It’s an interesting time to consider the status of corruption research, given the current level of discussion and criticism of existing research approaches. Although the volume of research has increased exponentially over the last 20 years, many of the approaches to combating corruption have yielded few documented results and we are yet to find ‘THE recipe for fighting corruption’.

Corruption in the sense of misuse of an entrusted position for your own ends has always existed, ever since people have been entrusted special responsibilities. Ever since that time, attempts have also been made to control it. But the subject of combating corruption in development cooperation only emerged after the end of the Cold War.

Since the early/mid 90s, corruption has no longer been a taboo, largely thanks to Transparency International ‘s Corruption Perceptions Index and other international opinion and experience indexes which have put the topic on the agenda of international politics. However, these indexes are poorly suited to measure progress in combating corruption. The definition is too broad and it is not sector-specific. The methodology is unsuited for comparison over time, which to some extent is responsible for the frustration after the first 10 years of intensified development cooperation on this issue: the Scandinavians and New Zealanders still lead these indexes, while countries in the South and East occupy the lower ranks, and at the bottom lie conflict states.That’s overall. There have been small improvements here and there, but overall little has changed.

In fact we are probably all aware of initiatives, particularly at the local level, that have led to at least a temporary increase in transparency and involvement of the community and user groups in improving service provision. Examples can be found in project progress and evaluation reports and occasionally in academic case studies. However, it often proves to be difficult to replicate these experiences in other contexts. And it’s precisely that which is currently the focus of increased discussion among academics: What works in which context and what constitutes the context anyway?

Paul Heywood, Professor at the University of Nottingham, published a very useful review of the current discussions on what kind of corruption research we need. His study has the thought-provoking title Rethinking corruption: hocus-pocus, locus and focus.

  • Hocus-pocus refers to the fact that we don’t have a magic formula. Explanations of concepts around corruption (and then combating them), such as Principal-Agent Theory Robert Klitgaard’s formula:
Corruption = monopoly position + discretion - accountability

This apprach has influenced the approaches of the World Bank, the International Monetary Fund and other international organisations. In practice the problem is that it’s not only the agents who are corrupt, but also the principals and even newly established regulatory authorities if they are not well enough insulated from existing corrupt practices.

Other academics have formulated it as a Collective Action Problem, in which one continues to act corruptly as long as others are corrupt. A lot of paper has been covered with concepts and typologies of corruption, although they haven’t led to any real practice-relevant recommendations, because the national state is often in the foreground, while the increasingly global character of corruption networks and informal patronage systems at the local level are neglected. Heywood therefore proposes aligning the locus and focus in research:

  • Locus: Indexes have led to an understanding of corruption as a global problem. In fact there has been very little research into areas such as the globalisation of corruption, illicit finance flows, transnational networks and how they influence national politics and economics. We need more research in these areas. Others, such as Professor Mustaq Khan at the School of Oriental and African Studies in London, see the locus primarily at the subnational level.
  • Focus: sector-specific analyses: what are the predominant types of corruption, how is it organised, who is involved, who tolerates it, what are the interdependencies with other sectors and the social environment?
We need theory-guided, empirical and systematic comparative research across disciplines.

The trend towards this began about 15 years ago:

  • In the USA: J-PAL (Abdul Lateef Jameel Poverty Action Lab) am MIT which has conducted over 80 field experiments since 2003, and Innovations for Successful Societies at Princeton University with accessible case studies.

  • U4 Anti-Corruption Resource Centre is an initiative by the Utstein group (the four development ministers from Germany, Norway, the UK and Netherlands) which was founded in 2002 at the Michelsen Institute in Bergen. U4 currently has eight bilateral partners for applied research on corruption in development cooperation to identify and implement effective measures and as far as possible to develop a practical guidelines and policy recommendations. Training courses and workshops bring us into direct contact with staff from our partner organisations and also with their civil society and state partners. In cooperation with TI we also offer a Helpdesk for urgent questions. We are currently working in consultation with our partners on the following main areas: general corruption in development cooperation and coordination of donor countries; natural resources and corruption, the private sector, international finance flows, sectors (in particular: education, health and justice) and now also on informal institutions and networks, as well as migration and corruption.

  • EU Antikorrp (Anticorruption Policies Revisited: Global Trends and European Responses to the Challenge of Corruption) is a big research project financed by the 7th framework programme of the European Commission. The first phase, from 2012 to 2017, was just completed. The main objective is to research the factors that support or impede effective anti-corruption measures. The project involves 25 research teams in 15 EU member states.

  • In 2015, the British Academy and the UK’s Department for International Development spent 3.6 million pounds on establishing the Anti-Corruption Evidence (ACE) initiative which supports eight leading international research teams to study successful approaches to combating corruption in developing countries.

In 2014, the DFID commissioned the Overseas Development Institute (ODI) and U4 to conduct a meta-study on corruption research. The study Why corruption matters: understanding causes, effects and how to address them systematically surveyed the literature on the causes and consequences of corruption and the effectiveness of certain anti-corruption measures (i.e., not only the measures by donors, but also initiatives by governments and the civil society). The availability of sources (volumes and methodological quality) on certain tools was put in relation to the effectiveness of these tools. For example, there are many studies on whether decentralisation can reduce corruption, but the results are contradictory and inconsistent.

Anti-corruption measures are most effective when they are supported by other factors/measures and integrated in a reform package.


In the following I list five exemplary points for which we have a relatively consistent body of evidence:

  1. „Corruption is bad“: it has a negative impact on existing inequality and on fundamental services, domestic investment, tax revenue and the environment; but there is no evidence on the negative effect on macroeconomic growth.

  2. Public Financial Management (PFM) reduces corruption. There is plenty of evidence for public expenditure tracking measures in particular, espeically when the civil society is involved.

  3. On the subject of involvement of the civil society, in particular community monitoring, transparency and access to information, there are relatively numerous studies, but most of these initiatives and the reporting are focused on improving services rather than reducing corruption. There are certainly some positive experiences in this area, particularly when it involves cooperation with responsive government institutions.

  4. Laws on transparency and access to information alone are not enough; but transparency and a free press can have a positive effect on other measures.

  5. Anti-corruption measures are most effective when they are supported by other factors/measures and integrated in a reform package.

Now you’ll perhaps be wondering: I knew that already, but what exactly is a functional reform package? And if there is no universal recipe, how do I cook up a fitting reform package?

To extend this culinary analogy a bit: I like using a cooking app. I can put in specific ingredients that are still in my fridge at the end of the week and then I get a list of recipes which use these ingredients. The recipes are uploaded and rated by others, with photos, comments, and alternative suggestions. We need this type of experience exchange in corruption control, as well as relevant documentation within development cooperation, but also dialogue with academia. Allow academia to glance over your shoulder while you’re cooking, perhaps even engage in joint experiments! At U4 we are planning to conduct field experiments in selected projects in the coming years together with our partners.

We have identified a need for iterative, problem-based approaches, which take an experimental approach from the outset and have a special focus on learning.

This means that failure must be allowed and the feedback loops must be knitted more tightly than for the usual project reporting. Technical cooperation is particularly well suited to this as it is so close to the processes on the ground. Invest in data collection with your local counterparts, in developing indicators (not the CPI, or at least not only!) and document, what works and what doesn’t work and under what conditions. Take the time to exchange experiences with colleagues in development cooperation, but also international colleague, partners and academics.


“This post is based on a short presentation which I gave in Berlin on 20.3.2017 at a meeting of development counselors at the German Federal Ministry for Economic cooperation and Development on the subject of “Reducing corruption: the potential for transparency and new forms of involvement”. When I use ‘you’ in the following, I am addressing staff from international development cooperation organisations.”