Humanitarian organisations have traditionally supported crisis-affected people with physical commodities such as food and shelter. But today, many NGOs are experimenting with an alternative approach. Instead of in-kind aid, a family may receive an envelope of cash, a plastic card or an electronic money transfer to a mobile phone, with which they purchase what they need locally.

During the 2011 famine in Somalia, for example, which killed more than a quarter of a million people, aid agencies used remittance companies to provide cash transfers to more than 1.5 million people. Refugees in Lebanon are also using smart card vouchers to buy goods as local shops and an existing government social protection programme in the Philippines was extended to send cash to those in need after Typhoon Haiyan.

Although we do not know the exact amounts, the best estimate is that cash and vouchers together have risen as a proportion of total humanitarian spendingtoday from less than 1% in 2004 to about 6% today. But this is much less than it should be.

The hold-up isn’t the strength of the evidence which is increasingly clear and compelling. Cash transfers are among the most rigorously evaluated and researched humanitarian tools of the last decade. People spend cash sensibly: research suggests they are not likely to spend it anti-socially (for example, on alcohol) and cash is no more prone to diversion than in-kind assistance. Local markets from Somalia to the Philippines have responded to cash injections without causing inflation. With the growth of digital payments systems, cash can also be delivered in increasingly affordable, secure and transparent ways.

People often prefer receiving cash because it gives them greater choice and control over how best to meet their own needs, and a greater sense of dignity. When people receive in-kind aid that doesn’t reflect their priorities they often sell it to buy what they really need as, for example, 70% of Syrian refugees in Iraqhave done. The difference in what they can sell food or other goods for and what it costs to provide is a pure waste of limited resources and means that in-kind aid is often just an inefficient way of giving people cash.

It also usually costs less to get cash transfers to people than in-kind assistance because aid agencies do not need to transport and store relief goods. A four-country study comparing cash transfers and food aid found that 18% more people could be assisted at no extra cost if everyone received cash instead of food.

So, given the strong case for cash, why is it still at only 6% of humanitarian spend? What’s the hold-up in getting to 30%, 50% or even 70%? Part of the answer is the long tradition of governments and organisations deciding what people need, and assuming that they cannot be trusted to make sensible decisions themselves. These priorities often reflect organisational mandates and interests hard-wired into the humanitarian system. Fears that cash will be misused are deep rooted and do not simply fade away on the first sight of evidence to the contrary. Organisational inertia is also an important factor; faced with uncertainty, agencies default to familiar forms of assistance, which largely remain in-kind. Donor governments and aid agencies also fear that western publics will not support cash – seeing it as less visible or more prone to corruption.

Degan Ali, one of the Somali leaders of the cash response to famine in Somalia in 2011, recalled: “Despite eight years of successful cash assistance in Somalia, aid agencies and their donors were reluctant to support cash transfers. Many donor countries had enacted anti-terrorism legislation and so aid agency staff feared prosecution should cash be diverted. In the end, we implemented the largest NGO cash transfer programme in history. But this was done despite, rather than with the support of, the humanitarian system. It was a battle, every step of the way, and in the months it took to fight it, many thousands of people died unnecessarily.”

Ultimately, it comes down to trust. The aid sector in general is bad at trusting people in poverty and reluctant to hand-over power and control. It’s fundamentally premised on the idea of the external experts deciding what is needed. Cash challenges that and so creates a series of professional insecurities. What will happen to logisticians and food aid professionals? And if it makes sense to deliver one cash grant to meet a range of basic needs, do we need so many organisations focused on delivering humanitarian assistance?

Giving people cash is not always the best option, nor is it a cure-all. Sometimes markets are too weak for people to be able to buy what they need and sometimes government policies make it impossible to provide cash. But these situations are rare and often temporary. They should not be used as bad excuses for providing relief in-kind after these constraints no longer apply. Cash transfers still need to be complemented by the provision of public goods that markets will not provide efficiently, such as clean water, sanitation or immunisation. The use of cash transfers does not mean that humanitarian actors should give up their key roles of proximity, presence and bearing witness to the suffering of crisis-affected populations. Indeed, making delivery more efficient protects the time and resources of humanitarian actors to do just that.

Paul Harvey is a partner at Humanitarian Outcomes and a research associate at theOverseas Development Institute (ODI).