How can social assistance better embrace uncertainty in protracted crises?

Published on 2 September 2022

Jeremy Lind

Professorial Fellow

In recent years, there has been a push backed by some of the largest donor and relief organisations to plant the seeds of social protection systems in even the most difficult protracted crisis settings. The case for doing so seems compelling enough: to identify alternatives to costly, perennial humanitarian aid operations so that more reliable delivery under highly variable conditions can take place. However, do such approaches – which are premised on anticipation, early warning and planning – work, or should we adopt more agile approaches suitable for uncertainty?

Woman fetching water in Ethio

The appeal of establishing social protection systems is to develop a policy, institutional and delivery architecture for providing regular support. This should ideally, over time, be handed over to states or other trusted public authorities. Yet, these efforts face the same entrenched challenges that plague previous efforts to create more reliable assistance in protracted crises. These challenges are: weak state capacities and/or political will; inadequate financing; low trust and public confidence; inadequate or destroyed infrastructure for aid delivery; continuing insecurity; and threats to the safety of frontline workers as well as beneficiaries.

Conditions of uncertainty and ignorance are therefore the dominant features of protracted crises. However, even while this is obvious, too often a control-focused, risk-assessment and management approach dominates at the nexus of social protection and humanitarian assistance. Risks are attributed either to individual characteristics (idiosyncratic risks such as illness or the death of a family member) or to an exogenous factor with wide effects (co-variate risks such as a drought or conflict). The ideal outcome for any intervention is seen as a stable state, and a return to conditions of predictability. This might be aided by the implementation of sequenced interventions and expert inputs. The focus here is on predicting and managing known risk events through centralised and (state-led) formal systems.

From risk to uncertainty

Standard risk-management approaches can paper over cracks that might open when providing social assistance in protracted crises, causing major shortcomings. The mismatch between the assumed design of systems and programmes, and actual practice on-the-ground by field-based implementers is often stark. Differences in outcomes on-the-ground are attributed to implementation gaps; targeting errors; exclusion by design; lack of co-ordination; poor governance; and sometimes vague, black-boxed notions of “political economy.” Arguing that all that is needed to fix these challenges is better planning, implementation, correcting errors, and improving co-ordination does not deal with the core issue: the probabilities of future outcomes are unknown and cannot be predicted.

What if embracing uncertainty rather than predicting and controlling risks was more central to the design of assistance to those living in protracted crises? Our recent open access paper in Development Policy Review – Providing social assistance and humanitarian relief: the case for embracing uncertainty – responds to this challenge. We argue that embracing uncertainty can help us rethink the intersection of social assistance, humanitarian relief, and disaster response in fundamental ways, resulting in a greater alignment of on-the-ground practice with wider policy and institutional approaches.

Toward a “high-reliability” approach

Based on case study material from Ethiopia and Libya, we outline four features of a “high-reliability” approach that works with uncertainty and complexity:

  • High-reliability professionals:” This refers to individuals that must move between the “nuts and bolts” of the system and wider horizon-scanning and modelling of unknown futures. In the process, they make sense of complexity, experimenting, adapting, and responding quickly. For example, in Libya, mosque leaders and Imams use their networks to reach the local population that might require assistance through the Zakat Fund (sunduq al-Zakat) – an independent entity operating outside the jurisdiction of the Ministry of Social Affairs and autonomously from the two governments operating within Libya. Mosque officials identify the neediest in a particular area. Volunteer “social researchers” (baheton ajtemaeyon), who are known for their integrity and honesty, then follow-up to visit the people who have been identified and to determine their eligibility.
  • The predominance of hybrid arrangements in reliable systems, which combine actors and practices that cut across formal/informal and state/non-state: While formal systems provide a broad infrastructure, it is locally embedded networks within organisations at the operational frontline that become important for learning and adaptation. In Ethiopia, targeting guidelines for the Productive Safety Net Programme (PSNP) explicitly include the involvement of local traditional leaders in the predominantly pastoral lowland regions of Afar and Somali. Local elders and clan officials make use of vernacular understandings of risk and need. Their approach is conditioned by understandings of fairness and fills the gaps in provision.
  • Accumulated learning and sharing of information across and between peers within networks: Learning from (non-catastrophic) failure can help in this process as networks of professionals navigate uncertainty together, gathering experience along the way. Such experiences may result in innovative practices and procedures, developed from the bottom-up in ways that become vital for assuring safety and reliability. In Ethiopia, the involvement of local traditional leaders in targeting decisions has made the PSNP more responsive to cultures of sharing, as well as to the implications of for targeting.
  • Organisational and funding flexibility: Since futures remain unknown, fixed budgets, standardised plans, and uniform protocols are inappropriate. Instead, plans provide frameworks not diktats and contingency plans, and anticipatory financing can allow rapid response if these remain flexible and not linked to pre-defined risks.

Realising a “high-reliability” approach requires fundamental changes to everything from programme design, professional practice and skill sets, organisational culture, incentives to funding, as well as monitoring, and evaluation regimes. However, ignoring uncertainty—and not accepting inevitable ignorance and surprise—can be misleading, and even downright dangerous.

Fortunately, there is already wide recognition – and even celebration – amongst policymakers of the need for adaptive flexibility and contingent responses within wider thinking about the nexus of social protection and humanitarian assistance, at least at a rhetorical level. Everyone is all too familiar with settings that are completely unpredictable.

More importantly, in seeking ways forward that embrace uncertainty, we can learn from the actual practices of those already involved in many existing social assistance implementation efforts. These are usually hidden, informal, and outside the regularised, accepted practice of standard programmes and policies, involving actors both inside formal agencies and within local communities. These people and practices can point to a way to doing things differently – and embracing uncertainty in social assistance.

The views expressed in this opinion piece are those of the author/s and do not necessarily reflect the views or policies of IDS.


Supported by


About this opinion

Ethiopia Libya

Related content