Systems change in practice: learning from the Start Network’s Drought Financing Facility

Written by Ian McClelland, Start Network

The Start Network’s Drought Financing Facility project sought to revolutionise the way the humanitarian system responds to major droughts. Outreach and Engagement Adviser Ian McClelland from Start Network found out about the challenges they’ve faced in trying to change the humanitarian system, and what they’ve learned in the process.  

To date, much of humanitarian innovation has focused on product or service innovations that solve specific, well-defined problems. But the big, high impact challenges are often far more complex, requiring changes to the supporting “ecosystem” for a given innovation. A great idea might fail to find its footing because of barriers to uptake such as procurement processes, government regulations or cultural norms – what the RSA has termed the “system immune response.”

Addressing these systemic challenges is a core concern of the HIF’s new strategy – and has been key to the Start Network’s plans to establish an innovative drought financing mechanism, with funding from the HIF and in partnership with GlobalAgRisk (now known as GlobalParametrics). The Start Network’s Drought Financing Facility (DFF) project has sought nothing less than to revolutionise the way the humanitarian system responds to major droughts – shifting it away from a reactionary model based on voluntary contributions, and towards an anticipatory model that enables early, pre-planned and pre-financed responses.

Embracing this highly-complex challenge meant that not only did the Start Network have to overcome the technical barriers to designing a viable financing model, but they also had to work towards changing the systems which the DFF depends on for its success, navigating changes to ways of working in NGOs, and advocating for changes in how government donors fund humanitarian efforts. Ian McClelland spoke to Emily Montier, who leads the DFF project, about the challenges they’ve faced in trying to change the humanitarian system, and what they’ve learned in the process.

What is the problem you are seeking to address?

The initial problem statement was: We know that we have the science to predict droughts and develop an early warning system, and yet when early warnings happen, the response is still always late. We know that it’s more cost effective to respond early, but the current system means that too often we have to wait until we see suffering before money is allocated. There’s this long period of procrastination and missed opportunities. The Drought Financing Facility was set up as a way to tackle these issues.

How did you set out to address this problem?

We were very interested in the potential for disaster risk financing and insurance, and we had been talking to contacts in the insurance industry for several years. While the Start Fund’s structure as a global contingency fund is well set up for smaller under-the-radar events, it isn’t suitable for large-scale drought. We wanted to better enable early responses to all instances of drought by modelling risk in a given country context, establishing triggers for early pay-outs, and supporting this with a contingency fund for small-medium events and an insurance product for much larger events.

Our idea was that a donor would pay to cover the risk for, say, 50,000 vulnerable people in Pakistan, and so that donor’s money goes into a fund in Pakistan, managed by in-country NGOs. Some of that money will sit in a contingency fund, and some of the money will be used to buy drought insurance from a commercial insurance provider. There’s no point in buying insurance for something that happens every three years as it would be prohibitively expensive, so the contingency fund covers small to medium size events. But with competing funding priorities it’s difficult to justify sitting on the amount of money necessary to respond to an event that happens once every 10 years. In these instances, insurance works as a way of smoothing payments over time, by paying smaller annual premiums.

Insurance is often seen as the shiny new innovative thing, but to me, what’s interesting is not insurance, it’s the risk financing approach. It’s the idea of not waiting for something to happen, but understanding risks, modelling them in advance, knowing what you need to do, and then positioning the funding to do it. It’s about getting the money where it needs to be in the right quantity and at the right time, and the way in which we do that is through a blended approach involving both the contingency fund and insurance. The contingency fund is triggered off the same model used to trigger the insurance, so the idea is that the it is the same model that underpins both.

To take this idea from theory to practice, there were three main pieces of the puzzle that we needed to put in place: (1) the science to model the risks of drought and to set triggers for early response payouts; (2) the tools for NGOs to plan what they would do in different drought scenarios and how much it would cost to protect communities; and (3) the financing to support this new kind of funding mechanism.

Let’s look at the challenges involved in addressing each piece of the puzzle. What were the challenges that you faced in developing the science of drought prediction and payment?

The main challenge we encountered was that there’s no one definition of what a drought is; there’s no one data set that tells you, these are the droughts that happened in this country, and this is why they happened. There are lots of political and humanitarian dimensions that influence drought and the potential for food crises. So how do we validate what we create in terms of scientific modelling for drought risks, and how do we ensure that we’re using the best science possible? For the DFF, we worked very closely with GlobalAgRisk, a for-profit social venture specialising in building disaster risk financing models. These questions have, however, spurred us to launch a research project in which we are working closely with academics and scientists. Its objective is to create a framework to assess the effectiveness of the scientific models we use in parametric insurance schemes.

A second challenge was how to bring human vulnerability into the model. Currently, the model is primarily climate-risk driven, but if you have a drought that interacts with conflict or a crop pest, then it’s going to be much worse than a drought on its own. For the insurance product, that creates a situation called a ‘basis risk’, which is when the model that tells you the insured losses on the ground are different to the actual losses, meaning that you could get an over-payment or under-payment compared to the funds needed for early response. Basis risk is a problem of all ‘parametric’ or ‘index’ insurance products – where insurance is based on automatic triggers rather than an independent appraisal of losses after the event. To address this in the DFF, when the insurance payout is insufficient, money from the contingency fund can be used as a top up, and when the insurance payout is in excess of the damage, this can be reinvested in the contingency fund.

What were challenges for contingency planning with NGOs?

We had the assumption that if you give people money six months earlier to reduce the impact of a drought, they would know what to do with it. But when we spoke to NGOs at the country level it caused everyone to stop and pause, and say, actually we’ve never had the money that early, so we don’t really know what we should do. People say that the importance of early action is that it’s not the same action, just earlier, but instead you’re enabling different actions to reduce potential impact.

We quickly learned that when you’re pushing people to work in new ways you need to accompany it with the right tools and support to enable that understanding about how to respond differently. Within the contingency planning tools supporting the DFF, we’re therefore encouraging NGOs to think specifically about what objectives they are trying to achieve as part of their early actions. The actions that are being recommended, however, need to be tailored very specifically to the context. Through the roll-out of these contingency plans, we’ll be able to build learning and knowledge of which activities are the most successful.

What were the challenges of financing?

This is a real challenge. We’re designing new financial instruments, so we’re challenging donors to give their money in different ways. We can’t just wait for a call for proposals and then apply, because this doesn’t fit in the box. Although there are more and more donors who are interested in risk financing, this interest hasn’t been translated into clear thinking about how donors give money.

There’s always a slight fear of proactive funding among donors, but these same donors fund very costly drought responses once situations worsen. We can put the efficiency arguments in front of them, with cost-benefit analyses, but it requires a large shift in mentality and willingness to change ways of working. It’s not impossible, because there are donors who are starting to look at things this way, but it takes a lot of work.

The Start Network is well placed to do systems change innovation because we can pool resources across our 42 member NGOs to try new things and develop new partnerships. But as an aggregate organisation, our interest is in scale, and in order to understand whether a new idea can bring about system change, a large-scale pilot is needed for validation. While you might be able to find a donor willing to pilot something in a small area of one country, we’re asking for a lot more in terms of risk appetite. This is an area we’re activity working on as we speak, as we are fundraising for the DFF roll out.

Where are you now with the project, and how are you taking it forward?

The recent funding from the HIF enabled us to refine and improve the original design concept for the Drought Financing Facility, and develop the drought risk index and operational plans in two country locations where there was appetite among a number of NGOs – Pakistan and Zimbabwe. That experience has now been compiled into a design report and a supporting 60-page document that brings together all of our learning from developing the necessary tools and approaches in those two countries. We are currently implementing a lot of the learning from this project into a separate drought risk financing programme called the ARC Replica Initiative, and we’re fundraising for an active pilot of the Drought Financing Facility.

Read more about the Start Network project here