In early April 2026, the Observatory convened an ‘evidence capitalization workshop’ with partners to synthesize insights and messages on advancing early action for dryland resilience and prosperity in the Horn of Africa. This short video records insights on the roles of trust in early action and approaches used to overcome gaps.
Our appreciation to the following colleagues who shared their insights: Abdishakur Diriye, University of Edinburgh; Bilach Jimale, Drylands Learning and Capacity Building Initiative; Brenda Lazarus, Food and Agriculture Organization of the United Nations; Camilla Schynoll, Welthungerhilfe; Claire Bedelian, Conservation International; Jackson Wachira, Centre for Humanitarian Change; Jarso Mokku, Drylands Learning and Capacity Building Initiative; Kelvin Shikuku, International Livestock Research Institute; Marcos Sugastti, Jameel Observatory for Food Security Early Action and University of Edinburgh; Nelly Bosibori Namweya, Mercy Corps; Oliver Wasonga, University of Nairobi; Robert Muiruri, Kenya Meteorological Department; Tahira Mohamed, Jameel Observatory for Food Security Early Action and ILRI.
To keep the video short, we could not include the full interviews so here, since they provide more detail, we provide the full interview transcripts:
Abdishakur Diriye
In my work, I experienced some kind of mistrust towards the insurance, index-based livestock insurance programs by the communities when I was doing my fieldwork in Ethiopia.
The mistrust was actually caused by the lack of feedback mechanism to address the grievances of pastoralists who have been insured in that context.
Although the program may have been working as it was designed, people are using their human judgments to basically talk about the drought that they are in, although the program is using index and waiting the threshold to be met to release payouts.
There should be a mechanism basically to talk to the communities and tell them that the threshold is not being reached.
Bilach Jimale
I personally work with pastoralist communities and mostly, I’m actually a pastoralist woman so sometimes standing in front of a pastoralist woman and talking to her as a pastoralist woman really helps.
I also tend to be looked at as a youth, I am not one, but I tend to pass off as one. But I usually have a soft spot for the women, youth, and also persons with disabilities in our community.
Once you stand in front of them and talk to them as a person who comes from that community, you build trust with them.
When you don’t build trust, any initiative or anything you say or anything that you bring to the community is not received well by the community.
You find that any message that comes from that organization, institution is not received positively because trust has not been there. And most of the time you build trust and any small misinformation breaks the trust. So it’s very important to ensure that. You build trust and ensure the trust between you and the community stays on so that you can be able to build that community and ensure that there’s inclusion in the conversations that is taking place with the community.
Brenda Lazarus
I’ve been doing basically food security early warning work for almost 15 years, and I can’t tell you how many times I’ve heard people say, well, the early warning was wrong. And I think to me, and then that leads to deterioration in trust.
I think to me, an issue is kind of our failure in communicating early warning information, that I think we often say the forecast is for above average rains or the forecast is for below average rains, but we don’t communicate the probabilities or the uncertainty relating to those forecasts, and I think we need to be better at communicating what exactly are those forecast saying, particulaly relating to uncertainities and probabilities.
I think similarly with reaching communities and reaching the last mile with early warning information, I think there’s also challenges in trust there as well, and I think we need to be focusing more on kind of how do we communicate with communities, particularly identifying trusted intermediaries, people, perhaps religious organizations, local leaders who the communities trust, and then disseminate that early warning information through those trusted intermediaries.
Camilla Schynoll
We as WHH, we try to work closely with the local communities and through local partners. And so we’re close with the people and what we often hear or see is when for example, there was a climate warning or any early information and it did not occur, so let’s say they say there should be a flooding and it did not happen, the trust kind of goes down into these formal information systems.
But we see that if you integrate traditional knowledge and their own perceptions and have it as a participatory approach, really trying to involve the communities more in these kind of early warnings that maybe go through the radio, etc. Trust doesn’t erode as quickly if it comes from the community itself.
Claire Bedelian
In our work with Weather and climate information services (WCIS), we found that it’s not only important for the information to be accurate and relevant to people’s livelihoods, although those are really important. It needs to be both trusted information and it needs to become from a trusted information source.
People are more likely to use information if it comes both from a credible, trusted source and the information itself is trusted.
We found that information is most likely to be trusted if it comes from familiar sources. And often these are from people’s social networks, from their family and community networks, such as the faith-based groups or women’s groups.
Having these intermediary layers of information makes the weather and climate service information more trusted. We also found that indigenous knowledge tended to be more trusted than the scientific knowledge, especially in pastoralist communities. If there are ways to prioritizing ways where we can weave and blend the different sources of knowledge, both scientific and indigenous can make the information more trusted by the community.
One important way to overcome distrust is through co-production approaches, both in the generation, delivery and dissemination of weather and climate information service. So this means really integrating pastoralists and other users into the design and delivery of the information.
That is an important approach to increase the trust over the information and where it comes from.
Jackson Wachira
Trust is a huge issue in pastoral drylands. With my work in Northern Kenya, I’ve seen how communities trust or do not trust early warning information, either from the local actors, such as the local seers. But also from the formal early warnings such as those from government agencies.
The lack of trust is mainly the result of inadequate engagement and shared understanding between communities and the agencies or actors that are providing the early warning information.
One way of dealing with that which I think has been quite successful in Northern Kenya includes bringing together the informal or traditional early warning experts together with the formal early warning experts.
In the context of Northern Kenya, places such as Marsabit for instance, we have people called the Uchus who provide early warning information to communities by reading intestines compared to the government-led early warning information providers such as the Kenya Meteorological Department who rely on scientific models to provide early warning and bringing them together through a multi-stakeholder scenario planning has always been very very useful in helping bridge the gap and provide a shared understanding of what different interpretations of indigenous forecasters as well as the more formal forecasters.
Jarso Mokku
The question around trust is very critical with communities. And the situation of, for example, trust and distrust is often because of the experience.
I’ll give you an example. When the meteorological department says it’s going to rain, and that rain fails, then the distrust creeps in.
Sometimes the accuracy of the information is what creates trust or distrust. So the information must be factual and timely so that it generates either trust or distrust.
The second round comes around. And this is very popular with the pastoralists. They have their own system of looking at weather, for example. And they have their own mechanism and system of preparation. So if their system contradicts, the mainstream system contradicts their information system, they start distrusting it.
Because they trust their own, because they have lived that for many years and they have believed it. So if you give some innovative, creative information that is not in the mainstream of community system, you need to be very careful because that’s how trust and distrust is divided.
And if for some chance in some situation the information is contradictory to what they know, what they have lived, then they will distrust it. And if it fails the test, they also will not trust it.
Kelvin Shikuku
One of the innovations I work on is index-based livestock insurance, which monitors pasture conditions and then informs payouts if vegetation conditions are not good.
One of the issues that cause lack of trust is basically, for example, if payouts are delayed, people kind of question whether this is helpful for them and that can also mean that subsequently it becomes difficult for people to want to take up the policy again. Delayed payments can be one cause of loss of trust.
The other one is that it’s a very complex solution in the sense that you have to monitor vegetation remotely and then payouts happen. But sometimes what the naked eye sees as severe drought may not be what the index shows as severe drought.
And that means that sometimes the communities might expect a payout, even though the index shows that the vegetation condition or the drought wasn’t as severe to kind of trigger a payout.
So that kind of misalignment between what the index is showing and what the community thinks can also be an important source of lack of trust.
So, how we have addressed both is ssues is, one is of course supporting the private sector to make sure that the monitoring process, the calculation that needs to be done happens as fast as possible and therefore the issue of delayed payments is addressed.
The other one is just also making sure that customers have a kind of grievance address mechanism so if they wanted to check on what’s happening or they have any concerns they know where to go to.
And in terms of the mismatch between what people believe and what the index shows, the communication part becomes very important. Investment in making sure the capacity of communities to interpret information is improved, so they know there’s a difference between drought and severe drought.
And so generating materials that simplify or present that information in a simplified way is one of the approaches we’ve used to help communities be able to understand what’s happening.
Marcos Sugastti
I have a couple examples that I think about when I think about trust in pastoral regions just from my time working there. One has to do with weather forecasts. So there’s different kind of forecasts that pastoralists use to make decisions about their livelihoods.
Some of them are more reliant on traditional methods, others are more reliant on scientific methods as shared by meteorological agencies.
Oftentimes people find it more natural to distrust forecasts that come from government agencies or NGOs, which in a way I understand because you know this is something that you might not understand how it’s produced and that comes from an external source. Also you might have some experience with it having failed you in the past.
If for instance you were exposed to a forecast that you’ve found pretty incorrect, you might anchor on that experience and from then on not trust forecasts anymore.
But there I think that exposure to repeated forecasts for instance really allows you to build that trust. Also, having someone that you know that is able to communicate those forecasts to you or explain those forecasts to you a little bit better also helps.
I know that the Kenya Met Office is doing some work around having participatory meetings with the communities there to promote their forecasts, but also to have conversations with the communities so that they better reach the intended audience for them.
Another example that I think about when I think about forecasts in pastoral regions has to do with insurance products. So, the index-based livestock insurance is a product for people to protect their cattle. When something goes wrong, they can get a payout out of that.
Now, with any kind of insurance product, we actually see very low take-up rates in different developing country contexts. And this makes sense because it’s a product that might be foreign to people. They might not understand how it works, and they might not have the appropriate resources to do so.
I think that a key way to build trust and to promote take up of these products is to have intermediaries that can promote the products and that these intermediaries are people trusted by the intended beneficiaries in the first place
In some early work that we’ve done trying to promote the AB product in the drylands, we have found some early results that when insurance agents are already familiar with the communities and when they’re more trusted by the communities that there are higher take up and renewal rates for this product. So we think that just changing the way that agents are recruited, trained, brought into the forefront of the product marketing can really make a difference in helping people pick up innovative solution that can be really beneficial for their livelihoods and resilience.
Nelly Bosibori Namweya
For me, trust is one of the most underrated factors that needs to be critically looked at when you’re looking at interventions for the drylands. And from my experience, I’ve experienced this firsthand where, you know, big programs, the success of that is really affected by distrust by the communities.
Some of the things that have really caused this is the social integration of the communities because drylands and pastoralist ecosystems have been in existence for decades and decades of years.
Really integrating traditional knowledge and traditional systems into any program that is being designed is absolutely critical. When this is not integrated in a socially cohesive way, then the community develops some sort of distrust to the program which affects the Implementation of the program. Or the intervention.
Also, governance does play a big role in building trust with communities. For instances where there has been a history of bad governance, then the community also does develop some distrust to the government systems, and therefore, by default, any program or intervention that will be implemented through the government systems would face that trust issue as well.
Ensuring that there is local integration and locally led solutions in all the interventions that are being designed and implemented is also very important. And this can be achieved through ensuring that during the project design or intervention design, that both men, women, and youth from the traditional systems are adequately integrated into the design of the project, and more importantly, the implementation of the project.
All these three things are very interconnected and have to be looked at for one to be able to have a system where the trust of the communities really is embedded in all the interventions that are being rolled out.
Oliver Wasonga
The issue of crisis is quite common when you’re dealing with communities, especially when you’re talking about climate information.
And most cases, the community raises the issue of trust because of their past experiences with the kind of information that they get from, for example the met department, in the sense that, probably they predicted the rains, and they never occured. Such kind of experiences make them have low trust in what they get from outsiders.
One way of dealing with this is to first build their trust and confidence in what they do and what they know first before you try to make them trust what you present to them. And this through getting to appreciate and recognize the indigenous knowledge so that when you introduce your own knowledge, then they are able, already by that time, already you have some rapport with them.
As researchers, we also experience trust issues with the community, especially when you want to get particular information from them. And this all depends on how you approach them.
The approach matters a lot. In order to gain their trust, you have to use approaches that help to build a rapport with them. And mostly we use participatory approaches to do this. Where they actively participate in the process of co-ideation and then jointly co-producing knowledge and co-designing solutions together.
Robert Muiruri
I’m a weather forecaster, but at the same capacity I’m also an advocate of the indigenous knowledge and traditional methods.
Trust is a very important term in terms of disseminating our weather information to those people who use indigenous knowledge.
To bridge and to work on the trust, we have tried to do participatory scenario planning whereby we do co-production, whereby we blend the indigenous knowledge with the scientific way.
Doing so, we do come up with a co-produced forecast. which streamlines the friction which is previously there, therefore boosting the confidence and the trust between us and those people who consume our information.
Tahira Mohamed
For me, the issue of trust is two ways. Number one, it’s about the data, the data that the scientific community is providing to the community. And the second level is also from the community, the indigenous knowledge and also the way they predict the future.
These two levels of information, they have to agree at a central point on whose information can be taken. For example, a scientific institution, if it’s informed by indigenous knowledge about a certain event, let’s say doubt is coming, how flexible is this institution to accept and change and use this information?
For the second level, for the community, trust issue is attached to just not providing the information, but then what to do with information.
So if I’m told there is flood coming on, and then I’m not told where to go, how to go, and what resources to use, then that brings about mistrust.
So for me, the point is… Attaching the information that we are providing to the community with the activity and resources that help communities move, that can build the trust because this person has not just informed me about disaster, but they’ve informed me where to go and how to go and probably facilitated me to move to a safer place.