Exploring Behavioural Insights At AAFC
At Agriculture and Agri-Food Canada (AAFC), we like to explore options. For instance, how could we apply behavioural insights to public trust policy? In 2019, we saw that the Treasury Board of Canada Secretariat’s Experimentation Works program could help us with this. This was a chance to develop departmental capacity in experimentation. Experimentation could be a way for us to better understand consumers. We could add experimentation to our existing Public Opinion Research (POR) work.
AAFC’s AgriAssurance Program
We want a competitive, innovative and sustainable Canadian agriculture and agri-food sector. This sector includes everything between the farmer to the consumer. It starts at the farm and reaches to global markets. All phases of producing, processing and marketing of farm, food and bio-based products are part of agriculture and agri-food. The department has a program, AgriAssurance, that supports industry in developing assurance systems. These systems aim to make how the sector produces foods more transparent. They do this by applying certifications based on standards and verification. To the consumer, these kinds of assurances usually show up as some sort of logo or claim on a food product (e.g. organic, fair trade, etc.).
When assurance systems are working well, everyone who uses assurances is aware of them, understands them, and can easily get the information they want from them. Strong assurance systems help promote transparency and accountability about methods of production. This makes agricultural processes more clear to consumers. And it helps producers’ credibility.
Citizens can vote with their wallets for the production methods they most want. Assurances welcome the consumer into the food system, encouraging this stronger relationship between buyers and sellers. For example, you can choose to buy organic-certified meat if you have strong preferences about how your meat comes to your table. The Canadian Organic Standards require that the producer meet certain animal care expectations. This is in addition to adhering to principles of ecology, health, and fairness. So, when you buy organic, that is, verifiable and credible organic claims (e.g. Canada Organic, USDA Organic, Bio Quebec or British Columbia Certified Organic — the list goes on), you know that the way the food was made meets the standard. You are assured.
Robust assurance systems are democratic, helping consumers identify how to buy in accordance with their values. And, these systems can help make the market economy more responsive. When market economies respond to the demand that consumers ask of them, they are working well.
Our Challenge
At AAFC, we know that industry is interested in our AgriAssurance program because we keep getting applications. We can work with applicants, industry stakeholders and provincial and territorial counterparts to improve the process and the program as needs evolve. What we don’t understand quite so well is how well the program meets consumers’ needs.
Academic literature and POR data can tell us some things about the effectiveness of assurances. This is often based on a concept called “willingness-to-pay”. This measure is a way to order products or their specific attributes by studying which attributes participants value or prefer the most. It allows a standardized format, under specific conditions. But often, willingness-to-pay studies do not account for time pressure, monetary penalties, or other constraints that exist when shopping for food in the real world. So, these studies may not answer public trust policy questions, such as understanding whether or not, and how much, Canadians have confidence that the food they are buying is being produced in the ways they want it to be (e.g., cage-free eggs, sustainably caught, Halal, etc.).
However, experimentation is challenging. Even just the idea of exploring which research methods *might* work is challenging. We need to practice experimentation, including exploring what options could be considered before deciding to go down the experimentation path. We need to do this not just to learn, but to become comfortable — with the process as well as with the outcomes.
How Experimentation Could Be Applied
And here is where experimentation might be useful. We can explore and try out some new ways to get information from people about assurances. But the research landscape is complex — attitudes can be explicit (i.e., self-reported) or implicit (i.e., laying below the level of conscious awareness). Behaviour can be observed or self-reported. This makes it difficult to land on which methods would be best, in trying to answer why someone does or doesn’t trust a product or a logo, and whether that influences their decision to buy it or not. Especially if they don’t even know why, they just know that they do. Or don’t. Trust is a concept that is inherently difficult to measure.
If we want to understand behaviour, then we need to know more than what someone does — we need to know why they do it. And for that, we need to observe. And ask. And adjust our questions. And dialogue. An experimentation approach helps us to explore what kinds of research methods we should try before investing in larger-scale research.
We’re looking at developing a multi-method approach that would help distill all of this confusing landscape down into some useful results. This would bring together different research methods that can each individually give us a piece of the puzzle. One method would hopefully get us some “aha” moments from candid reactions about assurance logos on food products. A second method may provide some ideas of whether one assurance feels more credible to someone than another. We’re thinking here about discrete choice models where participants pick from various pairs of products. We’re really interested in identifying if offering some information on one assurance makes any difference in awareness of unrelated assurances.
None of these methods by themselves will give us enough useful information to make any recommendations to decision-makers. But together, especially if they are combined with POR methods such as surveys and focus groups, they could be powerful and insightful.
For example, by asking participants about assurances, we could see whether they notice them at all. Or, if they noticed them but don’t know what they mean. Or, if they know what they mean but don’t have any faith in them. Then we’d have some data we could work with. We could even examine whether someone is able to notice differences between the credibility of assurances. Say, a logo that is more rigorous (backed by standards and independently verified), compared to a logo that is less independent, or is held to a lower standard, or is compared to a claim that is unsubstantiated. What if it turns out that, actually, this is one of the things that Canadians are most confused about and really want more clarity on? We’d then have a solid basis for finding ways to address that.
What We Could Use Results For
Insights like these could be beneficial for industry, consumers, and the entire Canadian food system. We could improve our assurance systems programming. We could look into raising awareness and answering questions about assurances. We could apply these methods to other topics, other programs, other challenges that we haven’t even thought of yet. Surely there are other things we could start doing in the food system, or stop doing, or do differently, that these methods could help identify.
***
Post by Rae Payette, senior policy analyst in the Public Trust and Assurance Systems Policy team in AAFC’s Strategic Policy Branch. AAFC’s EW2 project team included Rae Payette (team lead), Inge Vander Horst (senior policy analyst), Tim Rennie (policy analyst), Laura Stortz (economist), Carol Essenburg (senior communications advisor), and others.
Article également disponible en français ici : Explorer l’application de l’introspection comportementale au sein du ministère de l’Agriculture et de l’Agroalimentaire (AAC) | par L’expérimentation à l’œuvre | Juin, 2021 | Medium
