The challenges of managing development differently

Pablo Yanguas is a Research Associate at the Effective States and Inclusive Development (ESID) research centre.

On April 2nd ODI hosted a group of aid practitioners and public sector researchers gathered for the purpose of discussing whether the “Doing Development Differently” (DDD) agenda can in fact be managed by development organisations. It seemed like a very necessary step after the “Doing Development Differently” Manifesto and ODI report, which have focused on changing the basic assumptions and discourse of aid, as well as providing some supporting evidence. The last panel of the day was tasked with debating whether the institutional barriers to DDD could in fact be overcome. Having worked on a similar question for political-economy analysis, I was invited to participate in this conversation, and this is what I had to say.

Institutional barriers are here to stay

The administrative and institutional barriers to innovations in the aid industry are well known: disbursement pressures, career advancement standards, corporate incentives, the illusion of control over results, technocratic organisational cultures, hostile political environments, as well as the very basic notion about what development assistance should do. ESID’s research on the use of political analysis at DFID and the World Bank has documented some of these constraints, and how hard it is to overcome them.

“Success stories” need to be taken with a grain of salt

Despite all these barriers, there is a growing body of evidence – more or less anecdotal, more or less biased – that change is in fact possible. The trick is what to make of these stories, and determining what exactly they are cases of. Are they:

  • Isolated efforts or strategic approaches to reform? For instance, the World Bank’s innovative programme in Nigeria, which is a one of a kind, as opposed to DFID’s “Smart Rules”, which resulted from a comprehensive process of internal review.
  • “Lone rangers” or organisations as a whole? In other words, are these initiatives bureaucratic or extra/infra-bureaucratic? The use of trust funds in the World Bank is an example of change happening around and in spite of core incentives, just as an energetic and enlightened head of office can turn a dormant DFID mission into a proactive developmental actor.
  • Accidental or purposeful? That is, are these cases of planned innovation, in which actors decided to do development differently, or are they cases of adaptation, in which practitioners simply had to make the best of a bad situation?

These questions are crucial for determining whether ostensible success stories are relevant, let alone replicable. Given the prevalence of personality-centred and extra-bureaucratic change, some kind of critical institutional memory is necessary.

DDD is simultaneusly more and less demanding than conventional practice

The paradox of doing development differently is that it may simultaneously be more and less ambitious than conventional practice, putting both more and less strain on the gears and engines of development.

  • In terms of management, principles like “empowered accountability” may be more demanding for the individuals involved, as professional choices become riskier for programme designers as well as for the managers who will assess them. However, a greater space for personal initiative, learning and adaptation demands less of accountability systems, which may come to rely more heavily on managerial discretion than on ridiculously nitpicky tools.
  • In terms of results, the focus on complexity and processes, rather than conventional outputs, may be less demanding for evaluation, which will not need as many proxies and unreliable indicators in the questionable quest to measure everything. However, it will be more demanding for the greater public – and taxpayers in particular – whose understanding of development has been shaped for decades by the promise of the technical fix, as opposed to the reality of incremental, contentious reform.
  • In terms of information, the embrace of uncertainty and iteration will be more demanding for practitioners, who will have to devise ways of compiling more sophisticated and ongoing data in order to properly adjust to context, for instance by spending more time in the field cultivating contacts and partnerships. The upside is that programme design then becomes less stringent and artificial in its drive to establish certainty and hold projects accountable to original expectations.

There is not enough evidence to make the case just yet, but we may very well come to realise that doing development differently is not necessarily harder than conventional practice – just different.

Change entails working at different levels

The paradoxes and ambiguities of DDD hint at the existence of different levels at which management transformations may be pursued:

  • At the high level of principals and principles, proponents of doing development differently will have to persuade ministers and directors – many of whom are not development practitioners themselves – that their organisations need to adopt a more relaxed understanding of the results chain. But principals are also subject to external pressures, and so this level of change also requires a shift in public discourse about development in general, and aid in particular, moving from a quick-fix mentality to a more contentious approach based on longer time horizons.
  • At the middle level of systems and structures, doing development differently entails changing those organisational features which inhibit – and even punish –  innovation, learning and adaptation. This goes all the way from budget preparation to evaluation standards and performance benchmarks.
  • At the low level of individuals and careers, managing a different kind of development requires new competency requirements for hiring, a system of inducements that rewards practitioners for trying new approaches, and above all a decoupling of career advancement from disbursement. In a way, DDD calls for development professionals, not simply aid administrators.

Naturally, the question arises of whether adopting a new approach to development can be promoted at any one level independently from the others, or whether it requires coordinated efforts at two or even all three levels at the same time.

In any case, underlying this entire discussion is the very rationale for development assistance as an industry:money. In particular: where would the money go? To facilitators, consultants, local brokers? Would there be any conditionality? Any targets for disbursement? How big would budgets need to be? Money is the one cross-cutting issue across all levels of change, and DDD proponents probably need answers to these questions if their efforts are to succeed.

Institutional barriers can be overcome, but not equally by all actors

With all this in mind, we can begin to sketch an analytical framework for understanding what the chances are that institutional barriers will be overcome, and where exactly that is likely to happen:

  • DEFINITELY – by individuals willing to go an extra mile despite corporate incentives and cultures, or bysmall units of like-minded people operating within larger organisations. The World Bank’s country team in Nigeria would exemplify this type, which we can call “outliers/lone rangers”.
  • PROBABLY – by small organisations like consultancies and implementers, which are by definition more nimble, or by those with longer time horizons and more relaxed disbursement imperatives, likefoundations. The work of the Asia Foundation in the Philippines would fall under this type, which we can call “explorers”.
  • POSSIBLY – by bilateral donor agencies using taxpayers’ money or multilateral agencies subject to consensus politics, as long as there is a change strategy targeting the three levels: public discourse about development, management systems and structures, and human resource policies. We can call this type of actors  “insurgents” (when they operate within agencies) and “advocates” (when they operate from outside).

At the moment we mainly see cases of “lone rangers” and “explorers”, but it would be a mistake to infer that their achievements can be easily ported to the realm of “insurgents” and “advocates”. These are fundamentally different fights, and DDD proponents may do well to cater to the different needs of these various actors instead of offering a single message. Not without a certain irony, the success of promoting change within development organisations requires DDD managers to practise the same kind of learning, adaptation and experimentation that they want those organisations to adopt in their programming.

DDD may be easier for local actors, but also more dangerous

A completely different issue is whether the doing development differently agenda – which so far has been mainly expressed in “donorspeak” – can in fact translate into the practice of local actors, both public and private. In this case, the answer will differ according to the issue that we focus on:

  • Uncertainty will be easier to manage for governments and local actors than for donors, especially in contexts of low capacity/delivery, where uncertainty and complexity are just fancy words for “politics”.
  • Time horizons are naturally longer for local actors than for external ones, especially in the case of donor personnel subject to 2-3 year rotations and aid programmes with a 3-5 year cycle, which hardly ever map onto the more protracted dynamics of local reform.
  • Incentives, in contrast, may in fact be harder for governments and civil society than donors, as local reformers and advocates remain embedded in political networks, systems of norms, and competing social demands which rarely allow them to rock the boat through innovation and change.

Doing development differently may turn out to be a feasible but dangerous exercise for many local actors.

A final word about DDD prerequisistes

In all this I cannot help but wonder about the shadow that the public sector casts over the task of pursuing development differently. Learning requires institutional capacity and adaptation requires political coverage, both things that may be missing from weak state organisations in many countries around the world.

Does DDD require an effective public sector supported by strong political will? If so, does the agenda give us any hint as to how to achieve them?

Paths to development: Is there a Bangladesh surprise?

By and

Mobile tea stall Paths to development: Is there a Bangladesh surprise?

Bangladesh’s economy has recorded remarkable economic performance in the new millennium, though its per capita income has remained low. Even more spectacular has been the steady improvement in its levels of many social development outcomes. Popular commentaries have drawn comparisons with India and Pakistan in highlighting the significance of Bangladesh’s development achievements. This phenomenon has been termed as the “Bangladesh conundrum,” and has received extensive coverage in international media outlets, such as the New York Times, the Economist and the Wall Street Journal.

But is Bangladesh’s social progress surprising, and if so, then in what respect and to what extent? To answer this question, we must systematically investigate the country’s path to development. Bangladesh’s achievements in several dimensions of social development are indeed surprising when compared to other economies at similar levels of economic development. In a research paper published last month in World Development (Asadullah, Savoia, and Mahmud 2014), we present these findings by drawing upon data on Bangladesh and over 100 other developing countries for the past 4 decades (1971–2010).

Exceptional in many ways

Using regression analysis, we document that Bangladesh has performed better compared to other countries at the same level of per capita income on a number of social development dimensions: female education, child health, and fertility.

Starting with fertility indicators, Bangladesh has, since the 1970s, managed to reverse its abnormally high record of average total births per woman—and since the 1980s, it has even outperformed countries with similar levels of income. Between 1980 and 2010, Bangladesh’s ranking for fertility data within the developing world improved rapidly compared to only modest improvements by Pakistan and India. Fertility also declined progressively because it was paralleled by an exceptional increase in contraception prevalence and we estimate that over 2006–2010, women in Bangladesh were giving birth to an average of two fewer children than in other economies at the same level of income. Between 1980 and 2010, the share of women using contraception jumped from 10% to nearly 60%, while the 2005 figures for Pakistan and India were 30% and 53%, respectively.

In terms of health outcomes, Bangladesh was among the losers in child mortality reduction in the 1970s and 1980s, but reversed this record in the 1990s and 2000s. Excess infant and under-five mortality disappeared before the 1990s, well before the country saw a large-scale reduction in poverty. The immunization rate increased from 1% in the early 1980s to over 70% within 10 years, a development described by the United Nations Children’s Fund (UNICEF) as a near miracle. Also according to our estimates, during 2006–2010, Bangladesh was immunizing 17% more children against measles than other economies at the same level of income. Similarly, gender disadvantage in primary and secondary education disappeared by the mid-1990s. Since the late 1990s, Bangladesh has outperformed other countries at a similar level of economic development in terms of female primary and secondary schooling, although it still lags behind at the tertiary level. Our estimates show that between 2006 and 2010, Bangladesh was enrolling over 7% more girls in primary education than other economies at the same level of income.

Overall, the empirical evidence shows a clear trend: that Bangladesh has steadily progressed over the past 4 decades, transforming itself from a laggard into a leader. Today, the country outperforms on multiple social development indicators, given its level of economic development. Considering its unfavorable initial conditions (e.g., devastation caused by the 1971 war and the famine of 1974) and the existing challenges of poor public governance and political instability, Bangladesh’s achievements in social development are truly surprising.

Where does the exceptionality come from?

Further econometric investigation of Bangladesh in comparison with other developing countries over a long period of time (1970–2010) helps understand why its development progress is superior to other economies at similar levels of national income.

Our results find limited evidence that such progress simply came as a result of economic growth, i.e., through income-mediated channels. We also find no evidence that development was led by public expenditure channels (i.e., driven by foreign aid or government health and education spending). On the contrary, Bangladesh’s progress is exceptional because it was achieved despite low budgetary allocations, low levels of physical inputs, and widespread poverty, and in some cases, within very short time periods. Our research rather highlights three concurrent factors that may have simultaneously been the cause of Bangladesh’s exceptional development progress.

First, development policy exploited the complementarities between public policies and nongovernment organization (NGO) initiatives. Including various nongovernment stakeholders (including religious bodies, in the case of secondary education) as part of the development strategy was instrumental to the social progress achieved, as it complemented public education and health interventions. In partnership with the government and with the support from international development and aid agencies, NGOs played a significant role in reducing fertility and child mortality through the simultaneous use of low-cost solutions and social awareness campaigns.

Second, the development strategy benefited from the synergies among the dimensions of social development. Health and education indicators improved at varying paces and different intervals creating virtuous interaction effects between different social indicators. The fertility decline began during the 1980s, when income and schooling levels were very low. This set the foundation for later progress in education and health. Equally, gender parity in schooling was triggered by the introduction of demand-side incentive schemes.

Third, long-term factors, such as geography as well as historical and cultural heritage, might have favorably affected the context of development policy in Bangladesh. Regarding the role of geography, the proximity of settlements, for instance, facilitated the easy adoption of low-cost solutions and the quick spread of good practices. Historical and cultural heritage also played a role in shaping and consolidating the elite’s political commitment to social development. Such a role has been reflected in policy sequencing, which has seen consistency across various political regimes over time. Since independence, successive governments in Bangladesh have recognized the need for controlling population growth, the importance of female education, and the role of child and maternal immunization. Similarly, the prioritizing of women and gender balance, scaling up of innovation, and focus on resilience to natural disasters have also been significant.

Moving forward

A country that was once famously dubbed “the test case for development” is today an important example for others within the developing world. The progress achieved over the last 3 decades could place Bangladesh on a path of sustained growth, eventually starting a virtuous cycle whereby higher human and social development is followed by higher growth, igniting a positive feedback loop.

However, ineffective public governance, dysfunctional institutions, and limited budgetary allocations could prove to be obstacles. As Bangladesh’s gains from low-cost solutions are reaped, further progress will increasingly depend on higher public social spending and improvement in service delivery systems. Further reductions in child and maternal mortality will require more expensive interventions and the provision of relatively costly health services. While good progress has been made in improving the school participation rate, there are now serious concerns about the quality of education. At the same time, improvements in public services delivery across social sectors will be necessary, requiring governance reforms aimed at improving mechanisms for public sector efficiency, transparency, and accountability.

To consolidate the gains made in social development so far and to make further improvements, the challenge for Bangladesh lies in addressing governance failures.


Asadullah, M. N., A. Savoia, and M. Wahiduddin. 2014. Paths to Development: Is there a Bangladesh Surprise? World Development 62 (October): 138–154.

Photo: “Mobile tea stall“. Licensed under Creative Commons Attribution 2.5 via Wikimedia Commons

This blog was originally posted at:

Training in Political Economy Analysis

by Sarah Hunt.

Over the past year training donor agency staff in Political Economy Analysis (PEA), I have found the topic inevitably means facilitating a debate. The overt aim of training is to introduce practical tools for carrying out Political Economy Analysis. But, from the outset, talking about politics in development means engaging with ideas of change – and this touches not just on the contexts where donor agencies work, but on the internal systems of donors themselves. In my experience, practitioners value the space to reflect on these issues.

Development practitioners, even outside governance portfolios, all have a grasp of how and why the politics matters for development interventions. Beginning with a reflection on political economy concepts, starts to generate a basis of shared understanding. Building a collective, systematic approach to analysis is a central aim. But delving into the ‘political’ in development provokes deep debate about the way change happens, about the dynamics of particular country and sector contexts, and about how donor agencies work.

Channelling these discussions towards an agenda for action is the principal challenge. Demonstrating the potential of PEA is critical in order to overcome the tendency for discussions of politics to become a dismal consideration of constraints to development. Introducing and explaining the range of available tools and technical guidance for carrying out PEA is helpful only to a point; practitioners want to know the value added of yet another framework to be used in programming.

The growing body of research on PEA provides useful inputs. The 2012 report on the political economy of development in Africa has a provocative message for practitioners. In the words of one training participant, it ‘shakes our assumptions’ about good governance and the use of best practice and universal blueprints in development interventions. This report also has useful messages not just on the role of incentives, collective action problems and working with the grain, but also on  the challenge to take the message about how development happens to voters and pressure groups in the North.

Shaking assumptions is not sufficient: participants want examples not just of good PEA, but also of the impact on decision-making. New case study evidence developed by ODI – presented at a recent event – provides insightful examples of how politically smart, locally led development, driven by iterative, adaptive, problem-solving analysis can meet the demands of working and thinking politically. This new research begins to address a gap in evidence and offer insights for a range of contexts and programmes. From a training perspective, finding the middle ground between acknowledging the uniqueness of each context, and promoting the applicability of PEA based on lessons from other examples, can be tricky.

There is more to do to package this type of learning for practitioners. Beyond advocating flexibility and space for experimentation, there is a need to respond to the needs of donor agencies with large programme aid budgets in rapidly changing African contexts. For example, one participant challenged me:
“Give me evidence of when a first-best option was abandoned in favour of a second-best option as the result of Political Economy Analysis”.

Presumably, he also wanted an example where this proved to be the right thing to do. Learning from the past presents concrete examples of where going against the grain simply does not work. For example, recent ESID research evaluating public sector reform in Uganda identifies the persistent difficulties in moving from form to function – and the associated political challenges. More specifically for donors, evaluations ofGeneral Budget Support in Uganda and Tanzania, highlighting reforms that worked ‘with the grain’, contrasting with those that went against the grain, and underperformed.

Systematically using this type of learning to develop politically smart programming options – which deliver results – remains outstanding. Applying PEA tools and political knowledge in workgroups reveals theunderlying tensions facing development practitioners in really ‘thinking and working’ politically.

  • Practitioners, especially in-country, are acutely aware of the political context. Yet making the leap to using collective Political Economy Analysis as a filter for programming and not just for informing context analysis and risk management remains a challenge.
  • For working with governments, moving from ‘form to function’ – away from a focus on blueprints and best practice, and towards the gritty mechanics of implementation – is risky and still largely untested. Donors cannot run government agencies; at best they can influence change processes from the outside.
  • At the point of ranking options, internal incentives and conditioning become clear. Agencies and advisors are reluctant to abandon dominant ways of working. Problem-driven approaches can help drive more creative thinking, but agencies are still limited by internal management systems and home constituency constraints.
  • Donors are also intermediaries: would they allow their subsidiaries, such as NGOs and government partners, to plan and manage their programmes in this way?
  • Finally, achieving the balance between what David Booth describes as ‘principles of what is right and doing the right things’ will depend on a number of factors beyond the specific context. As highlighted inrecent ESID research, these all relate to the political economy of the donors agency – guidance from HQ, leadership and management choices, relations with other donors.


To deal with these tensions, I believe that donors need at least to invest more in internal reflection and documentation of how PEA-type analysis already – albeit for the large part informally – guides programming options and ways of working. This has started already: using this knowledge alongside training could mean the implications of thinking and working politically, and the barriers to change, can be mapped more clearly.

Sarah Hunt is a Lecturer in International Development at the Institute for Development Policy and Management,  University of Manchester. Sarah’s research examines governance and the political economy of aid and development, with a particular focus on the politics of poverty reduction in Latin America.


This post was originally published at on June 17th 2014

The zeitgeist of political economy analysis

By Pablo Yanguas

On Monday David Hulme and I attended a half-day workshop at ODI entitled “Aiding reform: Lessons on what works, what doesn’t and why”. It was basically a gathering of scholars, consultants and practitioners working on/with political economy analysis (PEA), of the sort that takes place every few months around the UK and Europe. What David calls the “aspiring” epistemic community of PEA was well represented around the table: all the familiar faces and usual suspects from ODI, The Policy PracticeThe Asia FoundationThe IDL GroupLSEAfrica Governance Initiative, or the Developmental Leadership Program, as well as representatives from DFID and the World Bank (the latter presenting their recent volume of PEA case studies).

Under the chairmanship of Leni Wild and Marta Foresti we discussed the latest on PEA research and dissemination, as well as what’s next in the process of teaching and persuading development practitioners outside the governance silo that political analysis is inextricably linked with aid effectiveness (see Alina Rocha Menocal‘s excellent summary). It was an interesting afternoon altogether, built around a thought-provoking attempt at synthesis by veterans David Booth and Sue Unsworth. Plus our own ESID working paper on PEAgot cited by a DFID representative (he called it “scathing, but in a good way”), a gratifying sign that we are actually making a contribution to these debates.

I live-tweeted some interesting tidbits from the event, which I am copying here so that everyone can get a sense of the current zeitgeist of the PEA community (Chatham House rules apply, so no names):

  • “We have moved beyond bland references to ‘context matters’ or ‘politics matters’ and into developing a body of evidence.”
  • “It is now widely acknowledged that the puzzle of development is not a technical one but a political one.”
  • “The challenge for us now is to think how development actors can work differently, including whether donors are fit for purpose.”
  • “How will we get pro-rich governments to support pro-poor policies?”
  • “Transformative change comes from the accumulation of small victories that weaken the defences of the old order at strategic points.”
  • “It’s neither ‘one size fits all’ nor ‘every country is unique.’”
  • “Is it realistic to think that each and every aid programme can be designed in a flexible and adaptive way? If not, what’s the alternative?”
  • “We are moving on from the assumption that good analysis by itself can change practice.”

And my personal favourite:

  • “Just being able to say ‘yes, it has failed’ would create a great sense of relaxation in DFID.”


This post was originally published at: on 4 June 2014