Speakers

Dan Maxwell – Professor and Research Director at The Feinstein International Center, Tufts University, primary author of Network Paper 73

Nick Maunder – Independent Consultant

Simon Levine – Research Fellow, Humanitarian Policy Group

Duncan Barker – Humanitarian Resilience Adviser, Humanitarian Disaster and Resilience Policy Group, Department for International Development

Wendy Fenton – Coordinator, Humanitarian Practice Network

Introduction

Wendy Fenton introduced the panel and offered some initial thoughts on Network Paper 73 – Response analysis and response choice in food security crises: a roadmap, co-authored by Dan Maxwell, Heather Stobaugh, John Parker and Megan McGlincy. She opened the event by framing Maxwell’s paper as answering three questions: what response analysis is, what the food security community says it is doing and what it is actually doing.

Presentation of Network Paper 73 - Response analysis and response choice in food security crises: a roadmap

Maxwell introduced his paper by referencing an earlier HPN Network Paper titled “Missing the point: an analysis of food security interventions in the Great Lakes”, authored by Simon Levine and Claire Chastre, with Salomé Ntububa, Jane MacAskill, Sonya LeJeune, Yuvé Guluma, James Acidri and Andrew Kirkwood, as the genesis of his own research into response analysis. The key finding of the report was that practitioners ignored information where it existed and used pre-made interventions to respond to crises. Maxwell characterised response analysis before 2005 as being irrelevant and unnecessary because food security responses were largely limited to three tools: food aid, nutritional support and provision of seeds and tools. However, since 2005 Maxwell observed four major changes in food security interventions.

1.    Food aid sourced directly from the donor country has been replaced with cash transfers used to buy food from local and regional markets which is then distributed to recipients.

2.    A shift from in kind assistance to cash and voucher assistance

3.    A greater range of livelihood programs

4.    A revolution in nutrition programming

This widening of programming options, Maxwell stated, necessitates an evaluation of the efficacy of the different responses that are available to practitioners, i.e. a response analysis. The FAO and the WFP have both produced similar studies and response analysis methodologies. Both the “Response Analysis Framework” (RAF)and the “Response Analysis Process” (RAP) is designed to assist practitioners in evaluating the needs of the recipients and then choosing which food security tools would be most suitable. These studies focused on feasibility and appropriateness. Maxwell argued that it is too simple to view these response analysis frameworks as being the link between needs assessment and program design as other factors such as organisational mandate and institutional design have a bearing on how certain tools were chosen over others.

Maxwell’s study involved interviews with approximately 600 aid workers, government and local officials, focusing on the Horn of Africa to ascertain how organisations decide which responses to employ and how evidence is incorporated into the process. Instead of finding that organisations designed responses solely on empirically-based needs assessment and causal analysis, Maxwell observed that organisational mandate and capacity and a range of other factors such the cost of compliance and risk assessment were central to what tools were chosen.

Maxwell highlighted the importance of the different levels of analysis needed to evaluate different responses. In particular, cash and market based interventions had to meet higher standards of analytical rigour, with practitioners required to demonstrate that a cash or market based intervention would not disrupt local production or significantly distort local markets. The differing burdens of analysis thus was seen to influence the likelihood of a certain tool being chosen over another.

The study divided the factors that impacted on decision making into two groups: feasibility and appropriateness. The feasibility of an intervention included considerations such as access, donor resources, government policy and previous intervention whereas factors such as staff security, safety of recipients and the risk of corruption fell under the category appropriateness of an intervention. What was largely absent was the importance of recipient preference in any intervention.

Using the results of his interviews Maxwell divided the decision making process of response analysis into three tiers. These tiers were set out as being intersectional and continually informing each other. Maxwell argued that:

·         The first order of response to a crisis is an organisation’s objectives, whether they mainly focus on providing nutrition, livelihood protection etc. As most organisations have specific mandates, much of the response is already set, regardless of the information coming from the field.

·    The second order of response, the place where most of the response analysis takes place, is the manner in which the objectives will be achieved, for example a livelihood intervention could use asset distribution to increase food security

·    And finally the third order is the program design (or response choice) and how the chosen response is implemented in a given context.

However, the dominance of 1st order - institutional power - was highlighted by Maxwell. Agencies can argue for certain programming choices regardless of the data from needs analysis; for example, some organisations supported different programming decisions based on the same data. Maxwell thus argued organisational ethos and mandate was the dominant factor in dictating the type of response.

Maxwell’s study differentiated between evidence and non-evidence based factors. While a significant amount of the considerations that influenced the decision making process were based on evidence, the majority of the factors were largely based on assumptions or perceptions. Maxwell pointed to the dominance of an organisation’s ethos which dictates how evidence is viewed and analysed. The dominance of first order choices was also shown to marginalise other factors, most notably recipient preferences and knowledge, from the decision making process.

In addition Maxwell was sceptical of much of what his interviewees saw as evidence that guides decision making, for instance organisation capacity and resource availability. He concluded by noting the need for further research grounded in the context of decision making processes and by stating we may still be “missing the point” in response analysis.

Panel Discussion

Evidence

The panel closely scrutinised the role of evidence in forming interventions. Maunder argued that while there is a gap between situational analysis and the range of responses, if evidence is used to design simple tools that had a good history of effectiveness and there was a focus on a learning cycles, default responses could provide the solution to many of the problems present in current response analysis. Barker contended that while improvements in incorporating evidence into the decision making process have occurred, there is still a need to integrate knowledge gleaned from one crisis into future responses.Agreeing with Barker, Maunder further stated that when there is a solid evidence base for a specific practice, operational and organisational barriers can shift in response, but that sharing information remains a challenge.

Echoing Maxwell’s problematisation of evidence in response analysis, Levine critiqued the way in which recipients are incorporated into response analysis. He instead suggested that we move from a needs based assessment framework and ask three questions when approaching interventions:

1.    What are people doing for themselves and what should happen to support these people?

2.    What should we collectively do as organisations/governments/institutions?

3.    What should we do as individual organisations, if anything?

Levine stressed that if organisations asked these questions, food security interventions would be more grounded in the context and not just collect data to prove the validity of specific projects.

Holistic interventions

Maunder, Levine and Barker all emphasised a holistic approach to food security. Maunder and Levine insisted that humanitarian actors need to fit into the recipients’ worlds and not build aid structures that privilege projects over recipients’ needs and strategies. Maunder advocated, where possible, for a government-led rights-based approach to food security, which would need any intervention tools to be simple and easily applied due to the difficulty large bureaucracies have in coping with complexity.

Barker addressed the relationship between development and humanitarian organisations. Stating that humanitarian problems do not necessarily require a humanitarian solution, Barker suggested that in order to solve protracted crises development and humanitarian responses should become more closely integrated. Turning to DFID’s work on resilience, Barker further contended that to solve the chronic underlying causes of crises the humanitarian and development agendas should be combined by seeking to increase resilience throughout the food system, even if this is an uncomfortable position for humanitarian organisations.

Integrating the Nutrition and Food Security Sectors

Levine and Barker discussed the failure to integrate the nutrition and food security sectors. Levine differentiated between food security and nutrition by arguing that because nutrition operates on the level of the individual and is more easily quantifiable a stronger evidence base for effective practices has emerged. In contrast food security demands a much larger analytical lens and is thus more difficult to quantify.

Barker reasoned that a need for nutrition response is often caused by an earlier development failure and by the time a nutrition response is needed, there can only be a small range of appropriate responses. Earlier responses to food insecurity would in contrast allow for a greater range of responses.

Q and A

A member of the audience# asked the panel to address the issue of aid being used to bolster governments who are actively persecuting their own population. Levine recognised that it is a difficult area to negotiate but that every intervention has a political dimension and these should be carefully judged. The moral hazard lies, Levine said, in ignoring the long term political consequences of an intervention and not factoring in the political changes that may result from a response into the decision making process.

One question asked the panel to address the difference between single and dual mandated organisation, how they differ in responding to chronic and acute food crises. Maxwell responded by noting that there are very few single mandated agencies and that most agencies are able to adapt to address both chronic and acute crises. Barker argued that development organisations areoften best placed to intervene and that the humanitarian organisations fail to address underlying development causes for chronic crises.

A speaker from the floor asked the panel to offer an organisational framework that could respond to the challenges in the paper.

Maunders highlighted the importance of improving coordination of analysis and saw opportunities for collective ways of working which allow more efficient use of aid. He also claimed that contingency planning should occur before a crisis and there is a need for a collective vision that could be orchestrated through the cluster system. 

Levine contended that greater coordination is important if it results in building a vision of what should happen but not if it merely creates more meetings. Additionally he recognised that broad based discussions are important but that there is a danger that interventions could suffer “death by coordination”.

Maxwell added to Levine’s critique of the cluster system by observing that when acute crisis are finished cluster systems tend to disappear, reducing interagency co-operation and resulting in less longer term strategic coordination. He followed on the topic of coordination, noting that there is little evidence that coordinated situational analysis produces better results and that there is almost no evidence supporting the claim that better organisational cooperation produces better overall results. Maxwellhighlighted that response analysis tools are logically based rather than empirically, because if all agencies were performing the same analysis and producing the same response, market distortions would emerge. He further observed that there does seem to be a logical claim for better co-ordination but there is a little evidence that either supports or disproves this claim.

Maxwellfollowed on by arguing that there is “no right way”to work and all decisions re context specific. In addition that the architecture of humanitarian response means that agencies they often avoid working with host governments and states. Maxwell concluded by noting that change is often driven from the top as opposed to due to better food security analysis. He pointed to how the WFP in Kenya rebuilt itself to do in kind or market based responses, leading to a reorganisation of the office, including finance and logistics departments as well as the programming.

The final question from the floor asked Maxwell to address thestrategic donor choices versus the operational decision on the ground in causing the slow response in Somalia in 2011. Maxwell responded that the slow response was less about strategic and modality choices, although differences between HQ‘s and the ground did not help, and more about fears associated with counter-terrorism legislation and insecurity. He highlighted that the lack of security in the country meant that many agencies were working remotely and were thus unable to guarantee the whole aid delivery cycle. Moreover, many agencies feared they would inadvertently violate US counter terrorism laws, leading to the WFP pulling out of Somalia.