On 1 September 2023, in the concluding session of a global conference at the UN building in Vienna, I had the honour of launching the Vienna Principles Towards a Global Framework for the Measurement of Corruption. The outcome of deliberations among four international organisations, the fifteen principles offer guidance to the many diverse teams that develop and use indicators of corruption nowadays.

The Vienna Principles were arrived at through discussions during the global conference among the four co-organisers: the United Nations Office on Drugs and Crime (UNODC), the United Nations Development Programme (UNDP), the Organisation for Economic Cooperation and Development (OECD), and the International Anti-Corruption Academy (IACA). As Head of the Global Programme on Measuring Corruption at IACA at the time, I held the pen as we debated what should be included, a process that saw plenty of ink spilled and many versions revised before we reached agreement.

Why do we need principles?

Corruption measurement has really taken off in the past few decades. Before the 1990s, the main demand for corruption indicators was from foreign investors assessing the different types of political risk that might jeopardise deals or undermine operations. Then Transparency International (TI) created the first dedicated measure in 1994: the Corruption Perceptions Index. The international NGO’s annual release of country scores and rankings quickly became a powerful advocacy tool, persuading governments and international organisations to take corruption seriously.

Measurement works as an advocacy tool because a bad score can harm a country’s reputation and even have a material impact – whether through credit ratings being downgraded, foreign investment going elsewhere, or voters calling for change.

But as the field grew more specialised and technical, it became clear that the ‘one country, one score’ approach does not meet the needs of the anti-corruption community. Experts have come to recognise that corruption needs to be understood in its context and tackled with tools that are sector specific or harness local social norms. It makes much less sense for corruption indicators to persist in assuming a ‘one size fits all’ approach. That is why principle 2 on ‘comprehensiveness’ suggests that measurement frameworks should ‘seek to take account of the differences in political, legal and institutional structures’ and why principle 4 emphasises the importance of being transparent about methodology so as to promote the informed use of data.

Against this backdrop, both the demand and the supply of corruption indicators have rocketed. A whole host of anti-corruption actors in government, the private sector, and civil society demand measurement instruments that are more precise and serve varied purposes: indicators to diagnose the nature of the problem, indicators to monitor how it changes over time, and indicators to assess whether policy is working. There is demand for measures that are sensitive to the characteristics of different business sectors, localities or government functions, and which can adapt to the constant shifts in corrupt behaviours. This is why principle 3 on ‘relevance and usefulness’ calls for methodologies to be ‘designed with the needs of users in mind’ but also why principle 6 recommends ‘prioritisation’, so that measurement efforts focus on the areas where corruption risk is greatest and most harmful.

The supply of corruption indicators has also grown. The open government movement has meant that much more data about government functions is published, while technology has made it easier to automate data collection and analysis. We’ve seen massive innovation in methodologies for turning data into interactive analytical tools. In not much more than a decade, the use of big data analytics to build corruption risk indicators for public procurement, for example, has gone from an emerging frontier of research to a mainstream anti-corruption tool that many governments and the World Bank use and champion.

A crowded field

The corruption measurement space has become crowded. But far from all these users coalescing around one favoured methodology, the trend is towards different actors developing their own tools to suit their specific needs. Indonesia’s Corruption Eradication Commission (KPK) uses a survey to monitor integrity levels in public institutions. Brazil’s Office of the Comptroller General (CGU) employs artificial intelligence and machine learning to analyse complaints and better target audit resources in the public sector. The Italian National Anti-corruption Authority (ANAC) runs an open-source platform that integrates procurement risk data with socioeconomic contextual risk factors (eg, relating to education, criminality, and social capital) at subnational level.

Some innovations still aim for country comparisons, but tackle difficult aspects of the problem where evidence is still greatly needed, like grand corruption. Daniel Kaufmann has published a new state capture index that reveals a very different pattern across countries compared to traditional measures of petty corruption, and the Center for the Study of Democracy has trialled its state capture diagnostic in several European countries. The World Bank has launched an Anticorruption Lab focused on using new data sources particularly to monitor corruption risks and quantify illicit financial flows, complementing the conceptual framework developed and piloted by UNODC. And there is recognition of the need for better research to assess the effectiveness of anti-corruption agencies and programming.

Other contributions are regional rather than global in scope, like the Western Hemisphere Anti-Corruption Index (WHACI) which focuses on implementation of the 1996 Organization of the American States (OAS) Anti-Corruption Convention, or the European Public Accountability Mechanisms Index (EuroPAM) on anti-corruption regulations in Europe. Another recent initiative collates many sources of subnational data from around the world into one database.

Principles to inform practice

It is fantastic to see one thousand flowers bloom in this space. The aim is that the principles support the field in a number of ways. For developers, they offer a checklist of things to consider during the design of methodological frameworks and principles on ethical data collection, processing, and use. There are also specific recommendations on different data collection methods. For users, a key challenge is assessing the credibility of a particular indicator. Here, the principles provide a starting point for making this judgement. And for those whose anti-corruption work is assessed by indicators, the principles can help with interpreting a score and thus gaining insights to inform future efforts.

We do not yet know if the principles are informing practice, and they will almost certainly need updating over time. But ultimately they may provide some practical help for those navigating an exciting, dynamic and influential part of the anti-corruption field.