Organisational learning for aid, and learning aid organisations

29 October 2010

Although many aid agencies claim to be learning organisations, a recent review found that they still need to address some major challenges, especially at field level. Ben Ramalingam asks why this is the case, and what aid agencies can do to learn more effectively.

In their efforts to promote organisational learning, many aid agencies have embraced two influential approaches – the learning loops model of Chris Argyris and Donald Schön, and the learning organisation model of Peter Senge. Here I draw on the findings of research undertaken by the Active Learning Network for Accountability and Performance (ALNAP) and the Overseas Development Institute (ODI), which illustrate some of the problems aid organisations face in applying these approaches. Based on these findings, I suggest two reasons why learning in aid agencies has proved so problematic, and what we might be able to do about it.

Single-, double- and triple-loop learning

According to Argyris and Schön, organisational learning can be characterised in terms of a three-level evolutionary model consisting of single-, double- and triple-loop learning:

  • Single-loop learning is undertaken in line with explicit practices, policies and norms of behaviour. Learning involves detecting and correcting deviations and variances from these standards.

  • Double-loop learning involves reflection on the appropriateness of underlying practices, policies and norms. This approach addresses the basic aspects of an organisation, such that the same things are not done in response to changing contexts.

  • Triple-loop learning represents the highest form of organisational self-examination. It involves questioning the entire rationale of an organisation, and can lead to radical transformations in internal structure, culture and practices, as well as in the external context.



In most aid agencies single-loop learning happens at individual and group levels. However, evidence suggests that this is usually in an informal and ad hoc manner. In my research on knowledge and learning practices in the development sector, all the organisations studied saw value in informal learning, specifically in small acts of informal knowledge sharing and daily reflection. But there was no clear sense that such activities were actively supported by, or even related to, organisational learning strategies, even though they were generally regarded as key.

[IMAGE size=original align=left path=var/capacity/storage/images/media/images/cap-332/feature_diagram/275989-1-eng-GB/Feature_diagram.jpg]

For many aid agencies, formalised single-loop learning, as promoted in organisational learning strategies, is problematic. My research found that even in organisations where learning is central to the overall mission, systematic learning-based approaches are not widely accepted and applied. Formal learning is frequently seen as a non-essential support function – one, moreover, that is dominated by training and technology.

Consider, for example, the ‘after action review’, a popular facilitated learning process adopted from the US Army. Experiences with this simple tool suggest that it is often applied ineffectively in aid organisations. In one extreme case, the simple notion of a regular, blame-free group reflection process became a ‘lessons learned’ box in an electronic form to be filled out by individual managers at the end of a project. Such stories are not unusual. Rather than identify specific processes for organisational and group reflection, there is a tendency to point to information systems and documents as the ‘end products’ of learning initiatives, despite the widely held view that information is simply part of the overall organisational learning picture. Only a small minority of organisations I have encountered focused their efforts on human dimensions of knowledge and learning. This can lead to mistakes being repeated, time and time again.

Double-loop learning – questioning practices, norms and policies – is actually in direct conflict with the immediacy of ongoing organisational processes. Emerging cultures of learning and innovation frequently overwhelm existing cultures of compliance. In part this is due to entrenched power inequalities, meaning that mistakes cannot be admitted to those who provide resources, whether they are institutional donors, international NGOs or UN agencies. When mistakes are not admitted, lessons clearly cannot be learned.

[IMAGE size=original align=left path=var/capacity/storage/images/media/images/cap-332/feature_table/276031-1-eng-GB/Feature_table.jpg]

External relations also have particular implications for adjusting underlying norms, policies and objectives – especially between donors and implementing agencies. There is evidence to suggest that decision-making processes in aid agencies involve establishing common ‘narratives’ that fit the priorities of the agency and donor alike. From David Ellerman’s perspective, such a relationship can be seen as an example of mutually supportive ‘Official Truths’. Given the potential organisational interests of both parties in the acceptance of one shared Official Truth over another, this can, and does, lead to imperfect analysis and inappropriate responses.

Such relationships risk circularity, whereby aid problems are ‘constructed’ and ‘solved’ in ways that may bear little relation to actual needs. This makes it difficult to determine what ‘really’ works in practice, and therefore constrains double-loop learning. If mutually supporting Official Truths dominate no matter what aid organisations do, then even single-loop learning may be problematic. The risk would be that they primarily work to ensure that resource-providing relationships are not affected, and that the continuity of the organisations is not threatened.

There is some indication of a degree of triple-loop learning in aid organisations, given the frequency with which new leaders are recruited, and new strategies are launched. However, this does not appear to be particularly successful in achieving transformation. As one commentator put it, no matter what the situation, you can always predict which agencies will do what, when and how. Such predictability suggests that the deeper commitments to change called for by the concept of triple-loop learning are unlikely to be present internally within the majority of aid agencies.

The learning organisation

Building on the work of Argyris and Schön, Peter Senge outlined his vision of a learning organisation as an adaptive entity that is responsive to past errors and able to transform itself continually. To achieve this rarefied status, an organisation needs to apply five interrelated disciplines, as outlined in the figure below.

It is useful to consider these five disciplines in the context of the operational work of international agencies in the field. Effective international action is in large part dependent on the ability of operational staff to manage and implement programmes and projects. Therefore, the operational level should be where much of the learning that is crucial to the success of international action takes place, and where critical improvements are made. To test whether this is so, in 2004 ALNAP carried out a review of field-level learning among humanitarian aid agencies. The findings highlighted some of the fundamental issues these agencies need to address in applying the five disciplines of the learning organisation approach. These are summarised in the table below.

[IMAGE size=original align=center path=var/capacity/storage/images/media/images/cap-33/273998-1-eng-GB/CAP-33.jpg]

Table: Learning organisations – the ideal and the reality in the field

According to Senge, these disciplines have a ‘synergy’, such that organisational learning cannot thrive unless all five are present. Given this, and on the basis of the ALNAP findings, few international agencies can legitimately claim to be learning organisations at the operational level. This carries serious implications for the effectiveness of aid agencies. At least part of the problem is that the preferred learning mode of operational staff – which is profoundly social, and based largely on tacit knowledge – is not matched by formal learning approaches, which tend to focus on classroom training, information strategies and guidelines.

The way forward

What can we conclude from the above? Organisational learning in the aid sector is fraught with problems, whether we are talking about single-, double- or triple-loop learning. Moreover, at the operational level, where much learning that is critical to aid work should be happening, we are witnessing an inability to put in place the disciplines and capacities required to become a learning organisation. Problems exist at the aid sector level in general, at the level of individual organisations, and at the level of specific tools.

Why is there such an apparent gulf between the ideal of organisational learning and the reality of aid organisations? I suggest that there are two underlying reasons.

First, the models and approaches borrowed from other contexts have proved less than relevant, and even inappropriate, for aid work. To understand why, a comparison with the private sector may provide some insight. In the corporate sector, where many of the influential approaches to learning originated, the purpose of organisational learning is clear – to build profitability and competitive advantage in the global marketplace.

The aid sector arguably lacks clarity, coherence and consensus relative to the tight integration of the corporate mission. This lack of clarity plays out in terms of goals, objectives, responsibilities, relationships and outcomes, and at the individual, team, organisational and inter-organisational levels. It blurs the reference points and frameworks necessary for understanding and assessing effective performance, which in turn limits the scope for learning. As a result, the aid sector has been, at best, only partially successful in effectively applying the models of organisational learning from the corporate sector. Yet these ‘external truths’ have continued to play a substantial role in shaping thinking.

Given the above, I think learning initiatives could be further strengthened by paying more attention to ‘home-grown’ approaches to learning. This means that we accept that learning is not best arrived at through ‘external truths’, but through the approaches that have emerged from the experiences of people who have lived and breathed the complex realities and multiple perspectives that aid organisations face on a daily basis. It means taking greater pride and working harder to develop and disseminate those approaches to learning that have emerged from within the aid sector itself. Some of these are well established, such as participatory approaches; growing in use, such as the Most Significant Change (highlighted on page 13 of this issue) and Outcome Mapping approaches; or they are emerging, such as the framework presented by Niels Keijzer (page 14). It also means applying learning approaches to new areas such as advocacy, and identifying the new challenges that emerge (page 11). Finally, it means not applying incoming ideas blindly, but challenging their assumptions and testing their relevance, and by doing so arriving at new and more considered ways of learning to deal with development problems.

[IMAGE size=original align=left path=var/capacity/storage/images/media/images/cap-332/feature/275898-1-eng-GB/Feature.jpg]

Second, efforts to promote learning within aid organisations have underestimated the complexity of aid, leading to unrealistic expectations about what learning can achieve. As David Ellerman argues, aid organisations are attempting to address the most complex but ill-defined questions facing humanity, and in many different settings. In working towards change and improving the lives of poor people, aid agencies are dealing with huge numbers of interacting problems, factors and actors. There are inevitably degrees of non-comparability across, and unpredictability within, these complex systems.

The assumption that ideas can be transferred as ‘best practice’ from one place to another has driven much organisational learning. Rather than scanning globally and reinventing locally, as Joseph Stiglitz famously suggested, most learning initiatives in the development sector have tried to scan globally and apply locally. This ‘pipeline’ approach to learning seriously underestimates the complexity of aid work.

Therefore, best practice needs to be replaced with good principles that can provide the context for local reinvention, inspired by global learning. Some argue that this implies that aid agencies should abandon prescriptive, goal-oriented decision making and prediction about future states. This doesn’t mean a laissez faire approach to learning – quite the opposite. The most appropriate way to bring lessons from one context to another may be, as Patrick Breslin suggests, for ‘development workers to become facilitators … enabling representatives of different communities … to see first hand what in the successful project they would wish to replicate’. Another way to support local reinvention, Nour-Eddine Sellamna proposes, is for agencies to focus on ‘understanding the dynamics of change and promoting a collective learning framework through which concerned stakeholders can constantly, through dialogue, express their respective interests and reach consensus’. Home grown-approaches may prove useful here too.

In closing, I would like to quote an old Chinese saying, ‘learning is like rowing upstream: not to advance is to drop back’. The articles in this issue – grounded as they are in the complex, diverse and human realities faced by aid agencies – are useful. We may not yet see effective learning aid organisations. We may not yet be good at organisational learning for aid. But in our efforts to learn how to learn, and to engage with the complexity of learning, we are certainly moving upstream.

Further reading

ALNAP (2004) ALNAP Review of Humanitarian Action in 2003: Field Level Learning. Active Learning Network for Accountability and Performance (ALNAP).

C. Argyris and D. Schön (1978) Organizational Learning: A Theory of Action Perspective. McGraw-Hill.

P. Breslin (2004) Thinking outside Newton’s box: Metaphors for grassroots development, Grassroots Development, 25(1): 1–9.

J. Darcy and C.-A. Hoffman (2003) According to Need? Needs Assessment and Decision-making in the Humanitarian Sector, HPG Report 15, ODI.

D. Ellerman (2005) Helping People Help Themselves: From the World Bank to an Alternative Philosophy of Development Assistance. University of Michigan Press.

A. Krohwinkel-Karlsson (2007) Knowledge and Learning in Aid Organizations. SADEV Working Paper 2007:1.

C. McNamara (2005) Field Guide to Consulting and Organizational Development with Nonprofits: Collaborative and Systems Approach to Performance, Change and Learning, Authenticity Consulting.

B. Ramalingam et al. (2008) Exploring the Science of Complexity: Ideas and Implications for Development and Humanitarian Efforts, Working Paper 285, ODI.

N.-E. Sellamna (1999) Relativism in Agricultural R&D: Is Participation a Post-modern Concept? Working Paper 119, ODI.

P. Senge (1990) The Fifth Discipline: The Art and Practice of the Learning Organization. Doubleday.

Ben Ramalingam, Active Learning Network for Accountability and Performance in Humanitarian Action (ALNAP), London