4.1 Leadership and advocacy

The most significant influencing work that Australia was involved in during the Horn of Africa (HoA) response was its political advocacy with other donors and its push to change their cautious approach. Australia's foreign minister visited Somalia with the executive director of the United Nations (UN) World Food Programme (WFP) in late July, to publicise the urgency and severity of the crisis. This marked a moment of genuine international leadership by Australia. Not only did Australia call for others to fund generously and take greater risks in Somalia, it led by example—releasing funds early and focusing assistance on Somalia. Following this and other appeals (together with key allies such as the United Kingdom's Secretary of State for International Development), donor 'caution' was relaxed and more aid flowed into famine-stricken areas.

During the crisis, the Australian aid program in Nairobi worked with UN agencies and other donors to have influence in the areas that affected how the 'system' could work most effectively. Stakeholder interviews showed clearly that Australian leadership in this area was appreciated.^e Australia leveraged its investment in WFP, making it move faster than it might have done otherwise. A useful partnership formed between the Post and colleagues in the Rome office whereby a rapid exchange of information allowed for sensible advocacy to help WFP improve its response (for instance, making sure the deployment of emergency staff was speeded up). Australia also pushed WFP and the food cluster for better reporting that gave actual rather than planned beneficiary numbers. Such advocacy work is often essential in emergency response to ensure the largest multilateral agencies are performing. The Australian aid program needs to ensure that staffing levels and expertise are sufficient to facilitate such work.

Conclusions

• Australia advocated for action and led by example through releasing funding early and taking some bold decisions.
• Active engagement with multilateral agencies by donors is essential to guarantee performance.

4.2 Coordination

Coordination of the international response was viewed as poor in the HoA crisis, possibly because to work in Somalia, individual agencies needed to negotiate with Al-Shabaab and could not share sensitive information.²⁷ Coordination 'hubs' established inside Somalia suffered staffing problems and found they could not easily organise meetings because security issues made it difficult for people to attend. The real-time evaluation from the UN Inter-Agency Standing Committee (IASC) found that there was a lack of strategic leadership from the key coordinating bodies, which resulted in no overall response strategy.²⁸ The IASC evaluation saw coordination as mostly occurring through the cluster system^f centred in Nairobi. This produced multiple substrategies at the individual cluster level. The cluster system had limited non-government organisation (NGO) participation, prompting one NGO consulted in this evaluation to say 'clusters have been a disaster here and should be scrapped'.²⁹ Parallel meetings were set up adjacent to the cluster system resulting in an overall crowding of the system.³⁰ Other factors contributing to dysfunctional coordination were the access constraints in Somalia and the consequent relocation of the majority of the response management to Nairobi, about a thousand kilometres away from the famine epicentre. The breakdown of coordinated action meant that there was not a common operational picture of needs or who was doing what.

In Nairobi, the Australian aid program took the view that coordination was critical, and looked to support the UN Humanitarian Coordinator's efforts to direct the overall operation.^g They tried to ensure Australia's partners, such as WFP, engaged properly in coordination activities, as well as pressing other donors to do so. A good example of the former was the Australian aid program lobbying WFP to make the food cluster active and effective. In terms of donor coordination, Australia took on the leadership of the informal donor coordination group in early 2012. Implementing partners and other donors considered that Australia had effectively improved coordination and played a valuable 'brokering role' between large donors, the UN system and smaller donors.

Conclusion

• In a complex conflict-affected environment, coordination is difficult. Despite this, the Australian aid program achieved some success in improving coordination and played a lead role in the coordination of food security among donors.

4.3 Staffing levels and support

The Australian response was characterised by the incredible dedication of aid program staff. The Africa program was tiny but growing fast when the crisis hit. In Nairobi, most staff members were new. In Canberra, staff already had a large and challenging portfolio. In these circumstances, Australia's fast, effective and well-regarded response is a testament to the hard work of those involved.

High volumes of aid handled by a small and relatively new team inevitably created challenges. There was little institutional experience in African famine and drought to call on. Neither was there a system in place to deploy extra support—especially humanitarian expertise—for complex conflict-related slow-onset crises. This affected the team's ability to monitor programs, adapt interventions and focus the portfolio. This is a key lesson for the Australian aid program.

This evaluation found that the staffing levels deployed to support the response were inadequate. Staffing levels were not appropriately increased in Canberra or in Nairobi, despite a huge increase in workloads. Comparative staffing numbers for other humanitarian responses can be seen in Table 6.

Table 6 Staffing levels for major Australian humanitarian operations

Category	HoA famine	Pakistan floods (2010)	Cyclone Evan—Samoa
Funding	$112 million ( in 2011)	$75 million	$9 million
Number of grants	44	29	3
Staff at Post	4	11	10
Dedicated staff at Post	none	2–4	5
Additional staff at desk	2	none	1 full time (1 week), 3 part time (1 week)
Crisis centre	Not activated	Activated: 20 staff initially, but then 4–5 at any one time for 6 weeks	Activated: up to 10 staff initially, but then 3–4 officers at any one time for 4 weeks
RRT officers	2	1 to assist in overall response	3
Other support	HER support for HPA and Dollar for Dollar Initiative	4 taskforce officers (4 weeks)	5 HER staff full time (1 week)

HER = Humanitarian Emergency Response Section; HoA = Horn of Africa; RRT = rapid response team

Despite the financial commitment to the HoA crisis being much greater than responses to the Pakistan flood in 2010 and to Cyclone Evan in Samoa, the HoA received the smallest amount of staff support (Table 6). Understaffing appears to be partly due to the designation of the crisis as 'slow onset' and therefore managed by the geographical desk. The Australian aid program has developed a well-functioning system, with a number of key tools for emergency responses to rapid-onset crises. Once an emergency is declared, a 'crisis room' system operates by putting a new team in place to manage it day to day, under the supervision of the Humanitarian Emergency Response Section. This can call on the rapid response team (RRT)—a group of staff across the organisation trained in emergency response and support roles. In the HoA crisis, the crisis centre was not activated and only two RRT members were deployed. Support from the Humanitarian Emergency Response Section was limited to use of the HPA mechanism, and the designing and running of the Dollar for Dollar Initiative.

The fact that money was being exclusively channelled to partner agencies seems to have been another reason why staffing levels were not increased. Some staff who worked on the response expressed the view that another factor may have been that staff with the required humanitarian expertise were not available within the wider organisation.

This evaluation has identified a number of issues that are broadly applicable in responses to slow-onset crises.

• Partner-mediated responses require additional human resources: The assumption that because there is no direct response the workload is less is not the case. In fact, the oversight required of partners can be just as time consuming and the briefing, high-profile visits, policy work, and responses to public and political interest remain the same for both types of emergencies.
• Some of the skill sets needed for complex humanitarian emergencies are different to those for 'natural' disasters: Complex humanitarian emergencies often demand deep knowledge of how to work in conflict, and the international architecture deployed in such situations. There is a need for more generic 'humanitarian adviser' personnel, who are familiar with the architecture of the humanitarian system and the niche of the different responding agencies. These skills are important for help in strategy setting and for steering the response once resources have been committed.
• The deployment times are longer in slow-onset crises: Typically, there is a need for heightened staffing levels for at least six months, meaning RRT deployments might be less useful for this type of scale-up as the removal of staff from their day jobs for this long is disruptive. The organisation may need to expand its cadre of humanitarian staff (who will be available for response in their regions, as well as deployable). Moreover, ways to scale up staffing rapidly, including through consultancy type arrangements, need to be found.
• The intense period for the first six weeks is very similar to rapid onset: In this period, the types of skills the RRT has to offer are as useful as in rapid-onset contexts—briefing, disbursal, strategy setting, contracting and so on.
• The management should stay with the desk, but the largest responses should have corporate oversight: There is a need to recognise that in the largest emergencies there should be a 'corporate response'—some sort of trigger that means senior management endorse the correct resources and tap into the skills of the humanitarian branch properly. A simple activation procedure—such as a meeting with some standard agenda items—would suffice.
• Working on humanitarian crises stresses staff through both workloads and exposure to traumatic suffering: The Australian aid program should recognise this as a matter of policy and make sure procedures are in place to support staff properly.

Conclusions

• There were few, if any, extra staff deployed to help with the HoA response, either in Canberra or in Nairobi, despite it being the largest ever response by the program in financial volume. Heavy workloads, as well as witnessing the devastating effects of the famine, led to extremely high pressure and stress on staff.
• The fact it was a slow-onset crisis meant that rapid-onset models did not apply. In the future, a system needs to be put in place to resource such responses rapidly and properly.

4.4 Flexibility of funding

Australia worked with partners to identify broad goals but gave partners scope to adapt programs to specific or evolving circumstances. Interviews found that Australia's flexibility as a donor was appreciated. Flexible funding is important, as it allows those closest to the work to make decisions on where the greatest need lies and how needs can be addressed most effectively.

Conclusion

• Australia was noted for being a flexible donor and should continue to provide flexible funding to partners.

4.5 Monitoring partner performance

This evaluation found a lack of capacity within the aid program during the HoA response to monitor the performance of partners. The lack of capacity was compounded by the difficulty of obtaining good data in a conflict zone and the general lack of transparency within the sector. As a result of limited staffing, resources were quite sensibly focused on monitoring the largest grants, and in particular on WFP. The effective work that the Australian aid program did in pushing WFP for greater transparency (about lack of access), and pushing for coordination mechanisms such as the food security cluster to be established, shows how donors are often obliged to 'actively manage' their funding portfolios. The greater the financial commitment and the more complex the situation, the more active management is needed. A conclusion of this evaluation is that during crises Australia needs to deploy routinely more capacity to engage with partners and monitor their operations.

The level of detail of information needed to evaluate retrospectively the performance of implementing partners is not available. Agencies, particularly UN agencies, did not report in any detail on what was achieved as a result of flexible funding. Nairobi Post accepted that there was a direct trade-off between flexibility and reporting on Australia-specific results. As a result, by the end of the crisis, knowledge of the impact of Australian funding was limited.

An illustrative example of this issue is provided by the UN High Commission for Refugees (UNHCR) who were given $15 million for a 'regional refugee response'. UNHCR did not report on this expenditure beyond its standard reporting (annual country reports, global reports), as is the agreement. When more information was requested for this evaluation, the agency confirmed that all of the funds had been allocated to the Dadaab refugee camp in Kenya, but what the funds had been spent on could not be identified. The evaluators were told that if Australia had specified in advance where funding should be spent a detailed expenditure report would have been provided, but because funding was given flexibly funds had not been tracked.

Beyond standard reporting, little additional information about what was delivered and what was achieved was available to this evaluation or forthcoming on request. Some UN agencies stated that requests for information were not met as they created extra work and were not in the original agreement. NGOs routinely provide more information about their work, both ahead of time and retrospectively, than the UN agencies. Asking UN agencies to provide this level of information would help Australia to better understand who is routinely performing.

Flexible funding need not require a trade-off with reporting. Based on the HoA experience, there are some key principles that will help Australia's aid be flexible without reducing the usefulness of reporting:

• Agencies do not need to be overly prescriptive at the proposal stage. Proposals can be solicited on the basis of broad strategic fit, with a commitment to achieve high-level outcomes.
• Reporting needs can be specified in some detail in agency level granting agreements. All agencies have detailed internal accounting systems; donor funds are allocated codes and reports can be generated against this to show precise expenditure on all lines. This type of financial accounting should be required as routine.
• Australian aid can specify reporting on outputs or outcomes in its agreements. Reporting requirements can be based on what is practical and how the information will be used.

Standards and policies requiring sex disaggregation of beneficiary numbers also continue to be ignored or deemed infeasible by many agencies. While this is usually done at project level, numbers are then aggregated for higher-level reporting. This precludes monitoring and reporting on the most basic level of equity, comparing the value of relief items distributed to men and boys with the value of relief items distributed to women and girls.

As the humanitarian work of the Australian aid program has grown in complexity, scale and ambition, the organisation has begun to build humanitarian expertise. Recently, a humanitarian adviser has been appointed in Canberra and there are also a growing, but still insufficient, number of humanitarian professionals in Posts. A mechanism to redeploy these types of advisers rapidly, or to be able to bring in others at short notice, would give the organisation access to robust monitoring capacity, in turn allowing for better informed decision-making and follow-up.

Conclusions

• Efforts to monitor partner performance during the crisis focused largely on WFP.
• An inadequate deployment of staff to work on the crisis meant that the Australian aid program was not able to monitor its portfolio in detail.
• NGOs typically reported in more detail than UN agencies.
• A number of UN agencies receiving significant levels of funding did not provide adequate information on outcomes, and did not disaggregate data by sex and age.
• Flexibility has become conflated with a lack of scrutiny, but this should not be the case. Future agreements with UN agencies should continue to provide funds flexibly but also require greater specificity in reporting and accountability.

4.6 Choice of implementing partners

The Australian aid program's strategic intent was sound but the operational reality in the HoA complicated humanitarian responses. Al-Shabaab controlled the worst-affected areas, where they banned many aid groups and made delivery dangerous and difficult for the rest. The complex contextual constraints and lack of detailed reporting by some partners make it difficult to evaluate the choices about implementing partners.

Largely, Australia made good choices about which partners to fund. The level of WFP funding and the small numbers partners who used of cash-transfer programming (see Box 1) are the exceptions. The following covers the main funding decisions.

World Food Programme

The background to the decision to focus funding heavily on WFP is worth setting out to understand the constraints and opportunities that existed at the time.

At the time of the famine declaration, the international aid community was trying to find ways to get a general distribution of food going. Widespread free food distribution is an obvious thing to do in famine. Increasing the general availability of food is also beneficial as it helps ensure that the special nutritional foods are used to treat clinically malnourished children as intended, as well as reducing the need to sell assets for food. WFP is the pre-eminent deliverer of emergency food globally. It is consistently rated as one of the best-performing humanitarian agencies, with a reputation for fast response and slick logistics. They scored highly in Australia's Assessment of multilateral agencies and in the United Kingdom equivalent, the Multilateral assessment review. WFP is one of the few agencies that can deliver food to millions of people rather than hundreds or thousands.

However, in July 2011, the capacity of WFP to instigate large-scale food distribution in the worst-affected areas of Somalia was limited. In the years leading up to the crisis, the combination of a UN monitoring report criticising WFP for inadequate scrutiny of Somali contractors³¹ and unacceptable operating restrictions enforced by Al-Shabaab meant WFP ceased operations in south-central Somalia. Many donors suspended financial support for WFP. An additional problem faced by WFP was being banned by Al-Shabaab in January 2010.³² As a result, when the crisis escalated, WFP had no access to Al-Shabaab–controlled areas and had not received substantial funding, meaning it had few resources should it gain access.³³

Although WFP was not operational in the most severely affected regions,³⁴ other agencies were not able to fill the gap because contingency plans for the absence of WFP had not been developed. Consequently, despite WFP's lack of access and funding for almost a year, the agency was still both perceived and promoting itself as the agency to undertake large-scale food distribution in Somalia. It was hoped that Al-Shabaab, faced with a catastrophe in their area and a change in policy from WFP, would enable WFP to commence large-scale food distribution in the worst-affected areas.

Al-Shabaab did not allow WFP access to south-central Somalia, preventing any impact it might have had in the epicentre of the famine at the peak of the crisis. WFP also took time to get going, starting from a very low base, having lost much of its capacity. It was September 2011 before WFP Somalia was able to adequately scale up operations, and even then only in the border areas and Mogadishu were there were large numbers of internally displaced persons (IDPs). As a result, only 24 per cent of funding of WFP in Somalia went to the south-central area (Figure 6). Consequently, by financing WFP in Somalia, Australia arguably did not fund the most immediate life-saving interventions. At the time, it was not possible to know that WFP would not have access, although it was possible to predict it would take time to scale-up.

At scale, WFP distributed food across significant areas in Somalia, principally in Mogadishu, Puntland and the border areas, which received large numbers of IDPs. Most WFP assistance was provided outside the worst-affected region. Intrinsically, there is nothing wrong with this, especially if others are providing assistance in other regions—it can be seen as a division of labour. But it was not the Australian aid program's original intention.

Throughout the crisis, WFP continued to operate on a food-delivery model despite reforms in headquarters to increase the agency's use of cash transfers. The use of a food-delivery model to some extent reflected the agency's focus on IDPs in Somalia. However, this meant that the bulk of Australian assistance did not catch the wave of cash-based programming that that proved effective in famine areas. This significantly affected the efficacy of the Australian intervention in Somalia.

Figure 6 Regional breakdown of Australian financial assistance to Somalia through the World Food Programme

The strong focus on WFP did, however, deliver good outcomes in both Kenya and Ethiopia. In Kenya,Australia also gave $8.6 million to WFP Kenya for the refugee-feeding operation. WFP used local implementing agencies to support a general food distribution and a supplementary feeding program. Given the huge influx of refugees to Kenya, the massive challenge this represented for the Government of Kenya, and the strong performance of WFP and UNHCR^h in meeting these challenges, this seems like both a sensible investment and a well-founded priority.

UN High Commission for Refugees

Outside Somalia, the second-largest overall commitment was for refugees in Kenya, principally in the Dadaab camp. Once again, part of this prioritisation was intended, and part of it determined by the UNHCR. Australia donated $15 million to UNHCR for their regional refugee appeal, all of which was used in the Dadaab camps. As stated above, this funding was well justified and effectively used.

UN Common Humanitarian Fund

The UN Common Humanitarian Fund (CHF) was very useful in the crisis, as it was only mechanism through which local (Somali) NGOs could be supported. Its decision-making process uses operational agencies to determine priorities, making it theoretically closer to the ground. It is also administratively straightforward, with opportunities for donors to engage at a strategic level in decision-making. However, the CHF process was slow throughout the period under examination. The 'standard allocation' took an average of three months to get money out to partners, and even the 'emergency reserve' (approximately 20 per cent of the CHF) took on average 11 weeks.³⁵ Compared to Australia's two-week delivery³⁶ this is a critical flaw, especially in a famine where every day counts. The reason appears to have been administration based in Geneva (the CHF is run by the Office for the Coordination of Humanitarian Affairs [OCHA], part of the UN Secretariat and therefore subject to their rules)—an issue that constantly impedes such mechanisms.

UN Children's Fund and UN Food and Agriculture Organization

In Somalia, UN Children's Fund (UNICEF) and UN Food and Agriculture Organization (FAO) were widely perceived to be among the best-performing UN agencies. They both had access in Al-Shabaab areas, and significantly expanded their programs to attempt to cover the absence of WFP. UNICEF's cash program, and its de facto lead in this sector, was excellent.³⁷ FAO also supported affected communities with cash transfers and used innovative measures such as the approach to Dubai-based Somali food traders to increase their presence. Other forms of assistance were also exploited by both agencies: UNICEF undertook significant water and sanitation work, and was the lead UN agency for nutrition supplying Plumpy'Nut (a high-energy peanut-based food), and FAO worked to improve livelihoods.

International Committee of the Red Cross

The International Committee of the Red Cross (ICRC) appears to have performed particularly well. The free food distribution that took place in south-central Somalia was largely delivered by the ICRC, who distributed food to about a million people. This was possible because the ICRC:

• was very experienced in working in south-central Somalia
• managed to remain in critical regions for longer than most other international organisations, as it was not expelled by Al-Shabaab until January 2012
• engaged Somali program staff, who then subcontracted local assistance.

ICRC also supported the health clinics of the Somali Red Crescent Society (which enjoys a good reputation), as well as undertaking their traditional protection work.

Retrospectively, ICRC should probably have received more funding as, at the time funding allocations were made, they had access to the most severely affected areas of Somalia and, in interviews, indicated they had some capacity to use additional funding.

Overall partner choice

In hindsight, while partner performance was clearly related to their capacity to work in country contexts, the choice of implementing partners was appropriate overall. The funding of multiple UN agencies helped to achieve the strategic intention to focus assistance on Somalia and address food and refugee needs. A good learning point for the Australian aid program is the need to ensure that decision-making should place greater weight on the informed analysis of those closest to the operation including Post, program areas and other donors with extensive regional experience. Such an approach would have resulted in more funding for the ICRC, UNICEF and FAO, and proportionately slightly less for WFP.

Conclusions

• Effectiveness of implementing partners was highly reliant on context.
• WFP in Somalia took time to scale up and were constrained by access issues, which meant Australian funding did not immediately reach those most in need as intended. Later, once WFP became operational in Mogadishu and border regions, it did succeed in delivering food to famine victims.
• Funding to WFP and UNHCR was used to provide much needed assistance to the refugees in Kenya, which was effective under the circumstances.
• The Red Cross Movement (and ICRC in particular) performed consistently highly.
• Of the UN agencies, UNICEF and the FAO performed well in Somalia, and WFP and UNHCR in Kenya. The CHF provides an excellent mechanism on many levels—reach, analytical ability, coordination, local NGOs—but was severely constrained by slow disbursal.
• Decision-making should place greater weight on informed analysis provided by those closest to operations.

4.7 Appropriateness of NGO funding mechanisms

In the HoA crisis, funding mechanisms determined which NGOs were funded. Funds were allocated to NGOs that were in the Humanitarian Partnership Agreement (HPA) or funded by the Australian public.

Humanitarian Partnership Agreement

The HPA was expedient rather than strategic in the HoA crisis. It disbursed funds quickly with minimal paperwork but with limited volumes in a protracted African crisis for which it was not specifically designed. All six HPA agencies had presence in the region and were able to prioritise effectively the funding among them to ensure funds went to the members best placed to respond. The consideration of proposals by the HPA NGO partners was based on OCHA assessments, in-country partners and public sources.³⁸ Beyond this, there was no intention to coordinate and there was limited collaboration within the HPA. Oxfam championing the use of gender action plans is an example of the collaboration that did occur.

The HPA partners conducted their own evaluation into the mechanism, finding that 'projects generally met or exceeded the output targets in their implementation plans' and 'a number of examples of innovation and good practice can be found in project reporting'. Support provided to NGOs was used in Somalia but also Kenya and Ethiopia in the life-saving phase of the emergency immediately after the famine declaration.

Australian assistance provided through the Humanitarian Partnership Agreement was used by CARE to provide water in Dadaab camp in northern Kenya. Photo: Kate Holt, CARE Australia.

The NGOs in the HPA mechanism were probably among the Australian NGOs best placed to deliver assistance in the HoA. However, like WFP, at least two had been excluded from south-central Somalia by Al-Shabaab. Several other NGOs, which did not have an Australian base, were much better placed to deliver assistance in the worst-affected parts of Somalia—for example, Action Against Hunger and the Danish Refugee Council. This is not to deny the good work highlighted by the HPA's evaluation, but only to suggest that a mechanism designed for rapid response in Asia and the Pacific could not be expected to have the same 'fit' when applied to a famine in an area of conflict in east Africa.

The question of what might work better than the HPA is difficult to answer. One option is to let those closest to the crisis decide which agencies are best placed to respond to the most pressing need. However, this requires humanitarian liaison capacity and, potentially, means more administration (multiple grants), which could compromise speed. Working with locally based NGOs requires a mechanism to be in place to ensure due diligence. Other options include an adapted mechanism that allows for separate procedures for slow- and rapid-onset crises, or some way of assessing and funding potential new partners directly.

Dollar for Dollar Initiative

Opinions are divided on the merit of the Dollar for Dollar Initiative. It was organised at short notice and therefore had to be designed in a hurry. It required a significant amount of work and time to formulate and negotiate parameters (which donations could be accepted) and resulted in funding going to an inordinately large number of NGOs. By the time funds were being disbursed, the peak of the famine was over. Some NGOs were still spending funds when this evaluation was conducted. Funds were largely spent on recovery activities or long-term development rather than emergency assistance.

Like the HPA mechanism, the Dollar for Dollar Initiative was not necessarily the best way of channelling funds to organisations with operations in the heart of the famine zone. The majority of the Dollar for Dollar Initiative agencies, as with the HPA, acted as conduits for funding to sister agencies in the HoA, or into larger families of organisations. Where the funding went depended on where these organisations had presence. Much of the funding, for example, ended up in Kenya, because it had the greatest number of pre-existing agencies with links to Australian NGOs (usually engaged in long-term development work). These historical ties and agencies usually meant an intimate knowledge of the area, making for some excellent projects. In a small number of cases, however, it meant funds going to areas that were only marginally affected by the famine. One agency's rehabilitation of a water system in coastal Kenya falls into this category. In the future, the effectiveness of such schemes could be increased by the aid program directing that funds must be channelled to priority needs.

The large number of organisations funded through this initiative meant that it was administratively burdensome. A fundraising consortium or entity, able to receive and administer funds, could have streamlined the initiative. On the other hand, the Dollar for Dollar Initiative engaged the public and was a major fundraising success.i The reporting forms and formats seemed widely appreciated by the agencies, suggesting that work done in the initial scheme could be used or modified for any subsequent effort. A public engagement and fundraising strategy needs careful thought, in collaboration with NGOs, ahead of the next large-scale emergency.

What the Dollar for Dollar Initiative did offer, beyond public engagement and extra fundraising (two very big bonuses), was the opportunity to channel funds to NGOs that had no other opportunity to access Australian emergency aid. This is a serious consideration in the current funding architecture for emergency response, and one that needs to be looked at thoughtfully. There is no doubt that NGOs are one part of the large humanitarian architecture and they, like the other parts, need to function to contribute to an overall rapid and effective response. Learning from the HPA and Dollar for Dollar Initiative suggests the need for more such mechanisms, perhaps in conjunction with other donors, and perhaps on a regional- or crisis-specific basis.

Conclusions

• The HPA enabled a rapid disbursal of funds and the funded NGOs achieved some good results. However, the HPA was not designed for the HoA context and, as a result, had mixed coverage.
• The Dollar for Dollar Initiative raised funds and engaged the Australian public, but it was administratively burdensome and did not provide funding until late in 2011.
• Some NGO-funded programs delivered critical life-saving assistance. Other programs focused on the recovery phase and built resilience.

4.8 Appropriateness of sectoral funding

Although Australian funding was generally focused on key sectors, there was also considerable spread across all sectors, largely as a result of NGO funding. In particular, funding provided through the Dollar for Dollar Initiative was provided without direction and, as a result, was spread across all sectors. The severity of the crisis meant that people needed assistance in all sectors.

The evidence from this evaluation suggests that a certain amount of focus is desirable. Funds dispersed too flexibly may not end up addressing the most urgent needs, and certainly may not flow to where Australia intends. Funds can be dispersed on the basis of addressing a particular sector or area, but it can be left up to the agencies on how exactly to address the problem. A strategy can outline broad focus—the sectors, places and people that are priorities—and the detail can be worked out collaboratively later. This is flexible, but also focused.

Conclusion

• The aid program probably should have directed agencies towards the sectors that Australia wanted to fund and in so doing reduced some of the sectoral spread

4.9 Appropriateness of modes of assistance

With WFP as its main partner, Australian assistance was focused on funding of food aid. A 2012 evaluation of the WFP Somalia program for the past decade concluded that 'there is a need to open up the debate on the relevance and impact of food aid, including its impact on vulnerable households, the dependency it generates and the degree to which it complements or undermines the agricultural/pastoral economy'.³⁹

In the famine epicentre, Al-Shabaab control made it very difficult, costly and dangerous to get food in. Some humanitarian agencies that were still able to work in these areas exploited a novel way of helping people that was already being piloted in Somalia—cash. Cash-based programming has been gaining credibility in humanitarian operations for a number of years (see Box 1). In Somalia, it was carried out on a massive scale in 2011–12, with more than US$100 million in cash handed out, coordinated by UNICEF.

A recent evaluation of cash transfers, commissioned by UNICEF, concluded that,

The unconditional cash and voucher response … quickly achieved an impressive scale, building principally on international and Somali NGO field capacity. The evidence marshalled in this evaluation suggests that cash and vouchers made a quantifiable difference in reducing hunger and improving food security, enabling a more rapid recovery than would have been possible without assistance.⁴⁰

The UNICEF evaluation also found that even in the worst-affected areas of Somalia, cash transfers attracted merchants and people were able to buy food. Another benefit of this modality was that there was little diversion or stealing. If anything, the diversion rates appear to have been lower than with food. On balance, cash transfers benefited women and men equally and put decision-making at family level.⁴¹ Interviews conducted with beneficiaries of Australian aid in Wajir, Kenya, similarly found that cash transfers had been a valuable form of assistance (Box 2).

The fact that such a large-scale cash transfer program took place,^j in such a complex and risky environment, and was apparently successful, should be a clear signal to the Australian aid program.

Conclusion

• Cash transfers were an extremely important substitute for food distributions that worked well but were not fully exploited.

Footnotes

e Several donor agencies interviewed referred to Australian leadership on donor coordination; the United Nations was appreciative of the critical but constructive engagement.

f The 'cluster' system essentially arranges coordination by technical groups with a designated United Nations (UN) agency in charge. So the World Food Programme coordinates food, United Nations Children's Fund water, World Health Organization health and so on. The Office for the Coordination of Humanitarian Affairs (OCHA) and the humanitarian coordinator are responsible for bringing these various groupings together to provide overall coordination.

g The humanitarian coordinator is the most senior UN official in charge of the response on the ground. UN OCHA supports this position by providing a coordinating office.

h While there were undoubtedly many challenges, the mortality rate remained below emergency standards in the Dadaab camp despite refugees arriving in extremely poor condition. This suggests the aid operation was effective in meeting immediate needs.

i A report produced by Media Monitors for AusAID found that, in the 71 days between 5 October 2011 and 14 December 2011 (the period of the Dollar for Dollar Initiative), there were 137 separate media items. This coverage reached a cumulative audience of 9 062 444 people.

j The UNICEF cash evaluation suggests that 'the declaration of famine in July 2011 effectively created "consensus by compulsion in the face of famine" making significant funds available for a rapid cash-based response'.

References

27 VL Hammond, Humanitarian space in Somalia, Overseas Development Institute, London, 2012.

28 J Darcy, et al., 2012, p. 28.

29 Interviews for this evaluation.

30 J Darcy, et al., 2012, p. 32.

31 M Bryden, A Laloum & J Roofthooft, Report of the Monitoring Group on Somalia pursuant to United Nations Security Council resolution 1853 (2008), 2010.

32 A Seal & R Bailey, The 2011 famine in Somalia: lessons learnt from a failed response? Conflict and Health 7(22):2, 2013.

33 Interview with the WFP country director, March 2013.

34 A Seal & R Bailey, 2013.

35 G Taylor, B Willets-King & K Barber, 2012.

36 Evaluation interviews with HPA agencies and WFP.

37 K Hedlund, N Majid, D Maxwell & N Nicholson, Final evaluation of the unconditional cash and voucher response to the 2011–12 crisis in southern and central Somalia, Humanitarian Outcomes & UNICEF, 2012.

38 Australian Agency for International Development, Mid-term review of the AusAID-NGO Humanitarian Partnership Agreements (HPA) 2011–2014, AusAID, Canberra, 2013.

39 N Nicholson, K Longley, M Fisher & T Walters, Somalia: an evaluation of WFP's portfolio, volume I—full report, Country Portfolio Evaluation, Office of Evaluation, UN World Food Programme, Rome, 2012.

40 K Hedlund et al., 2012.

41 K Wasilkowska, Gender impact analysis in the unconditional cash transfers in South Central Somalia, Somalia Cash Consortium, 2012.