As well as thinking about types of intervention, practitioners should pay equal attention to how they are implemented. Conventional operational frameworks may not work well in the complex world of anti-corruption. Many anti-corruption initiatives could benefit by taking a more adaptive approach.

Corruption is a complex problem

In countries where corruption is deep-rooted, addressing the problem should involve collective changes in behaviour so that all actors move away from the corrupt practice. Transitioning to a non-corrupt form of governance is a complex challenge. This is because of the sheer number and different kinds of actors involved, and the unpredictable dynamics connecting them.

Consider trying to prevent what is one of the more straightforward corruption challenges: bribery in hospitals. Many individuals and groups may be involved in bribery: patients, doctors, nurses, and administrators. They could respond to clampdowns on corruption in unpredictable ways. These networks often have a strong interest in maintaining bribery. They likely will oppose or undermine the initiative, creating disruptive tensions with those actors who are leading change. Opponents may respond to reforms by creating new opportunities for bribery — or by seeking new forms of extraction within the health sector.

This complexity is multiplied when patterns of corruption are different depending on the hospital or region. In different parts of the country, there may be distinct rules, producing diverse attitudes.

The reform itself could have unintended consequences, prompting a reduction in the quality of care as employees — angry at losing the additional income from bribes — start to avoid work. Doctors may begin to leave for abroad.

While hypothetical, the example illustrates that anti-corruption programmes, like many other programmes, face two basic dilemmas:

(1) It is difficult to know all the important information about the nature of corruption in advance. For instance, in the example above, while it may be straightforward for investigators to find out about the number of bribes, it could be far more difficult to discover how those bribes connect to working routines, resource needs or networks of employees.

(2) It is difficult to anticipate how the reform will play out as a process. For example, it is challenging for investigators to predict how those losing out due to an anti-corruption intervention may seek to undermine or avoid it.

As many have commented before, classic implementation models of ‘plan-implement-report’ pre-define knowledge flows, activities, and outputs. They are not well set up to deal with the surprises and complexities involved in tackling corruption. It is then perhaps to be expected that anti-corruption interventions may be overwhelmed by realities on the ground, undermined by political pushback, or damaged by unintended consequences.

Adaptive approaches as a form of resilience

Adaptive approaches (AA) are implementation models that go beyond linear execution by including greater flexibility and emphasis on ongoing learning. These approaches may also be known as, for example, ‘adaptive management’ or ‘flexible and adaptive programmes.’

Adaptive programme design is an extended and continuing process. The programme planning framework and details of outputs, results, and timelines continue to be formulated and revised throughout the programme lifecycle. A 2019 World Bank report suggests that such approaches are: ‘predicated on the notion that solutions to complex development problems can only emerge through implementation itself; they are essentially impossible to identify at the outset of a program.'

Adaptive approaches can bring resilience to complex anti-corruption programmes where disturbance is likely, by helping them to absorb and respond to unanticipated dynamics.

Consider one of the few anti-corruption programmes to use AA — Strengthening Action Against Corruption in Ghana (STAAC) — a five-year Department for International Development (DFID)-funded public sector anti-corruption programme started in 2016. This programme operates with an adjustable logframe that has encouraged ‘learning on the go’, as well as experimentation. In the initial phase, the programme emphasised small grants and activity-focused support, as opposed to larger and more official calls for proposals with pre-specified results and limited flexibility. You can download the annual review summary here.

A DFID- and EU-funded asset-recovery support programme — Strengthening Uganda’s Anti-Corruption Response (Sugar) — has used AAs to immerse itself in the political context. The programme, which runs 2015–20, adopted an unusually long one-year inception phase. This has enabled it to better understand Uganda’s patronage system and identify institutions that will promote change.

These two exceptions tend to prove the rule that despite billions being spent by western donors on anti-corruption and integrity-building programmes each year, hardly any funding is channelled through adaptive approaches. There is no lack of practical knowledge on how to operationalise AA — well-developed practices and processes already exist (see text box below). So why is it not used more?

In a connected blog post on the challenges to adopting AA, to be published later this week, we explore how pursuing these approaches creates a set of problems for institutions. Overcoming thee requires a parallel set of deeper shifts in leadership structures, operational processes, and incentive frameworks. Such shifts are not easy to achieve in the anti-corruption field.

You can find the second installment of this blog here: Adaptive Approaches to Anti-corruption.

Further reading: operational guides to adaptive approaches

Bond. 2016. Adaptive management: What it means for CSOs. This is an introductory paper on adaptive approaches. It is aimed at managers and leaders in civil society organisations and funders who are not already familiar with the issue. It provides insights into what AAs are, when it may be appropriate to use them, and when it may be necessary for organisations to adopt AA.

Learning Lab website. Collaborating, learning, and adapting toolkit. USAID has developed a toolkit on how to implement project cycles for effective use of adaptive approaches. The toolkit includes a section on how to get started with adaptive management and important tips.

Overseas Development Institute (ODI). 2016. Doing development differently: Who are we, what we’re doing, what we’re learning. This document is a collection of how governments and other key players are embracing and getting onboard with AA. Moreover, the document tries to identify some key actions and constraints that have to be overcome in order to ‘do development differently.’

ODI. 2016. Putting learning at the centre: Adaptive development programming in practice. This paper outlines how learning needs to be at the centre of AA development programming. It is written with practitioners and others looking for structure, specificity, and practical entry points from AA.

ODI. 2017. Putting theory into practice: How DFID is doing development differently. This paper explores some of the experiences of DFID staff in supporting efforts to put adaptive approaches into practice within DFID throughout 2016.

Oxfam. 2018. Managing to adapt: Analysing adaptive management for planning, monitoring, evaluation, and learning. This report illustrates practical steps on how to introduce AA into programme processes, and outlines the main elements and practices of PMEL (planning, monitoring, evaluation, learning) for AA.

Oxford Policy Management. 2017. How to set up and manage an adaptive programme: Lessons from the Action on Climate Today Programme. This document provides an example of what an adaptive programme can look like, using practical examples. The document is intended for practitioners interested in designing and implementing adaptive programmes. It is important to mention that all adaptive programmes are different in nature; the document should not therefore be seen as a recipe, but as an example.