Between 12–13 April, U4 TRIAL switches on its co-creative process together with a wide range of motivated stakeholders to get this new initiative off the ground. In this blog post, Fredrik Eriksson explains some of the initial thinking influencing the work with U4 TRIAL.
The results of implementing the common anti-corruption agenda of policies and practices have so far not matched expectations in terms of effects. Over the last couple of years, critical research has pointed out weak results, policies and practices, measurements, as well as a poor understanding and precision when addressing corruption in different contexts. The latest critique even points to an irreconcilable mismatch between governments’ mandates in their formal systems of authority and their actual power to achieve change.
While the latter ctitique concerns an existential crisis of the system of nation states beyond the reach of anti-corruption, other critique can be responded to. The U4 Anti-Corruption Resource Centre at CMI has created U4 TRIAL as a way forward. The mission of U4 TRIAL is to facilitate the development of new approaches for addressing corruption and testing them in specific contexts. For reasons explained below, we do not think all solutions have already been thought of, adopted and tested. Also, there are a wide variety of problems outside of the type of measures intended to address corruption. How anti-corruption measures currently are being developed as well as implemented may also explain the poor results of the current version of the anti-corruption agenda.
Local Innovation Projects
The main instrument of U4 TRIAL to help advance that knowledge are Local Innovation Projects (LIPs). Through LIPs, U4 TRIAL will facilitate the development of new approaches to address corruption that can be tested in specific contexts. We do so by inviting stakeholders to collectively develop solutions that fit the local characteristics of a corruption problem. The LIP model process can be used by many different actors who on their own are unable to develop and test new approaches. This means U4 TRIAL helps to lower the transaction costs and risks for developing new approaches and learn how they work.
By adopting a double learning loop to U4 TRIAL’s work, we expect this thinking to develop as we gain experience and learn from its implementation.
What’s at stake?
Speed is important. The big global challenges of our time all require efficient and effective governance to be addressed. Those challenges are no longer in a distant future; they are already here. And this is why the urgency of addressing corruption has never been greater; the required governance capacities are undermined by corruption. In fact, it is impossible to imagine mitigation and adaptation to the many different consequences of climate change without such governance capacities. The effects of poor governance are not contained within the boundaries of nation states, but a great deal are. As long as the international community heeds the principle of Westphalian sovereignty, governance within a territory is primarily a national responsibility. At the same time, the current nation-state system as we know it may have outlived its benefits by proving its inability to deal with the big global challenges of today.
Besides understanding anti-corruption as an urgent necessity to address the big global challenges of our time, it is also a main concern of the global youth. According to the The Global Shapers Annual Survey 2017 capturing the views and values worldwide of young people, corruption is not only a top national issue but also perceived to be the top driver of inequality. Young people also value integrity and honesty in leaders and employers as the most important characteristsics. Still, it is hard to see how that is taken seriously given how secrecy is still both effectively enabled and accepted by many nations and corporate actors. While the need and desire for addressing corruption are strong, not all obstacles to progress can be explained by malign powerful interests blocking anti-corruption measures to protect their interests.
The many problems with anti-corruption
Even if there was a sudden global willingness to prioritise the reduction of corruption, it does not mean we know how to do it effectively. There are a multitude of potential explanations to current weak results: Perhaps it is because controlling corruption through the common approaches requires longer time to have effect. Or perhaps it is because they are simply ineffective also in the long run and cannot achieve any control of corruption. Other explanations refer to the need to appreciate the political barriers to change. The Thinking and Working Politically community makes a strong point by referring to that change involves the renegotiation of power and resources. It creates winners and losers, so there will always be people or groups who want to keep the status quo, and those who will welcome change because they stand to gain from it. That argument makes anti-corruption a political endeavour.
But there are many other alternative explanations for weak results. The method to develop anti-corruption interventions as well as how they are implemented also matter. An intervention that lacks legitimacy in the eyes of those who are to implement it is off to a bad start. When they also lack both understanding of why a certain change of practice is needed, and the actual performance incentives do not support any such change, the chances of success are slim.
During implementation, change is complex and often unpredictable. It is next to impossible to know how an anti-corruption intervention will unfold at the outset due to changes it brings in behaviour, incentives and interactions. One step forward may be followed by two step backwards, only to be followed by a new step forward, and so forth. A ‘political’ process of change should not be expected to be linear. Yet, it is only recently that the emergent learing from implementing interventions has been recognised by intervention funders as valuable to secure effectiveness and responsiveness to reality. So far, that adaptation to emergent learning has not been well accommodated for reasons that to some appear to be structural.
Others refer to a misunderstanding of the problem of corruption. For that reason, the current general solutions simply do not fit the problem. Rather than being a simple problem of individual exceptional deviance from a societal norm of no corruption, the problem can often be understood as systemic and complex.
Similarly, some claim a poor reading of the actual governance contexts in various countries leads to meaningless interventions. Where the required social norms in support of the proper use of entrusted authority have not yet been developed, there is no reason to believe formal institutions will be effective in addressing them. Some have referred to this as seeking to stop corruption relying on the corrupt.
Directly linking to this is the very recent discovery that there actually is no clear microfoundation for the rule of law, i.e. whatit is that is required for a society to establish rule of law when there is none. In other words, understanding is emerging for how to achieve a wanted societal change through legal prescriptions in different contexts.
Smaller social worlds and external validity of research
The strong call for taking context seriously also influences the validity of research outside of a specific context. The same policy can have very different effects in different populations. Similarly, measures that have been proven effective in small trials are not always equally effective in larger interventions, even in the same country, as different actors respond differently. This points to a very central, but often neglected fact: corruption consists of unwanted human behaviour in certain situations. To understand unwanted human behaviour in order to figure out the factors and dynamics that influence it, we need to have fairly specific theories of actors in speific settings. Many of the anti-corruption approaches of today do not build on any such differentiation, nor that there is any need to pay attention to smaller social worlds.
If we look at the crisis in economics after the onset of the latest global financial crisis, we have seen a scramble to dig deeper into the assumptions of rational human behaviour. Some stress how social actions are shaped and influenced by the context, while others stress the specific characteristics of actors (preferences, perception, and cognition) shaped by contexts. Sociology, the scientific study of society, which includes patterns of social relationships, social interaction, and culture, has also seen a shift. The importance of context and the social relationships and processes within it has consequences for anti-corruption.
Subsequently, as research has pointed out many of the crude assumptions in anti-corruption, this means we have a range of new opportunities to explore. The poor results of the global anti-corruption agenda as well as the critique in and of research tell us that we need to innovate new approaches to progress.
Developing and testing new approaches
Currently, the dominant approaches to developing and testing anti-corruption interventions represent considerable obstacles both in terms of risks and costs. Firstly, many policy recommendations based on empirical evidence have never been tested in specific contexts. The external validity -or replicability of studied successes- is basically unknown, although research can be clear about the mechanisms for change and the contextual factors and dynamics present where those successes have occurred. Should they be similar, the likelihood of replicated success may be quite OK. Otherwise, chosing to implement a new research-based policy recommendation represents a high risk.
Secondly, the approach to testing something new mostly occurs through pilot or research projects. In pilot projects, the new approach is tested as a full-scale stand alone project measured in the same way as most donor-funded interventions are currently measured. This means that the eventual results in terms of effectiveness will emerge after the termination of the project, which can be between three to five years, and with a final evaluation of impact perhaps one or two years after termination. This assumes that the expected effect of the intervention is reasonably identified so that the measurement actually captures any eventual effects. These pilots can of course consist of Randomised Control Trials (RCTs) to make sure effects are robustly assessed.
The pilot approach to developing and testing something new represents quite a high risk, high costs and slow emergence of results to learn from. It is an approach that differs considerably from how service and product innovation occurs through the so called ‘lean methodology’ commonly used in the private sector. It consists of coming up with an idea -a hypothesis- about how to solve a problem. This is followed by developing a most minimally viable way to test that hypothesis; a prototype with a test methodology able to measure effects in a reliable way. The lean approach to work builds on emergent learning from the feedback of data that comes from the testing, followed by adjustments to the tested prototype. The testing is conducted in iterations until there is a solution that really works. should several iterations prove it does not work, the prototype is abandoned.
The benefits of this approach is that it allows for limiting costs by investing as little as possible in early, often faulty attempts at solving problems. By assuming the approach of failing as quickly and as cheaply as possible, effective solutions can be found more quickly and cost effectively.
The potential of using this in anti-corruption is made possible by new methods to define, mapp and understand the problem of corruption, including how it is maintained in specific contexts. Instead of only seeking to evaluate the effectiveness of anti-corruption intervention by measuring the effects on acts of corruption (or the proxies thereof), these more granular understandings of corruption can help identify not only influencing factors and actors, but also dynamics and dependencies to corrupt practices. To learn about the effectiveness of an anti-corruption approach that seeks to disrupt or remove nodes or components of such systems could focus on measuring changes to the manifestations of them. If it does not work, new knowledge has been gained and potential losses from failed full-scale interventions averted while making available funds elsewhere. If it does work, the new knowledge can be used by others for scaling an intervention using the same tested approach.
Diversity of stakeholders
While reviewing several strands of critique agains anti-corruption and looking into the experiences and methodologies of social innovation labs, we quickly realised that what we need to do we cannot do on our own. Collaboration between a diversity of stakeholders is therefore a crucial component of U4 TRIAL.
We also learnt during our background research that collaboration is deeply influenced by stakeholders’ different objetives, motivations and obstacles to engage. Firstly, to choose to only work with those stakeholders that have certain types of motivations to find new effective ways to address corruption is counterproductive to achieving our overall objective: to advance the emergence of knowledge of effective anti-corruption practices and policies in specific contexts. Secondly, from our background research it was clear that different actors have very different obstacles to engage in the development and testing of new approaches. For some actors, costs are a major obstacle to the development process, including testing. To other actors, the political risks of failure prevent trying something new.
A minister or a head of agency/general director deciding to prioritise costly reforms to reduce corruption takes a considerable political and reputational risk. To superiors, it may be perceived a threat to vested interests. The same applies to staff, which may respond by cutting back performance to provoke a change at the top. In relation to the public, should the reforms not work, there is a risk of losing legitimacy and public support. Worse, it may result in disillusionment and loss of hope for improvement.
For a Chief Compliance Officer (CCO) at a company the situation is very different. There may be strong pressures for compliance, which requires that corruption risks are managed to be kept at tolerable levels. But some risks in a certain market may be difficult to manage without the collaboration of others. The CCO may see an opportunity to address an intolerably large risk in a market segment but struggles to secure the funding to try it out. The CEO may think the idea is too costly to test, requires considerable external expertise, lies too far from the corporate mission and the risk of failure is considerable.
Other actors have structural constraints making it impossible for them to detract from formal requirements as well as to fund non-established approaches generally acceptable. Consider a formal financial regulation that stipulates the requirement that the larger the risks to results, the greater the need for risk management to secure value for money. Similarly, as explained by a representative of a multilateral development bank during a recent workshop on rethinking anti-corruption, there is no scope for funding new approaches unless there already exists evidence of its value. Such regulations and policies effectively preempt any collaboration to develop and test new approaches to anti-corruption.
Then there is the issue of different performance incentives of different actors. Different operational objectives lead to the establishment of different performance incentives. To ask staff of an organisation to ignore their performance incentives and work on matters that do not count and put them in a worse position will not work. Also, different operational objectives lead to different understandings of what is important and not.
For example, for a political activist to wait for empirical evidence from academic research fits poorly with that role being deeply involved in seeking to influence the rapid and communication intensive movements of local politics. That does not mean that research cannot inform good political strategies. It simply means that collaboration that makes effective use of different capacities and resources of different actors needs to take into account their different objectives, motivations and obstacles. Asking them to change in a way that undermines their motivations or objectives will not help achieve good collaboration.
To find a way that makes effective collaboration possible is a central task of U4 TRIAL. Assuming the role of leading coordination and facilitation requires specific expertise. Effective stakeholder collaboration is often complex, but much has been learnt about how to do it well. That expertise will be required to help U4 TRIAL bring together a wide range of stakeholders under the same overall objective: to advance the emergence of knowledge of effective anti-corruption practices and policies in specific contexts.