The Department for International Development’s approach to managing aid in conflict zones is often weakened by a desire to protect its own reputation, with staff reporting confusion over how to implement DfID’s key policies, a report has found (pdf).

According to the Independent Commission for Aid Impact (Icai), which investigates taxpayer-funded UK aid, DfID staff in conflict-affected areas are satisfactorily assessing and mitigating the risk of aid misuse. However, the report found that weaknesses in the approach to risk transfer, as well as staff confusion over “zero tolerance” and “high-risk appetite” policies, require urgent attention in order to protect UK aid funds and make maximum impact.

DfID has committed to spending half of its budget in fragile states and regions – a sum worth £5.5bn this year – where government access is limited, and the risk of fraud, looting, aid diversion and informal taxation is high. As a result, it is crucial that DfID has central systems in place for staff to classify, rate, escalate and monitor these risks with confidence, said Dr Alison Evans, Icai’s chief commissioner and the report’s author.

“You’ve got so many different ways that food aid, for example, can end up in the wrong hands,” said Evans. “You have to have more than one strategy for working out where that asset ends up.”

Unlike most other donors, DfID decentralises its fiduciary risk management to country offices, where the staff responsible for programmes are known as “senior responsible owners”. Evans said this focuses attention on the changing risk environment, encouraging personnel to implement management systems autonomously and work with local partners to control fiduciary risk more effectively.

But there is a downside. Because DfID lacks a clear, overarching risk-management system, and country offices categorise and rate risks themselves, there is a huge variation in how risks are defined and assessed across the department’s work in conflict areas. This, in turn, can affect multilateral or corporate partners.

Icai’s report, published on Tuesday, assessed performance in five conflict-affected countries – the Democratic Republic of the Congo (DRC), Somalia, South Sudan, Syria and Yemen – using interviews with DfID staff and implementing partners, and country strategy reviews. DfID was given a “green-amber” rating, indicating that its fiduciary risk management in conflict-affected states, though mostly satisfactory, required urgent attention in some areas.

A number of senior staff expressed concern about the frequency with which decision-making was driven by “reputational considerations” rather than fiduciary risk itself, said Evans.

“How do you manage that political tension? We found that to be a very live discussion, particularly in environments where DfID is trying to work with local partners, or where DfID is working in unfamiliar territory, of which Syria is a good example,” she said. Staff can be torn between working with a smaller, local partner (which can mean a higher fiduciary risk) or a larger (potentially lower risk) private sector partner. Fearful of the “reputational risk of a possible loss or diversion of aid funds”, the larger partner is often chosen, despite a 2013 report by the Overseas Development Institute showing that local aid frequently delivers equal, or greater, development returns.

“There is not a long track record for this,” added Evans. “How far DfID goes in working with these organisations, and how quickly, is a matter of balancing both fiduciary and reputational risk … [But] there is a certain concern that DfID is putting a lot of its eggs in particular baskets with certain contractors.” This narrowed the pool of potential suppliers and possibly reduced value for money, said Evans.

DfID’s dependence on taxpayers’ money meant it was up to staff to protect and spend that money as effectively as possible, said a DfID spokesman, making the department likely to “err on the side of caution” in assessing risks. He added: “There are of course challenges to providing life-saving aid in some of the most difficult and dangerous places in the world, including disasters and war zones, which is precisely why we have such a rigorous system of checks and measures in place.”

The report found DfID to be at its most effective when using technological innovations to manage fiduciary risk, particularly in countries where access is limited. In Somalia, where the department has only a limited presence in the capital, Mogadishu, and is instead required to work primarily out of the Nairobi office, using third parties to deliver assets, tracking technology has been deployed to follow money and assets in real time. A similar programme, tracking pharmaceuticals, was used in DRC.

“That seemed to show to us how [DfID’s] decentralised system really plays up to its strengths,” said Evans. “We feel DfID could do a better job spreading this experience across its portfolio and learning how it might be applied in other contexts.”