A whole industry has emerged to foster humanitarian learning and, in the main, actors are increasingly collaborative and open when it comes to learning lessons from emergencies. But challenges persist, including a prevailing aversity to risk as the industry corporatizes, and a disconnect between individual and institutional, formal and non-formal, learning.
Humanitarian agencies and donors have increased their formal reporting requirements, and attempts to learn often turn into heavy databases, but this formal learning can overlook the rich seams of informal learning that have always driven much humanitarian innovation, said Luz Gómez, humanitarian planning, monitoring, evaluation, accountability and learning (MEAL) officer at NGO Intermon. “Lots of the real learning that takes place isn’t captured in annual reports or formal documents. It is exchanged in discussions held in the parking lot before and after cluster coordination meetings - or in the evening over tea or a beer.”
The question is how to share the learning from those conversations without killing the openness or trust; and spreading those messages from parking lot to parking lot, said Saul Guerrero, learning and evaluation head of the 19-NGO-strong Start Network, which is trying to revolutionize collective NGO funding and response.
While individuals learn emergency-to-emergency, as do teams, the challenge lies in higher-up institutional learning, said Gomez. “It is between the team and institutional level [of learning] that you get the most difficulty.”
To sink in across NGOs, learning must be engaging and efficient. Use video; push for webinars; encourage two-minute, not 20-minute presentations, ban power-points, were some of the suggestions made. “We don’t need more data. We need more time,” said Dayna Brown Listening Program Director at US-based learning non-profit, CDA.
Benedict Dempsey, learning specialist at Save the Children, says 70 percent of learning takes place on the job, 30 percent in formal training sessions. “To date there has been an over-emphasis on external, formal training, but now people recognize that on-the-job coaching and training is the way forward.”
The Start Network staff realized, for instance, that programme heads were often the first to spot mistakes, but did not have the time to investigate them or adjust programmes as a result. So the Start Network now asks agencies to highlight problems and appoints a third party to investigate them and find solutions.
But while discrete attempts are being made, a decreasing appetite for risk among both aid agencies and donors according to interviewees, continues to hamper agencies’ ability to admit failure. “They shouldn’t be so scared,” said Guerrero. “They need to see the sky won’t open up and swallow them - or their money - if they admit to failure. Failure is integral to our learning: failure, learn, next step, failure, learn, next step - we need to get better at that.”
Humanitarian learning must become less insular and draw lessons from outside the industry, said John Mitchell, head of the Active Learning Network for Accountability in Practice (ALNAP). He cites the Humanitarian Futures Programme as a group that has done this well: it collaborates with scientists, the military and private sector to elicit new approaches.
And learning is still too northern-centric: “It’s still difficult for northern-based agencies to learn from southern ones,” said Mitchell. “This neglect of national staff and local actors is a common theme in our work.”
Finally, complex learning - what Mitchell terms “triple-loop learning” which involves questioning the rationale of an aid approach or an organization’s ethos, like the shift from food aid to cash vouchers, which is currently under way - takes years, sometimes decades, and we must be patient.
(Single-loop learning involves sharing information about what agencies do; and double-loop: redesigning the way they do things based on that learning).
IRIN spoke to practitioners and learning experts to highlight some innovative initiatives:
Learning from mistakes
Some agencies or individuals are daring to be more open about mistakes. Action Against Hunger (ACF) publishes an annual learning review where it discusses mistakes and lessons. Swedish aid agency SIDA even used humour in a report on the abuse of the log frameworks.
Such initiatives are refreshing, said Francois Grunewald, head of French humanitarian think tank Groupe URD: “It’s not a problem to share a mistake, it’s a problem not to share it,” he pointed out.
Confidential talk shops can also be hugely beneficial: UK-based NGO network the Disasters Emergency Committee often does this following a collective appeal and response. “We need a safe environment to share what is not working,” said Dayna Brown of CDA. This works best if facilitated by an outside organization with no stake in the findings, said one observer.
Humanitarians can take a cue from outside the sector, said Gomez, citing website Admittingfailure.com, which was started by a US citizen who wanted to start a community to create a more transparent, collaborative approach to development. Examples of failures include corruption in a youth soccer initiative, and money drying up in a scholarship programme that caused students to drop out of their schools.
Donors do not give agencies enough space, or put enough pressure on them to change, said interviewees. “Donors want value for money, but we want intelligence for money,” said Grunewald. So in some cases NGOs are taking charge. The Start Network puts aside 1 percent of the donor funding each agency receives for flexible learning support; and will link future funding disbursements to what an NGO did as a result of learning about a mistake. “We wanted to ask: how can your performance and your ability to be transparent get captured and be reflected in future funding?” said Guerrero. After all, the key to learning is making a stronger link between the generation of materials [reports, best practice] and how they are used, said Brown.
Communities of practice
NGOs are increasingly gathering learning networks, in some cases “hubs”, around a theme - say Cash in emergencies (CaLP), Communicating with disaster-affected communities (CDAC) or the Digital Humanitarian Network. ALNAP also sets up communities of practice around different themes (urban disaster response, communicating with communities) which are starting to have a significant impact on the sector say observers.
Increasingly regional networks such as the Asian Disaster Reduction and Response Network are laying emphasis on inter-NGO learning as a primary goal. Online learning opportunities bring opportunities to their members that were out of reach a decade ago. Such networks may be the norm in the future, said Guerrero, “so we need to get much better at capturing lessons learned.” Start’s primary challenge in his view is: “what can you really achieve when you put 19 agencies to work together in an emergency? How can we really be greater than the sum of our parts?”
Local partners overlooked?
The aid sector is getting more complex, and given that, there will be “more missed connections and disconnections”, said ALNAP’s Mitchell. “The learning challenge will only get bigger and more complex.”
Missed connections apply to national and local partners: they are too often overlooked in training programmes, and of the thousands of pages of best practice, evaluations, standards and tools, just a fraction are translated into local languages, despite 90 percent of humanitarians being local staff. “It doesn’t add up,” said Gomez. Groupe URD translates all of the materials on its Haiti learning hub into Creole and focuses its Haiti training towards national government and NGO staff, but “more needs to be done,” admitted Grunewald.
Groupe URD was one of the first organizations to launch real-time evaluations - in Central America in 1994. Since then, the sector has come a long way: evaluations are increasingly shared (ALNAP has at least 2,000 evaluations in its database, estimates Mitchell); they are often led by independent parties such as the Emergency Capacity Building Project, and lessons from them are often synthesized into meta-evaluations or quick lessons-learned crib sheets. ALNAP’s document, synthesizing 30 years of lessons learned from earthquake response, was downloaded 2,500 times in four hours following the 2010 Haiti earthquake.
Evaluations are also going further. In its learning “observatories” in Chad, Afghanistan and Haiti, Groupe URD revisits projects every few months to evaluate whether or not changes have been made to previous evaluation findings; runs seminars to discuss the results; researches initiatives to look into solutions if they are unclear, and training to address gaps.
The Humanitarian and Leadership Academy (not yet running as it still awaits funding) plans to bridge the gap between the swathe of humanitarian materials available online and the lack of learning, development, capacity-building and expertise in the sector, by sifting through it all and separating the good from the bad. “There is a lot of material available but it is hard to judge what is relevant, what is good quality,” said Benedict Dempsey, head of Knowledge and Evidence at the Academy. The Academy, which hopes to have 10 centres across the globe, will also translate materials into local languages - on demand if needed. “If enough practitioners are asking for building regulation guidance in Bahasa Indonesian, we’ll translate them,” said Dempsey.