs that against the risks. But unlike with European farming businesses, for many smallholder households, money for farming is not separate from money for everything else. The same household budget needs to cover seed, food, school fees, healthcare and other expenses.
There are also many risks beyond crop failure. The planting season can coincide with periods when food and money are scarce. Illness can affect the availability of household labour, while unexpected healthcare costs can compete with agricultural investments.
We explored this through an experiment in which farmers could invest in hybrid maize seed under different risks. When we introduced the possibility of a health shock, women reduced their investment in hybrid seed, while men's decisions remained largely unchanged.
We cannot say exactly why this happened without further research, but it suggests that health risks may carry different weight for women because of their responsibilities within the household. It also shows why we need to be careful about treating farmers as a uniform group or assuming that agricultural risk is the only risk influencing agricultural investments.
What do your findings mean in practice for SeedNL and its partners?
For me, the main implication is that we should start with understanding farmers' behaviour before designing interventions. When demand for improved seed is low, we often move quite quickly towards solutions. Perhaps we need more demonstrations, better extension services or more affordable seed. But before choosing an intervention, we should ask what is actually preventing farmers from considering or purchasing that seed.
That is why I am particularly happy with my paper on the COM-B model and how it can be used to understand and stimulate seed demand. I think it is an accessible way for practitioners to look more critically at the assumptions behind their interventions. It brings together insights from behavioural science and seed sector development, and shows how they can help identify different barriers to seed uptake.
This also has implications for seed companies. Traditionally, we have often approached product development and marketing from the supply side. We develop a variety with certain characteristics and then consider how to sell it. But we should spend more time looking at the process from the farmer's perspective. How do farmers respond to the packaging? What catches their attention? Which brands or suppliers do they trust? What do they think when they see a certification label, and how does that compare with advice from a neighbour or a familiar agro-dealer?
Many seed companies have already made considerable progress in understanding their customers, but I think there is still much to gain. It may be worth spending more time testing packaging, branding, communication and distribution approaches with farmers, and then adapting these based on their responses.
Trust is particularly important. Formal certification is intended to give farmers confidence in seed quality, but it only works as a signal if farmers believe in it. The credibility of seed suppliers and agro-dealers matters as well. The same applies to development programmes.
I think my research helps bring some of these considerations together and makes them more specific to seed. Much of it is not entirely new, but we could be more systematic about incorporating it into how we design seed sector interventions.
What message would you like SeedNL stakeholders to take away from your research?
I would encourage people not to interpret low adoption of improved seed too quickly as a lack of farmer demand. A farmer can know about a good variety, understand its benefits and have access to it, yet never seriously consider buying it. Perhaps they already have seed at home. Perhaps they do not trust the agro-dealer or the certification label. Perhaps they have other household priorities, or social norms influence who is able to make decisions about purchasing seed.
We sometimes focus very strongly on developing and introducing better varieties. That is important, of course, but in some situations there are already plenty of varieties available and farmers are still not switching to them. We need to spend more time understanding how farmers actually make their decisions, and design our seed products, services and programmes around that reality. Not just asking whether the seed is available and affordable, but whether farmers recognise it as a credible and worthwhile option.
I hope the research encourages people working in seed sector development to look at farmer demand a little differently. Sometimes the answer is not another variety, another demonstration or more information. Sometimes we first need to understand why the options already available are not being considered.
Read Astrid's full PhD thesis: Smallholder seed choices under risk and uncertainty
