This blog was first published on LSE Blogs Activism, Influence and Change

Conventional anti-corruption approaches — transparency mechanisms, accountability frameworks, top-down enforcement — are often ineffective, because they are based on an inadequate understanding of why actors behave as they do. A recent report on corruption in health systems points towards a more powerful approach: “regulatory landscaping”, which works with the causal dynamics of actors’ behaviours in particular contexts.

Corruption in health systems causes serious harm, squandering resources, degrading care and deepening inequity. Its costs fall hardest on those who are already most vulnerable. For millions of people, this is not an abstract governance problem. It is the difference between receiving care and going without. And sometimes, between life and death.

Conventional responses are often ineffective, because they tend to view corruption as the result of individual moral failures and a lack of accountability that can be addressed through appeals to “Good Governance” and technical fixes; better monitoring, more information, stronger sanctions. The research reviewed for this synthesis — 40 papers produced by the SOAS-ACE and GI-ACE research programmes — points in a different direction.

Spanning four thematic areas — beyond “Good Governance”, health worker absenteeism, informal payments, and pharmaceutical procurement and pricing — the papers show that much corrupt behaviour is driven by survival necessity rather than greed, and that corruption is highly heterogeneous, taking different forms, involving different actors, and driven by different causes. Understanding why actors behave as they do is the essential starting point for designing responses that can actually work.

Inspired by the work of Mushtaq Khan and Pallavi Roy at SOAS-ACE, the concept of the regulatory landscape — the rules, incentives, norms, institutions and relationships that shape how actors behave and what they can get away with — provides a framing that can inform more effective action. Corruption does not just occur on that landscape; it emerges from it. Actors’ behaviours are shaped by the formal and informal aspects of the landscape they inhabit, and they reshape it, collectively, through their behaviours over time. Addressing corruption requires reshaping the landscape; improving the soil instead of simply pulling up the weeds.

The way that patterns of corrupt behaviour emerge from and are sustained by this recursive dynamic is precisely what makes corruption self-reinforcing and resistant to top-down fixes. But the logic runs both ways. The same dynamics that sustain self-reinforcing patterns of corrupt behaviour can, under the right conditions, generate self-reinforcing patterns of rule-following.

Regulatory landscaping involves four main steps:

  • Mapping the regulatory landscape around a particular corruption challenge
  • Observing actors’ behaviours and analysing the causal drivers of those behaviours
  • Identifying actors who are already following the rules — bright spots of positive deviance — and understanding why
  • Designing policies that extend horizontal peer checking and enable rule-following by a critical mass of actors, supporting the emergence of self-reinforcing patterns of rule-following behaviour.

Two examples from the research illustrate this logic.

Health worker absenteeism in Nigeria and Bangladesh, for instance, is not the result of individual moral failures. It is a systemic phenomenon, generated by upstream governance failures — inadequate funding, political interference, poorly enforced rules — that make being absent a rational response for many health workers. Sanctioning individuals without addressing those upstream conditions consistently fails. But addressing the structural drivers of what the papers call “survival absenteeism” — improving pay, security, accommodation and career progression — can create a constituency of workers willing and able to follow the rules.

As survival-driven absenteeism is reduced, this may erode the normative cover that sustains more brazen, politically protected absenteeism, with rule-following actors reshaping the landscape of norms and expectations for others through horizontal peer monitoring. Once this dynamic is in play, traditional rulers and women’s groups, closest to the problem and most motivated to address it, can support and extend it.

The Bangladesh pharmaceutical market presents a different starting point — not survival-driven rule-breaking but rent-seeking by powerful actors. Identical medicines are sold at price differentials of up to 200%, sustained by a vicious cycle in which pharmaceutical companies illegally incentivise doctors and pharmacies to prescribe expensive brands on spurious grounds of superior quality. Information asymmetries create the vulnerability; power asymmetries sustain it.

Rather than wait in vain for the regulator to discipline powerful companies, the research identifies a different solution: quality certification. If the regulator certifies that each brand meets published efficacy and safety standards, and requires pharmaceuticals to have a stamp of approval, cheaper-brand companies gain credible grounds to challenge spurious quality claims, and have a commercial incentive to do so. The elegance of this proposal lies in the horizontal peer checking it will trigger in a competitive market where nearly 200 companies make collusion hard to organise. Even small shifts in demand would have multiplier effects: as revenue from expensive brands declines, so does the marketing budget that sustains the vicious cycle.

A regulatory landscaping approach shifts the question that reformers need to ask: not “what rule should we create?” but “what rule might incentivise the changes in behaviour that are needed to trigger self-reinforcing patterns of rule-following behaviour among a wider set of actors?”

The seriously systemic and relational move is that when some actors change their behaviour, the regulatory landscape is reshaped for others, opening up possibilities of behaviour change that did not previously exist. As the complexity practitioner Jen Briselli puts it, the task is one of “sensing where the system is malleable, where small shifts might propagate meaningful change”, and then taking action to incentivise and amplify those shifts.

The distance between analytical insight and sustained reform remains large. The evidence on the effectiveness of specific interventions is still patchy, and moving from sophisticated diagnosis to operational implementation is genuinely hard. But Corruption in health systems: Context, incentives and the political economy of reform provides a firm foundation: a powerful way of thinking about the dynamics of corruption, a coherent strategy for identifying entry points, and a set of diagnostic tools for designing reforms that support incremental progress by working with the grain of existing political economy dynamics.

The 40 research papers underpinning this synthesis report — produced by research teams in Bangladesh, Nigeria, Tanzania, Uganda and wider regional contexts — are all available in Annex A of the report with links to full papers and summaries. Three papers are highlighted here, by way of example: Paper 23 on grassroots approaches to absenteeism in Nigeria; Paper 35 on a behaviour change intervention to address gift-giving in Tanzania; and Paper 38 on quality certification as an anti-corruption tool in Bangladesh’s pharmaceutical market.