Systems of corruption: Patterns, actors and (inter)actions

by Alan Hudson and Kathy Bain

Introduction

We have followed the work of the Corruption, Justice and Legitimacy program (CJL) on understanding and addressing overlapping challenges relating to corruption and conflict for a number of years. So, we were excited to see Peter Woodrow’s post about deploying a common patterns approach - maps of sub-systems of corruption, that can be tailored to particular contexts - which aims to fast forward the process of developing systems maps of corruption.

Informed by our engagement in related conversations about governance, anti-corruption and supporting change in complex social systems, and by our reflections on Peter’s post, in this piece we share our thinking about understanding and engaging with systems of corruption. We begin by outlining the value that a common patterns approach might bring to understanding system dynamics, before then explaining how it might be strengthened by complementary efforts to understand the actors, actions and interactions which drive the emergence of patterns in complex social systems, and why that is important.

Common systems patterns, actors, actions and interactions

As Peter’s post notes, CJL’s experience with supporting systems mapping has been mixed, with local participants, as well as CJL colleagues, finding the process challenging, time-consuming and often somewhat unsatisfying. Informed by this experience, the common systems patterns approach is intended to accelerate the process of getting to a useful systems map which can then be used to “identify possible points of intervention and subsequent program planning”. The approach aims to do this by developing and then deploying modular maps that represent the dynamics of particular sub-systems of corruption.

The common patterns developed so far relate to the following areas: systems of pervasive patronage; the diversion of government resources; corrupt patterns in procurement and contracting; and, common patterns of bribery (see figure 1, below). The idea is that these common patterns form a library of templates, which can then be deployed in various combinations, and tailored for use in and across different contexts, to better understand the contextually embedded dynamics of corruption in particular places and sectors, and inform actions that are intended to shift those dynamics and generate different patterns.

Figure 1: An example of a common systems map, from CJL

A modular and pattern-based approach to making sense of complex systems of corruption has considerable appeal as a way of accelerating the production of systems maps which can then be used to inform action. We were particularly intrigued by the potential of such an approach to provide a shared vocabulary for comparison across contexts and cases, perhaps as part of a wider infrastructure for collaborative learning about how to address the challenges of corruption and other complex social challenges. Such an approach could support both the cross-context and in-context collaborative sense-making and emergent learning that is needed to inform effective engagement in complex social systems, and strengthen synergies between learning on each of these fronts.

There is much to like about the approach to mapping patterns that CJL proposes. However, to the extent that it remains focused at the level of patterns, there are two important risks. First, that it gives insufficient attention to the ways in which those patterns emerge from the action and interactions of the actors that make up the system. And second, that it fails to address the causal dynamics, including the political economy dynamics - the distribution of power and resources, and the incentives faced by different actors - that shape the behaviour of those actors.

For instance, a map of a common systems pattern for bribery (see figure 1 above for example) that focuses attention on a self-reinforcing loop of things that happen might provide an accessible way of discussing system dynamics. However, on its own, such a map is unlikely to shed much light on the actors and actions that drive such dynamics. As a result, it may not provide the insights that are needed about causal dynamics to inform actions that might be taken to shift both the formal and informal dynamics of those systems.

The ability of a systems mapping approach to enhance understanding and inform engagement in complex social systems depends on whether it illuminates and helps people to grasp and reflect on the causal dynamics of such systems. If an approach to systems mapping is to deliver useful insights about feasible entry points and realistic ways forward, the analysis it provides must cover, in an integrated way, the emergent patterns and dynamics of the system as a whole and the parts (the actors) whose actions and interactions, shaped by patterns which have previously emerged, give the system its dynamics.?

Complex contextual causalities and the drivers of actors’ behaviour

Finding ways of addressing the causal dynamics of complex social problems is a big challenge; as big as shifting the behaviours of powerful actors who benefit from systems of corruption and the resources which flow to them as a result. However, facing up to the reality of this challenge is essential. Anything less is futile, leaving investments, initiatives and efforts to tackle corruption and address other complex social challenges stuck in the land of normative exhortations for “Good Governance”, greater transparency and accountability and less corruption which rarely lead to sustainable shifts in behaviour and outcomes.

Corruption, accountability, transparency, governance and complex social systems more broadly are fundamentally about webs of relationships. Relationships change when actors behave and interact differently. As such, approaches to address complex social challenges and support the emergence of different patterns must: be guided by hypotheses about why actors behave as they do; experiment with ways of encouraging and enabling different behaviours; and, leverage the findings of those experiments to improve the understanding of how change happens, to inform subsequent rounds of action and learning.

We do not, of course, have a simple solution to this complex meta-level challenge of how one might support behavioural changes that will lead to healthier and more constructive relationships and enhanced system dynamics. Neither do we imagine that the way forward involves external actors coming up with neat concepts or prescribing simple solutions for more local actors to apply. But by exploring this meta-level challenge in a way that centres on relationships - a way that aligns both with the relational reality of complex social systems, and our ability to shape the relationships we are in - we hope that we can encourage more focused and collaborative learning about ways of addressing complex social challenges.

The good news is that there is a wealth of practical experience across contexts and cases, and increasing interest in exploring how such challenges can be addressed, in ways that put actors, actions and interactions, and the drivers of those actions, centre stage. We know, for instance, that the common patterns proposal is part of a wider portfolio of analytical frameworks, approaches and tools for stakeholder and political economy analysis that CJL has developed in the conflict and corruption space. As such, we are keen to understand whether and how CJL program colleagues overlay or incorporate within their common systems mapping a focus on actors, behaviours, drivers and change.

There is also much to learn from the work of the SOAS-ACE Anti-Corruption Evidence research consortium. The approach that Mushtaq Khan and Pallavi Roy have developed to support the design and implementation of feasible and impactful anti-corruption strategies has huge potential for application as regards the effective implementation of public policies that might otherwise be distorted by powerful players and emergent patterns of corruption. In particular, we appreciate the focus on the ways in which actors’ behaviours in relation to specific sectoral challenges are shaped by the political economy dynamics of the contexts in which they take place, with systems patterns - and forms of corruption - emerging from the complexity of these actions and interactions.?

Systems mappings: Participants and use

Two related questions flow from our interest in approaches to understanding and addressing complex social challenges that put actors, actions and interactions at the centre of analysis. First, who should participate in that analysis, and second, how might that analysis be used.

As regards the question of participation, our view is that it is important to involve multiple actors in collaborative processes of mapping, analysis and crafting potential ways forward. Complex social systems are relational networks of actors with different positions, perspectives, resources and connections. Efforts to address complex social challenges that do not embrace this diversity will fail to grasp the essential nature of such challenges, fail to leverage the complementary contributions of different actors, and fail to nurture the radical collaboration, local relationships and collective ownership that will be needed to implement potential solutions. Well-designed participatory processes, with appropriate inclusion of relevant actors - including those with lived experience at the sharp end of the problems in question, as well as those with the power to make change happen - are, in our view, a must-have for effective implementation and impact.

We also encourage more explicit attention to the ways in which systems maps that focus on behaviours, actors and drivers are used. It seems to us that the real value of such maps comes from their use to support ongoing reflection, learning and adaptation, as actors continuously respond to each others’ actions in the context of the system’s emergent patterns. To this end, mapping processes should be designed not to produce a single static source of truth, but rather to establish a living systems map that supports collaborative sensemaking and learning, and adaptive action, by the participants in the system, on an ongoing and evolving basis.

Next steps: Actor-based mapping, collaborative learning and cross-fertilization

CJL’s experiments with developing common systems maps provide welcome food for thought, including as regards the role they might play in supporting collaborative learning across and within particular contexts. To realize that potential, and to inform effective action, it will be important - for CJL and others - to incorporate, or overlay, an analysis which focuses on the behaviours, actors and drivers that maintain and reproduce, and have the potential to shift, the dynamics of the systems in question.

In this regard, we feel that the SOAS-ACE approach has much to offer, along with other efforts to grapple with similar issues including the newly-established Governance Action Hub, Cynefin’s estuarine mapping framework, and other approaches to systemic design. We also think that investing additional time in exploring the value of frameworks which more explicitly seek to understand how system dynamics emerge from actors’ behaviours would be a very fruitful exercise. See for instance the actor-based change framework developed by Drew Koleros and colleagues, and Niki Wood’s helpful reflections on the application of that framework.

We look forward to participating in and facilitating further conversations, collaborations and cross-fertilization amongst various initiatives that are experimenting with different approaches to understanding and addressing the complex and power-laden systems dynamics that shape corruption and hinder the implementation of public policies. Such collaborative learning, bringing together people with approaches informed by political economy analysis, participatory systems mapping, design thinking and more, could support the emergence of a rich diversity of approaches to addressing the complex social, political, economic and environmental challenges that we face.

If you would like to be involved in those conversations, please do drop us a line. Or, if you know of existing fora when we might build in some purposeful discussions on this, please do let us know. We’d love to chat and find ways of collaborating and learning along together!

_______________________

This is a shorter version of a longer piece which is available here. The longer version explores in greater detail and with suggestions for further reading a number of issues including: emergent and cross-context learning (paras 8 and 9 of longer version); the SOAS-ACE approach (paras 18 and 19); participation and inclusion (para 22); and, the Actor-Based Change framework (para 28).