Behavioural science and rigour at the last mile of aid

A mother takes their child for their immunisations. After vaccine funds were suddenly cut, it has taken months to get the supply over the border, transport it over to their region, distribute it over the clinics and reappoint dismissed community health workers. The mother gets there but she has lost the child’s vaccination card. Her child won’t be able to get the vaccine this time. And no one can tell her if there will be another date.

Meanwhile, an overwhelmed community health volunteer tries to vaccinate hundreds. They are tired, frustrated and have no patience for more questions. A parent with doubt walks away. A local preacher comes into the waiting area to dissuade parents from vaccinating their children. There’s nobody there to counter misinformation. A number of concerned parents leave with them. 

The last mile of aid and policy delivery matters now more than ever. Getting it right can have life-saving consequences, especially when resources are scarce. 

Behavioural science can help identify and remove these barriers and frictions, and it can do so cost-effectively. 

One of our recent vaccination projects showed that a simple chatbot providing time-date reminders, with personalised information on nearby clinics and creating conversational commitments, almost quadrupled the rate of Covid-19 vaccination in Argentina. And following the most rigorous standards in academia, we successfully replicated this study and its impact.

This last point is important because learning what works, and what doesn’t work, at a time when funders are facing the trickiest of trade-offs, is one of the best tools to deliver value for money. 

So how can funders, policy-makers, local governments, NGOs and delivery partners put this into practice? Here are four recommendations:

1. Demand evidence and cost-effectiveness: Invest in interventions backed by rigorous evaluations, ensuring every pound spent is justified by measurable, evidence-based results.
2. Embed a human-centric approach from the start: Traditional aid programmes often assume that if you build a clinic or fund a school, people will automatically benefit. In reality, behavioural barriers like inertia, misinformation, or social norms can prevent life-saving services from reaching those who need them. By understanding these human factors, we can design interventions that nudge people toward positive actions at very low cost.
3. Allocate and prioritise budgets for rigorous testing: Allocate funds specifically for experimentation and RCTs. Testing small-scale interventions before large-scale implementation ensures resources focus only on strategies that deliver proven impact.
4. Build local behavioural capacity: Establish and support local behavioural science units within government agencies and NGOs in the Global South, promoting sustainable, culturally relevant innovation.

Global aid is under unprecedented pressure. Yet humanitarian needs and development challenges in the Global South have intensified. 

In an era of tough choices, strategic allocation of scarce resources matters more than ever. Behavioural science isn’t a solution to the overall shortage of resources. It cannot replace essential investments in healthcare, education, or infrastructure. However, by rigorously identifying how to remove last mile barriers, it can ensure limited resources have the greatest possible impact and deliver true value for money.