Gaps in the literature & Future issues

There are a few gaps in the literature on KM and learning that are of particular importance to agencies working in the international development field. In this section they are presented under the four headings responsiveness, impact on policy, impact of policy, and Southern engagement. These are issues that are relevant to development agencies in the current international development context, and which could usefully be linked to agencies' KM and learning strategies.

KM/learning and responsiveness

Viable civil society organisations (CSOs) provide a base from which the voice of the poor can be heard in decision and policymaking processes that affect their lives. However, this is only true if CSOs are able to assess and represent the situation of the poor accurately and to formulate appropriate responses. One of the factors determining the capacity of CSOs to do this is their ability to process information and use it in the most effective manner. Can improved KM and learning systems enable them to do this better, i.e. to respond to the situation of their 'beneficiaries' more accurately and effectively? Or is KM in this context an example of the self-absorbed practices that can sometimes make Northern development agencies or elite-based Southern CSOs revolve around themselves? Edwards (1994) answers this question by pointing to NGOs' democratic value base and their emphasis on openness and non-hierarchical communication channels. These inherent values, he argues, will enable NGOs to use their information systems and processes to the benefit of the grassroots communities they work with. King (2001) provides a less idealistic analysis of the situation, emphasising that KM and learning processes do not automatically or necessarily make NGOs more responsive to Southern needs. He voices concerns about the fact that Northern NGOs have so far implemented KM to alleviate their own information blockages - based on the same rationale of efficiency and profit as corporate businesses - rather than using KM to address key questions of how they can contribute to knowledge development in the South.

How is a development organisation's ability to be responsive strengthened? A useful distinction can be made here between 'step thinking' and 'web thinking' (O'Malley and O'Donoghue 2001). Step thinking can be illustrated, for example, by evaluation tied to the project cycle such as the 'before-during-after' approach (BOND 2003). 'Learning before' starts by looking at lessons from past projects; 'learning during' a project consists of continuously reviewing project objectives; and 'learning after' a project is carried out by drawing together general reflections and lessons for the future. Web thinking, on the other hand, can be illustrated by partnership learning. The mutual organisational learning that goes on (or ideally should go on) in a partnership is an example of web thinking where different elements are brought into the picture at different points, where ideas are bounced off each other, and where the aim is to try and see the broader picture, without having a step-by-step answer (Drew 2002). Any development agency will usually engage in both step thinking and web thinking as means of processing information and learning; the difference between agencies lies in the relative emphasis they give to one or the other of these approaches.

Further work is needed to examine the questions:

• Can KM/learning increase the responsiveness of Southern and Northern institutional processes to the situation of the 'beneficiaries'?
• Can KM/learning help to connect the voice of the poor with the institutional knowledge of development/civil society organisations?

KM/learning and impact on policy

Northern development NGOs are increasingly called on to do advocacy work based on evidence from the South, and to add value to policy debates both nationally and internationally. Yet experience indicates that NGO programme managers and policy officers are under perennial time and funds pressure to move quickly from concept to implementation, with less space than they would wish for undertaking comprehensive research to strengthen their evidence or undertaking analysis on how to effectively influence policy. Can improved KM and learning systems in development agencies enable them to influence policy processes more effectively? This is a key question that will need to be examined in future work on KM and learning in the international development field.

A related question concerns what type of knowledge and KM strategy will provide strategic advantage for a development organisation that wishes to strengthen its ability to influence. In theory one might distinguish between knowledge of the field (bottom-up learning) and knowledge of higher-level negotiation processes (top-down learning or centre-out learning). For development organisations it is important to have knowledge of the field in order to boost their legitimacy and influencing power (Fowler 1992). This means they have to have good information systems in place in order to process information from the field quickly and effectively - from the right people, to the right people, at the right time (Madon 2000). An emphasis on field knowledge also means that the higher levels of the organisation have to be willing to learn from staff in field offices and 'on the ground'. However, in order to have an impact on policy, agencies also need knowledge of higher-level negotiation processes, i.e. knowledge of the channels through which to influence and how to go about influencing (Keeble 2002). This is often a job for senior staff in a development NGO, or staff based at headquarters. They will communicate to field office staff what type of information they need and which channels the information should go through. In these situations learning takes place from the centre-out, i.e. field staff have to learn from the headquarters based on the past experiences of headquarter staff.

Development agencies most frequently have to display knowledge of the field as well as knowledge of negotiation and policy processes. They need to make use of both bottom-up learning and centre-out learning in their organisation. This situation brings with it much potential tension (Suzuki 1998). Some of the same tension is shared by large corporations who have offices in different countries. However, agencies in the development sector have an added tension in that they are accountable to at least two different groups of people (Edwards 1994). Business market analysts have a certain advantage over the NGO sector in that their clients are also their target group (for e.g. advertisements) and financial supporters (Roche 1998). This usually means that if they have accurate knowledge of the consumers and their needs/demands, they have a good chance of making a profit. For development organisations, on the other hand, the 'clients' are separate from the funders and the two groups require different 'advertising' strategies. They also constitute two different target groups that organisations need to influence in different ways. This creates a need for strategic knowledge in development organisations that is somewhat more split between knowledge of the field and knowledge of negotiation processes than would be the case in a corporate firm.

This is worth bearing in mind when applying ideas from much of the KM literature to international development organisations. For example, Senge (1990), based on analysis of businesses in the corporate sector, speaks of 'the learning organisation' as if the whole organisation was one harmonious entity where all sections and staff are willing to learn together. This conclusion must be modified in line with the particular tensions experienced by development organisations.

Against this background, further work is needed to examine the questions:

• Can KM/learning help to connect different types of institutional knowledge within the same organisation?
• Can KM/learning help to connect institutional knowledge and policymaking processes?
• Can KM/learning increase Southern and Northern development organisations' impact on policy?

KM/learning and impact of policy

It is usual to speak of capacity for 'research uptake' into a policy process. However, it is of little use to influence a policy process unless the policy actually has some impact. This suggests that one should also speak of the capacity for 'policy uptake': How much of a policy is taken up into practice, and how quickly does it happen? How are policies bypassed, re-interpreted, modified and sometimes drastically changed as they are implemented? Why do some policies simply evaporate? There are several factors that influence whether or not, and to what extent, policy ideas and formulations are picked up and acted on by the agents that are the official implementers of policy. The central issue in relation to KM/learning is whether improved KM and learning systems can enable development agencies to translate policy into practice more effectively.

There is increasing interest in examining these issues in the UK. In 1999 the Economic and Social Research Council established the Evidence Based Policy and Practice Initiative, a collaborative network of seven research units aiming to bring social science research closer to the decision-making process (http://www.evidencenetwork.org). One of the research units involved has produced a framework for understanding the evidence-into-practice process, emphasising, among other things, the shift from researcher-as-disseminator to practitioner-as-learner (Nutley, Walter and Davies 2002).

However, there are not very many case studies documenting what actually happens to development policies in practice. The examples that do exist, mainly within the sociological and anthropological field, only serve to highlight the need for more research on this issue. Lipsky (1980) examines what happens at the point where public policy is translated into practice in human service bureaucracies such as schools, courts and welfare agencies. He argues that policy in the end comes down to the people who actually implement it (the teachers, lawyers, social workers, etc). They are the 'street-level bureaucrats' and are able to change the planned impact of policy to a large degree. In many instances this is not an intentional action by the street-level bureaucrats, but rather a natural reaction to various pressures such as limited resources, continuous negotiation with headquarters, and relations with clients.

Another illustrative example is provided by Mosse (2002). At a recent workshop framed by concerns about how DFID could become a learning organisation, he presented a case study of policy-practice linkages in a rural development DFID project in India. He argued that in this case the policy of participatory development did not primarily serve the function of guiding action. Rather, it served the function of legitimising the action that was taken. Thus, the policy process in this project was not a process where policy was followed through in practice, but instead a matter of practice needing to be followed up by the correct policy model, in order to interpret and justify the actions that had been taken. The representations used concerning 'participatory development' served as successful marketing devices that convinced superiors, secured funds from donors, and garnered higher political support.

As these two cases show, the link between policy and practice can be tenuous, and frequently the policy-practice dynamic is played out in a way different to what the policy makers and donors had intended. There is therefore a need to learn more about the relationship between policies and practice in international development, focusing on the question of how and under what conditions practitioners take policies into account in their everyday work and in their dealing with donors.

Further work is needed to develop an understanding of the following issues:

• Can KM/learning help to connect development policy with implementation of development programmes and projects?
• Can KM/learning increase the ability of institutions to translate policy into institutional practice?
• Can KM/learning increase the ability of institutions to take practice into account in their policy models?

KM/learning and Southern engagement

Northern NGOs are called upon to develop new roles. They now have a role to play in a relationship of mutual exchange with Southern NGOs, which involves both information sharing and joint contributions to policy processes. In some situations Northern NGOs will be called on to support and strengthen the capacity of Southern CSOs to engage with national and international debates and decision-making (Keeble 2002). As Southern-based agencies improve their own capacity to produce and disseminate research, and as they gain increased access to research from Northern agencies and international networks, this can give power to engage more effectively in national and international debates and policymaking processes. As DFID's latest Research Policy Paper argues:

'The evidence suggests that the capacity of developing countries to generate, acquire, assimilate and utilise knowledge will form a crucial part of their strategies to reduce poverty.' (Surr et al, 2002)

Can improved KM and learning systems in development agencies enable them to bridge the gap between Northern and Southern development institutions more effectively, facilitating not only Northern contributions to Southern concerns, but also Southern involvement in international development debates? Or will KM prove to be a luxury that only Northern and international agencies can afford, thus widening the gap between North and South?

One way of avoiding a situation where KM primarily works to the benefit of Northern agencies while passing Southern agencies by, is to combine KM and learning concerns with an explicit focus on Southern knowledge needs and challenges. In some Southern contexts, the most obvious challenges relate to information infrastructure - including the need for increased and improved technical, financial, institutional and human resources (KFPE 2001). In the edited volume by KFPE, several means of addressing this situation - aimed at Northern agencies - are outlined. For example, Northern institutions can cultivate partnerships with Southern institutions - of which a central component can be information exchange; they can offer visiting fellowship positions in their institutions to Southern colleagues or offer to co-host conferences and workshops that bring Northern and Southern agency staff together; they can engage in more demand-driven research, in association with Southern partners; and they can specialise in institutional strengthening in specific geographic regions and/or related to specific international development themes.

Finally, then, further work is needed to examine the questions:

• Can KM/learning help to connect Southern institutions and Northern institutions/processes?
• Can KM/learning (in both Southern and Northern agencies) contribute to increased Southern engagement in international development debates?