Download this workshop report (PDF, 5 pages, 170 kB)

Workshop report

This is a summary of the key themes, points of discussion and broad consensus over the two days. The meeting was held under the Chatham House Rule in order to encourage a full and frank discussion.

Thinking Politically

Politics, or political economy, has come to the fore over recent years. All reflected that the status of TWP was varied across various
aid and development organisations (Carothers & de Gramont, 2013).
The general feeling was that aid and development organisations are much further along and more comfortable with the T than the W.
The key framing question for the CoP was: ‘Does TWP mean doing the sorts of things we do now but doing them in a different way’ or ‘does it mean we are working in such a different way that effectively it means we are doing – or will have to do – things vastly differently’? The consensus was both. Follow up questions included: What does it help us do differently? If we can’t act on it, what’s the point? The response to this will be different for each organisation.
The challenge for the CoP: need to move from a narrow formulaic, up-front analysis (PEA), to operationalisation, implementation, programming, strategy, and working politically on an informal and day-to-day basis.
There are significant organisational weaknesses/blind spots in how donors address politics. The prevailing view was that there was little overlap between high-level thinking about politics and on the ground renditions. Concerns were expressed about siloed thinking, lack of sharing knowledge and lessons, and the need for a vibrant and contested conversation.
Some discussion around a distinction between working on politics and working politically. The consensus was that there was a clear distinction between the two and that it wasn’t a fruitful distinction.
It shouldn’t be assumed that everyone is on board with TWP, even in governance cadres.

Political Economy Analysis

Recognition that Political Economy Analysis (PEA) is an important step in the right direction – though only a step – and provides an initial way of thinking politically.
Incremental improvements to PEA over last decade are heartening, but what we need is a greater step-change in how we do things.
Political Economy Analysis is / offers a useful set of tools, but there is more to politics of development than PEA. PEA needs to be brought together with other forms of political analysis, especially downstream in terms of implementation.
Danger that PEA tools have been narrowly interpreted as thinking politically. The risk is that a formal PEA becomes a comfort blanket, tick the box exercise, or a product. No use if analysis is done and put in drawer. PEA / TP needs to be converted into a continuous way of thinking and moving forward.
A big risk is outsourcing to consultants. This is inevitably leading to a small group of consultants responsible for much of the market.
But what we don’t need is a load of political scientists telling donors about incentives, rules of the game, etc. What we really need are good development specialists/technical sector specialists who can do this day-to-day.
Another concern expressed related to there being a lot of excellent PEA work, but a lot of it is not made public for fear of being too politically sensitive. The group considered whether results of analysis should be internal, shared or mixed; forming the view that the reality would always be that some PEA would always be internal and not for public arena. While this will always depends on context it does lead to problems about sharing lessons, collective learning, and iterative progress.
Two areas where it was felt that PEA still needed improvement: firstly, deeper and more disaggregated recognition and consideration about power; and, secondly, better analysis of the role of ideas: values, attitudes, norms. Ideas are crucial in terms of articulating and framing shifting interests.

From Thinking to Working Politically

Operationally TP does not necessarily lead to WP. The conclusion might be – don’t do more; for range of different reasons – context not conducive, too great risk, partners can’t help you do more. A decision tree approach can help clarify when to translate TP to WP.
What’s TWP for? Aid effectiveness, transformational change (versus transactional), redistribution of power.
Consideration was given to various ways of incorporating TWP into development work:
- a whole portfolio approach, where everything has a TWP dimension to it;
- TWP is embedded in a particular programmes, where deemed appropriate; and
- on a more ad hoc basis, operating politically in any spaces that become available, possibly under the radar.
Overall a strong view that TWP is not new – many development and aid practitioners already recognise and work in this way. Many have been working politically (consciously and successfully) all along, especially in-country staff. People are constantly asking ‘what the real politics are, below the surface?’ and are always looking for new politically nuanced ways to work.
Need to acknowledge that we all work politically already. Need to be aware of the consequences of work: TWP can be done well or badly, consciously or unconsciously.
There is an interesting difference in the way that donor staff instinctively work politically in their own organisations, understanding what will or won’t fly, but tend to be less good at using this lens when working in country, and instead sticking to the log frame.
Despite some efforts to explicitly use the concept of TWP there appears to be little awareness and varied understanding of the term in the field. That said, the notions behind TWP seem to be enthusiastically embraced.
While lots of staff do TWP –– they need more encouragement (and crucially authorisation) to work in this way. The consensus was that staff have to spend too much time meeting more bureaucratic and operational requirements, making business cases etc. rather that having time to TWP.
Discussions identified a number of TWP success stories, even in constrained delivery environment, generally where individuals have been given flexibility, and have the understanding and confidence to work outside of design and log frame constraints. A key outcome of the workshop was to establish a workstream to document these successes (and failures).
Where there are individuals who act as champions, in both donor organisations and partner countries, this tends to create space to TWP.
Some participants indicated that in certain cases, working politically was just good program design, flexibility, change, risk management?

Presenting TWP

The practice of working politically is the essence of what the CoP is interested in, but how we communicate this and the language
that we use is centrally important.
Need to be clear about what TWP is (and isn’t! i.e. regime change). Need to do so more succinctly, without jargon, and in a compelling way – so that it resonates and is clearly understood by different audiences.
Need to use the right words and convey succinctly. Terms such as “limited or open access orders”, “isomorphic mimicry”, and “problem driven iterative adaptation” are not understood by most development or partner country colleagues.
The expression of ‘TWP’ might be useful in the CoP, but consideration needs to be given to its usefulness in wider audiences. The consensus was to consider an alternative term. Suggested options included: ‘operationalising politics’, or ‘TWP for development effectiveness’, or ‘how change happens’. The CoP Secretariat will follow up and consult on this issue.
Attendees urged that more thinking was needed to consider how to frame TWP with other popular and contemporary discussions and ways of working that are having purchase. For example, complexity work (lots of synergies here), and DFID’s ‘golden thread’ (Ramalingam, 2013; Cameron, 2012).
The CoP agreed that what was needed were clear and succinct narratives of success for policy makers and aid programmers – 5-10 page papers rather than 50-100 pages.

Defining TWP

TWP is more a mindset and a way of working; it primarily focuses on understanding the interest, incentives and power of local actors, coalitions and networks; it focuses on the local institutional and structural context; it is about relationships, and brokering/facilitating local processes.
TWP is also about identifying key critical junctures (elections and electoral cycles) and having the capacity to move quickly and flexibly when the unexpected emerges (Arab Spring).
TWP recognises the importance of local power structures and ‘unusual suspects’ and working in unusual places, e.g. faith based organisations, business groups. Not least because the groups that look and sound like what we want them to may well not be connected to the sites of power.
TWP practitioners emphasise that brokering is not necessarily about getting people into the same room. In fact, bridging political divides will often mean working with groups with competing interests in separate political spaces and that in some instances these interests will not share a room together. Further to this point, reform can sometimes look more like an act of conspiracy than democracy. It’s all about working behind the scenes – ‘assembling a coalition’, ‘aligning interests’, establishing the relationships, building trust and understanding and commitment. Plus building belief in competence and understanding of mutual self-‐interest.
Discussion on whether to work with the ‘political grain’ revealed somewhat different perspectives. Some argue that it isn’t always about working with the grain. A central point being that if no one’s objecting to changes/reforms then the program was not getting to the ‘nub of the issue’. Others suggest that working with the grain does not mean not challenging powerful local interests. Rather it means identifying what is politically feasible rather than slavish addiction to technical ‘best practice’. Political feasibility does not mean not challenging the status quo.

Implications of TWP

Discussions identified a number of implications for donors and development agencies:
The need for skilled and experienced staff on the ground for a long time, lots of building relationships, deep trust and brokerage.
It’s about iteration and flexibility, less about log frames and upfront analysis. A general consensus for searching approach not planned approach.
A willingness to make an investment without knowing whether it would be successful. e.g. International IDEA’s work on cross-‐party dialogue, working with parliaments on drug money in politics. Started about 10 years ago, no indication about whether this would work, but willingness to make investment anyway, took 5- 6 years to tell that there was something that would come out of that.
A need for live feedback and constant adjustment. More frequent sense checks, shortened feedback loops and increased flexibility.
A recognition that it is less about money: in many instances money can be the enemy of a political strategy for reform. If you’re really trying to get others to expend their political capital, incentivising with money can do more harm than good.
The potential relationship between accountability and flexibility. Needs to be constantly monitored and managed. There was some discussion as to whether this is a necessary trade-off or not, i.e. whether some days require more focus on flexibility and others accountability.
Need to tap into social media and act quicker to react to more fluid social and political dynamics from the citizenry, for example recent revolutions and mass mobilisations.
Considering being more explicit about normative values and to advocate or even to take sides – and that these are two different things. There needs to be more thinking about the difference between just ‘taking sides’ (benefitting A over B) and making clear a value base of what donors are advocating for.
Need to bite the bullet to think about a zero-sum game, or negative-sum game. There are invariably political blockages to tackle before real change can take place. There are always losers in any political change process. Need to understand that and be comfortable with political contexts and have plans to confront blockers and losers.
Donors need to be better equipped if they are thinking about taking a more interventionist approach. A particular focus on networks and coalitions and considering whether to facilitate and broker locally, or just barge into someone else’s and take over. Can they be built? Can they be nudged?
There is a risk in approaching WP through a thin veneer, e.g. picking out perceived reformers because they speak a similar language to us or speak about common objectives. Are often reform-minded bureaucrats instead of politicians or political parties. And a danger of being overly-embedded. There is a danger of relying on same old networks. Risk of talking to the same people and the same circles.
Recognition of the need for much more humbleness – both in listening, but also imagining what can be achieved. Has to be locally legitimate coalitions. Which raises questions about how low-profile facilitators fit with some donors’ desire to brand and the requirements of the results agenda?

Evidence and Results for TWP

First, there are numerous examples where aid and development agencies do get it right, but more often than not by serendipity.
Importantly, there could be more sharing and lesson learning on success, so it often happens by chance or feeling the way instead of translation and application.
As mentioned earlier, there appears to be a disclosure issue. With, for example, PEA, there are problems where some reports are classified and aren’t shared and creating a barrier to a mechanism for learning and sharing success. For example, with regard Verena Fritz and Brian Levy’s (2014) new book: many are saying that the best PEA case examples are not in there.
Strong consensus from the CoP that there is a need to better document examples of WP. Need to establish a good evidence- and anecdote-base of how failing to take politics seriously leads to failure, how politically-informed [work] succeeds, and can be done. And where and why WP has failed.
Thus far an inductive approach has worked well. But now need to build a portfolio of individual cases and a broader comparative evidence base showing what A is likely to work in context X but not Y, and only when Y and Z are both present is B likely to work.
Do successful examples of TWP come from political issues, i.e. where politics is more out in the open or difficult to ignore? For example, peace processes, land, security sector reforms, working in fragile states. Or, can we have more success – in building trust, coalitions, and cooperation – on the seemingly mundane but non-controversial issues?
The donor ‘results agenda’ is potentially a crucial blockage to effective TWP (if results are conceived in narrow medical paradigm).
There was a view that good program management and portfolio design could address this problem by ensuring the TWP component (possibly the ‘spine’ or ‘platform’) of an overall portfolio could ensure that donor ‘results’ can be delivered.
It was agreed that the TWP approach needed to shape the debate based on robust evidence, better recognition of how real change occurs, and knowing recognition of what kind of changes are important. It was suggested that it was rare that transformative changes are tracked at all. This discussion raised the question regarding the nature of transformative change and the difference between institutional and policy changes.
TWP results and measurement need to be appropriate for the context and will likely require a combination of (1) patient, long term, pragmatism, and (2) short term, flexible opportunism.
Donors can often focus on the wrong results. There is also a tendency to ignore or not recognise important negative (and positive) unintended consequences of interventions.
There is also a need to measure things better in the interim between inputs and outputs: day-to-day TWP is largely a process (e.g. relationship building, facilitating, convening) and there is a need to value this process – it is also difficult to measure process.
For TWP, evaluation is important for not just measuring success, but also for ongoing tracking and adjustment. Real time evaluations and ongoing monitoring needed to help change course.
The consensus was that TWP needs combination of informed and systematic experimentation and compelling narratives. It was noted that compelling TWP anecdotes in the short term will be a useful contribution to the evidence base.
DLP presented a fruitful model of using ‘embedded’ researchers in a kind of hybrid participant observation / action research model to analyse the political economy of the relationship between the donor, modality and politically informed programs: essentially “to see how the sausage is made”. The example provided was of John Sidel’s (LSE) work on Coalitions for Change in the Philippines.
The Asia Foundation presented on the important of the role of theories of change and having an embedded researcher to constantly review and refine.
Some scepticism was raised as to whether better evidence will carry the day. For example, with Drivers of Change in DFID, back in the early 2000s, where a small committed group was able to secure senior level and political level buy in, which then permeated the system.

Who does the TWP?

Discussion ensued around who is best placed and appropriate to TWP. Who should do it? Who can do it? Consultants? Donor or development staff? Local actors?
Ensuring donors and development agencies had a good enough understanding of the local political context – to support and create the space for local actors to work – revisited the debate about PEA. It was agreed there was real value in ensuring donor staff had the capacity to undertake PEAs but these staff need the skills, time and authority
It was agreed that TWP requires a certain mindset, with the view that the TWP ‘maverick’ doesn’t tend to use theories – they ‘just do it’ by understanding the local context, assessing interests and incentives of local actors, building relationships, negotiating when opportunities arise, and building coalitions at critical points, and confronting local power structures as required.
The CoP debated the questions of whether TWP is ‘teachable’, or is more about engaging the right individuals, or more about creating the right context, or whether there is a need for having both the right people and right context. The example from DFID: 2006-7 was discussed where DFID clearly had a conducive and explicit authorising environment. Then Secretary, Hilary Benn, spoke persuasively about the relationship between politics and development in the security of a bipartisan environment around development. This was supported by a deliberate strategy to develop TWP/PEA skilled staff, initially offered as a core program for governance advisers, then also extended to non-governance advisers, and now provided to entire country programs. The core of TWP/PEA capacity is now within governance advisers, but also extended to others to avoid silos.
The alternative to TWP/PEA training is accompaniment i.e. mentoring. This approach allows staff to learn by doing, not just intensive TWP/PEA workshop immersions. Especially because there is evidence that training is well taken up and enjoyed, but when staff return to posts they invariably are buried in aid management/administration and can have difficulties in figuring out how to operationalise TWP and its relevance.
Further to training and mentoring of donor/development staff there are also recruitment implications. TWP specialists (entrepreneurs/mavericks) tend to be people who like ambiguity and uncertainty, who like ‘riding the wave’. The implication is to employ more development ‘searchers’ and fewer ‘planners’ and to better reward risk takers.
The CoP concluded there are significant advantages of external actors taking the lead on TWP rather than donor staff themselves. This was partly based on the issue of staff capacity, but much more on the issue of ‘arms length’ independence. More often than not, development ‘outsiders’ can be more appropriate where donor- partner government relationships are all very fraught, where a more neutral voice is needed, and ‘outsiders’ can say or write things that are franker and more sensitive. But it was agreed that there needed to be buy-in from donor country office and involvement of staff in changing the way they think for there to be any operational impact.
The most important point – and the lens through which all of the above should be seen – is that TWP is less about donors and much more about supporting local leaders spending their political capital to change their societies. Donors and development agencies need to ‘create the space’ for local political actors to shape local changes. Within this context it was noted that ‘transactional programs and contracts’ were inappropriate for TWP approaches and that greater autonomy and flexibility was required rather than pre-designed activities.

Mainstreaming TWP?

While it is now relatively uncontested that politics matters for development, and that political analysis is now a regular component of any country and program development and operations, there still remains some ambiguity about whether donors HQ fully authorise such an explicit approach. The CoP discussion was littered with terms such as ‘mavericks’, the need for ‘creative deviance’, of programs carving out space ‘off reservation’, of managers being ‘less than transparent’, or ‘working under radar’. It was agreed that these were not the ingredients for scaling up or mainstreaming TWP.
If donors are to TWP effectively then they have to have an eye on their own organisational incentives. There is a need to understand their own political economy, e.g. Ostrom et al.’s (2002) work on Sida and look to influence it.
In the context of key decision makers being ‘receptive’ or ‘getting’ politics and development, (e.g. Ministers, senior country staff, experienced sector specialists) many felt that country managers, country directors, heads of office etc are the key constituency in pushing the TWP agenda forwards. They often like to know there’re not alone.
It was also noted that management cannot ignore sector specialists. As such it is politically smart to create demand for TWP from the bottom up.
As mentioned throughout the workshop, language (see Presenting, above) is always central.
The CoP discussed opportunities to build a more conducive environment for TWP by identifying critical junctures and changing incentives. For example, in the context of greater integration of aid into Foreign Affairs departments there may be opportunities for less reductive long term planning and more short-term politically informed approaches. There could be opportunities to significantly reduce over- investing in front-end program preparation and design. With the large-scale international aid/development organisational changes that are occurring (e.g. World Bank, UNDP, AusAID/DFAT), and the changing geopolitics of aid, especially around China, there may be policy opportunities in the High Level Panel, post-2015 process for the TWP agenda.
It was noted on the question of whether or not it is possible, or desirable for TWP to go mainstream, that there are advantages of using more technocratic programs: these can create enabling environments for more politically smart approaches.
Finally, in reference to the public, there has been a tendency to present development to tax payers as an easy process – an overly simplistic message that aid money gets rid of poverty. Donors may have raised expectations about what they can do, but in not delivering the outcomes (MDGs) there presents a huge challenge.

Conclusion and moving forward

The CoP agreed that there was a need to convince political masters, senior officers and sector colleagues of the value of TWP. It was also agreed that the best way to approach this was not through an extensive advocacy program or a dictate from the top. Much better to create a demand for TWP from the field supported by robust evidence and a compelling narrative including politically smart anecdotes. This approach would rely heavily on CoP members working with field managers and program staff on overcoming programming and operational hurdles.
The inductive theory building to TWP should be based on a systematic approach to global comparative research and lessons focussed on key aspects of programming cycle including:
1. Various analytical approaches;
2. Selection of entry points;
3. Consideration of designs and theory of changes;
4. Consideration of modalities and instruments;
5. Effective management and implementation; and
6. Results
The CoP agreed to establish three workstreams to develop the evidence, narrative, and anecdotes to promote the TWP agenda:
1. Workstream 1: Case studies of TWP approaches that have worked/not worked;
2. Workstream 2: Portfolio of 12-15 current/future program that have explicitly been designed to TWP;
3. Workstream 3: Mentoring, training and support program for field staff to TWP.
The workstreams will be focused on building evidence and knowledge of successes and failures, and unintended consequences. Workstream 2 and the portfolio of programs will draw from the DLP methodology and approach to Philippines Coalitions for Change. The aim is to build long term action research around a number of TWP designed programs to learn and develop a knowledge base of what works and doesn’t.
The future work will also explore whether adding an analytical framework to what TWP mavericks/entrepreneurs do actively and demonstratively helps them do things better and better supports mainstreaming and scaling.
The CoP also agreed to consider whether TWP agenda needed to include looking beyond partner countries and sectors to the international, cross-boundary and transnational arenas and political issues (e.g. resource management, regional integration processes).
The CoP agreed to meet again in London in late January in association with the DLP Adrian Leftwich Memorial Conference and DLP/TAF agreed to coordinate and prepare the framework for the on-going TWP workstream.