Anti-corruption needs a radical rethink. After decades of effort, the massive costs of corruption continue to harm many countries, and corruption appears to be increasing in some. Even worse, anti-corruption efforts have often been corrupted, with anti-corruption and enforcement agencies extracting from citizens or using their powers to harass and pick up the opposition. Why have anti-corruption efforts not delivered stronger results, particularly when they seem to work in some countries? We argue:

  1. Evidence of violations is only likely to trigger action for accountability if there are actors with the power, capabilities and interests to make sure this happens. Anti-corruption is costly and takes on powerful interests, so it is only likely to succeed if sufficiently powerful interests want to fight a specific type of corruption.
  2. To make anti-corruption ‘real’ we have to look for opportunities where actors have, or are likely to have, the power, capabilities and interests to act against particular types of corruption. Anti-corruption can be effective if it supports and strengthens these activities, in addition to supporting transparency and accountability. We call this the Power-Capabilities-Interests (or PCI) Approach to Anti-Corruption.
  3. Transparency and accountability measures are effective on their own in some countries, but only because a large number of organizations already exist across these societies with the power, capabilities and interests to ensure that rules are enforced. These are countries that already have, or are close to, a ‘rule of law’. Here, when transparency improvements reveal violations, interested actors are likely to take action to ensure that violators are punished, and accountability processes are therefore likely to work. But a rule of law is rare. Anti-corruption has to be designed to be effective in the general case where there is at best a rule by law: rules are enforced in some areas, but violations and informality persist in many others.
  4. In the general case, where the existing configuration of power and capabilities does not support a rule of law, anti-corruption immediately targeting all types of corruption is likely to yield limited or even negative results. Failures in many areas can undermine the effort as a whole. Instead, anti-corruption should focus on areas where it has feasibility and impact. Feasibility means the strategy is implementable, and this is only likely in these contexts if we can identify actors connected to that activity who can and will ensure its implementation. Impact means that the reduction of that corruption is not just useful for these actors but has wider social benefits. Successful anti-corruption of this type can create the conditions for anti-corruption in other areas.
  5. The most promising areas of feasible and high-impact anti-corruption are usually to be found at the sectoral level, where developmental policies and service delivery are often affected by damaging variants of corruption. These are often also areas where we are likely to find actors with the power and capabilities to support the enforcement of rules in their own interest, particularly if we can identify feasible policies to strengthen and support their activities.
  6. The political settlements framework shows that the implementation of any policy depends not only on the power and capabilities of the actors directly involved as enforcers (principals) and potential violators (agents), but also of other actors they have transactions with. The activities and responses of these ‘horizontal actors’ can change the calculations of both principals and agents. When self-interested horizontal actors act against principals and agents who violate rules, transparency and accountability measures become effective. We describe such horizontal relationships as effective horizontal checks. However, horizontal relationships can also support collusion and prevent enforcement, so it is important to identify the types of relationships that dominate in specific sectoral contexts.
  7. Our research shows there are many opportunities for enhancing and creating effective horizontal checks in sectors where anti-corruption can also improve development and welfare. The latter is important because anti-corruption cannot just be about enforcing rules without investigating what the rules are. The rules must support developmental outcomes and actors in that activity must be able to follow these rules. In contexts where actors have very different capabilities, this condition is often not met. As a result, many actors involved in an activity may be violating rules, but for different reasons. Some may be deliberately violating rules, despite being able to follow them, because they are extractors (thieves) or free riders. But others may be violating rules because they cannot follow these rules even if they try. The second type of violation is quite different and is happening for ‘reasonable’ reasons. Unless we address this, the true free riders and extractors cannot be identified and isolated.
  8. We identify three anti-corruption strategies meeting these conditions, based on the presence and strength of already existing horizontal checks:
    1. Enhancing effective checks is a feasible strategy when some actors are already checking violations in ways that support development. Here anti-corruption should support and enhance these activities with policy, together with continuing support for transparency and accountability measures.
    2. Creating effective checks is necessary when effective checks are not already in evidence, but where the numbers of rule-followers can be feasibly increased by addressing reasonable reasons for some violations. This directly reduces corruption but more importantly, allows the emergence of the horizontal checks that can then be supported by the first strategy. Finally,
    3. Mitigation and transformation strategies are necessary when neither of the first two are immediately feasible. This is the case in activities where many actors have low productive capabilities and corrupt activities are the only feasible way of making a decent living. Think of poor people engaging in poppy farming. Horizontal actors, both rich and poor, are likely to collude in networked corruption making anti-corruption very difficult. If anti-corruption is supposed to help the poor and vulnerable, a longer-term strategy is required to first develop the capabilities and opportunities necessary for engaging in other activities, while mitigating the immediate negative effects of the corruption. Only then can the second, and eventually the first strategy, become feasible. We illustrate these strategies with findings from the SOAS-ACE research programme.
  9. Anti-corruption cannot therefore only be isolated transparency and accountability pillars. Instead, anti-corruption has to be built into the design of all policies, programmes and institutions that we are targeting. An effective policy must itself trigger activity by interested parties of sufficient power and capabilities to ensure that the relevant rules are enforced. When this is the case, our evidence shows existing transparency and accountability processes work quite well.
Read full text